John S. Gulliver (Auth.), John S. Gulliver (Eds.) - Transport and Fate of Chemicals in The Environment - Selected Entries From The Encyclopedia of Sustainability Science and Technology-Springer-Verlag

Transport and Fate of Chemicals
in the Environment
This volume collects selected topical entries from the Encyclopedia of Sustainability Science
and Technology (ESST). ESST addresses the grand challenges for science and engineering
today. It provides unprecedented, peer-reviewed coverage of sustainability science and
technology with contributions from nearly 1,000 of the world’s leading scientists and
engineers, who write on more than 600 separate topics in 38 sections. ESST establishes a
foundation for the research, engineering, and economics supporting the many sustainability
and policy evaluations being performed in institutions worldwide.
Editor-in-Chief
ROBERT A. MEYERS, RAMTECH LIMITED, Larkspur, CA, USA
Editorial Board
RITA R. COLWELL, Distinguished University Professor, Center for Bioinformatics and
Computational Biology, University of Maryland, College Park, MD, USA
ANDREAS FISCHLIN, Terrestrial Systems Ecology, ETH-Zentrum, Zürich, Switzerland
DONALD A. GLASER, Glaser Lab, University of California, Berkeley, Department of
Molecular & Cell Biology, Berkeley, CA, USA
TIMOTHY L. KILLEEN, National Science Foundation, Arlington, VA, USA
HAROLD W. KROTO, Francis Eppes Professor of Chemistry, Department of Chemistry
and Biochemistry, The Florida State University, Tallahassee, FL, USA
AMORY B. LOVINS, Chairman & Chief Scientist, Rocky Mountain Institute,
Snowmass, USA
LORD ROBERT MAY, Department of Zoology, University of Oxford, Oxford, OX1
3PS, UK
DANIEL L. MCFADDEN, Director of Econometrics Laboratory, University of California,
Berkeley, CA, USA
THOMAS C. SCHELLING, 3105 Tydings Hall, Department of Economics, University of
Maryland, College Park, MD, USA
CHARLES H. TOWNES, 557 Birge, University of California, Berkeley, CA, USA
EMILIO AMBASZ, Emilio Ambasz & Associates, Inc., New York, NY, USA
CLARE BRADSHAW, Department of Systems Ecology, Stockholm University,
Stockholm, Sweden
TERRY COFFELT, Research Geneticist, Arid Land Agricultural Research Center,
Maricopa, AZ, USA
MEHRDAD EHSANI, Department of Electrical & Computer Engineering, Texas A&M
University, College Station, TX, USA
ALI EMADI, Electrical and Computer Engineering Department, Illinois Institute of
Technology, Chicago, IL, USA
CHARLES A. S. HALL, College of Environmental Science & Forestry, State University
of New York, Syracuse, NY, USA
RIK LEEMANS, Environmental Systems Analysis Group, Wageningen University,
Wageningen, The Netherlands
KEITH LOVEGROVE, Department of Engineering (Bldg 32), The Australian National
University, Canberra, Australia
TIMOTHY D. SEARCHINGER, Woodrow Wilson School, Princeton University,
Princeton, NJ, USA
John S. Gulliver
Editor
Transport and Fate

of Chemicals
in the Environment
Selected Entries from the Encyclopedia
of Sustainability Science and Technology
Editor
John S. Gulliver
Department of Civil Engineering
University of Minnesota
Minneapolis, MN 55455, USA
This book consists of selections from the Encyclopedia of Sustainability Science and Technology edited
by Robert A. Meyers, originally published by Springer Science+Business Media New York in 2012.
ISBN 978-1-4614-5730-5 ISBN 978-1-4614-5731-2 (eBook)

DOI 10.1007/978-1-4614-5731-2
Springer New York Heidelberg Dordrecht London
Library of Congress Control Number: 2012953699
# Springer Science+Business Media New York 2012

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or
information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed. Exempted from this legal reservation are brief excerpts
in connection with reviews or scholarly analysis or material supplied specifically for the purpose of being
entered and executed on a computer system, for exclusive use by the purchaser of the work. Duplication
of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the
Publisher’s location, in its current version, and permission for use must always be obtained from
Springer. Permissions for use may be obtained through RightsLink at the Copyright Clearance Center.
Violations are liable to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of
publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for
any errors or omissions that may be made. The publisher makes no warranty, express or implied, with
respect to the material contained herein.
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)

Contents
1 Transport and Fate of Chemicals in the Environment,

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
John S. Gulliver
2 Chemicals in the Environment, Diffusive Transport . . . . . . . . . . . . 13
Edward Cussler
3 Toxic Organic Chemicals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Simanga Gama, Jon A. Arnot, and Don Mackay
4 Transport in the Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
John S. Gulliver
5 Chemicals in the Environment, Turbulent Transport . . . . . . . . . . . 89
John S. Gulliver
6 Chemicals in the Environment, Dispersive Transport . . . . . . . . . . . 113
John S. Gulliver
7 Transport with Jets and Plumes of Chemicals
in the Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Wenming Zhang, Nallamuthu Rajaratnam, and David Z. Zhu
8 Atmosphere-Water Exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Bernd Jähne
9 Sediment–Water Interfaces, Chemical Flux at . . . . . . . . . . . . . . . . 195
Louis J. Thibodeaux and Joseph Germano
10 River Fate and Transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Zhen-Gang Ji
11 Lake and Reservoir Fate and Transport of Chemicals . . . . . . . . . . 241
Heinz G. Stefan, Xing Fang, and John S. Gulliver
v
vi Contents
12 Oceanic Fate and Transport of Chemicals . . . . . . . . . . . . . . . . . . . 287

Robert P. Mason
13 Subsurface Fate and Transport of Chemicals . . . . . . . . . . . . . . . . . 335
Frank T. Barranco Jr., Samantha L. Saalfield,
Frederick J. Tenbus, and Brian P. Shedd
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
Chapter 1
Transport and Fate of Chemicals
in the Environment, Introduction
John S. Gulliver
Glossary
Concentration The quantity of a compound or chemical per unit volume, unit

mass, or unit mole, where 1 mole = 6.02 1023 molecules of the
chemical or compound. In this text, concentration in mass or moles
per volume of water, mass per mass of solid, and moles per mole of
gas will be discussed, depending upon the media of interest.
Convection The movement of a constituent with the movement of the fluid.
Density Total mass per unit volume.
Diffusion The spreading of fluid constituents through the motion inherent to
atoms and molecules.
Diffusion A coefficient that describes the tendency of molecules to spread
coefficient a constituent mass.
Dilution The mixing of a more concentrated solution with one that is less
concentrated. The adage “The solution to pollution is dilution” is
still used, sometimes appropriately, for many pollution and miti-
gation processes.
Kinematic The fluid viscosity divided by the fluid density, resulting in units
viscosity that are similar to a diffusion coefficient, or length squared per time.
Turbulent The mixing of chemicals by turbulence, such that a turbulent diffu-
diffusion sion coefficient can be defined separately from the temporal mean
convection.
This chapter was originally published as part of the Encyclopedia of Sustainability Science
and Technology edited by Robert A. Meyers. DOI:10.1007/978-1-4419-0851-3
J.S. Gulliver (*)
St. Anthony Falls Laboratory, Department of Civil Engineering, University of Minnesota,
500 Pillsbury Drive S.E, 55454 Minneapolis, MN, USA
e-mail: [email protected]
J.S. Gulliver (ed.), Transport and Fate of Chemicals in the Environment: 1

Selected Entries from the Encyclopedia of Sustainability Science and Technology,
DOI 10.1007/978-1-4614-5731-2_1, # Springer Science+Business Media New York 2012
2 J.S. Gulliver
Introduction
Estimating the fate and transport of chemicals released into the environment is an
interesting and challenging task. The environment can rarely be approximated as
well mixed, and the chemicals in the environment often are not close to equilibrium.
Thus, chemical fate and transport in the environment requires a background in the
physics of fluid flow and transport, chemical thermodynamics, chemical kinetics,
and the biology that interacts with all of these processes. The goal is to follow
chemicals as they move, diffuse, and disperse through the environment. These
chemicals will inevitably react to form other chemicals, in a manner that
approaches, but rarely achieves, a local equilibrium. Many times, these reactions
are biologically mediated, with a rate of reaction that more closely relates to an
organism being hungry, or not hungry, than to first-order kinetics.
The global environment is large, on the chemical fate and transport scale.
Individuals attempt to apply the mathematics of diffusion to the atmosphere,
lakes, rivers, groundwater, and the ocean, depending on the system for which
the material is most applicable, and to transfer between these systems. Volatilization
of a compound from a water body, condensation of a compound from the air, and
adsorption of a compound from a fluid onto a solid are all interfacial transport
processes. Thus, the fate and transport of chemicals in the environmental media of
earth, water, and the atmosphere will be the topic. Contributions in this section
will first attempt to formulate fate and transport problems such that they can be solved,
regardless of the media or of the transport process, through the mathematics of
diffusion (see chapters on Chemicals in the Environment, Diffusive Transport;
Chemicals in the Environment, Dispersive Transport; Chemicals in the Environment,
Turbulent Transport; Toxic Organic Chemicals; Transport in the Environment).
Applications of these principals to specific media and between media will then
be described (see chapters on Transport with Jets and Plumes of Chemicals in the
Environment; Atmosphere-Water Exchange; Sediment–Water Interfaces, Chemical
Flux at; River Fate and Transport; Lake and Reservoir Fate and Transport of
Chemicals; Oceanic Fate and Transport of Chemicals; and Subsurface Fate and
Transport of Chemicals).
Transport Processes
A transport process, as used herein, is one that moves chemicals and other
properties of the fluid through the environment. Diffusion of chemicals is one
transport process, which is always present. It is a spreading process, which cannot
be reversed (without the involvement of another media such as in reverse osmosis).
Convection or advection is the transport of chemicals from one place to another by
fluid flow. The convection and diffusion of a chemical cloud, as represented in
Fig. 1.1, is the movement of the cloud and spreading of the cloud over time.
1 Transport and Fate of Chemicals in the Environment, Introduction 3
Fig. 1.1 Illustration Diffusion

of convection and diffusion t=0
of a chemical cloud along Convection
the x-space coordinate
(x-axis) [2] t=T
x=0 x x=X
Fig. 1.2 Illustration

of longitudinal dispersion
t=0 t=T
of a tracer “plane” at t = 0
to a dispersed “cloud” at
t = T. is the cross-sectional 0 X
mean concentration [2] x
t=T
C
t=0
0 X
x
Turbulent diffusion is actually a form of advection, but the turbulent eddies tend to
mix fluid with a random characteristic similar to that of the diffusion process, when
viewed from enough distance. The representation given in Fig. 1.1 could also be
used to represent convection and turbulent diffusion, except that the pace of turbu-
lent diffusion is normally more than one order of magnitude greater than diffusion.
This higher pace of turbulent diffusion means that diffusion and turbulent diffusion
do not normally need to be considered together because they can be seen as parallel
rate processes, and one has a much different time and distance scale than the other. If
two parallel processes occur simultaneously, and one is much faster than the other,
the second process can normally be ignored. This is discussed further in this section.
Dispersion is the combination of a nonuniform velocity profile and either diffusion or
turbulent diffusion to spread the chemical longitudinally or laterally. Dispersion is
something very different than either diffusion or turbulent diffusion because the
velocity profile must be nonuniform for dispersion to occur. The longitudinal disper-
sion of a pipe flow is illustrated in Fig. 1.2. While there is diffusion of the chemical,
the nonuniform velocity profile creates a dispersion that is much greater than would
occur with diffusion alone. The other important difference is that dispersion reflects
the spreading of a cross-sectional mean concentration, while diffusion represents the
spreading of a local concentration. In some contexts, typically in atmospheric
applications, turbulent diffusion is also considered to be a form of dispersion. This is
only a semantic difference, and herein, the differentiation will continue to be between
turbulent diffusion and the dispersion of a mean concentration.
4 J.S. Gulliver
Liquid and air bubbles are

both transported in a
sparging (aeration)
Air process.
Compressor Air is transported along with

the rainfall in a convective
rain cloud.
Oil on water
flowing into a
storm sewer ends
up as small oil
drops in water.
Fig. 1.3 Illustration of sparged multiphase transport. In these two cases, air bubbles create a water
flow, and rain drops create an airflow. The oil drops and rain drops create an airflow. The oil drops
do not have a significant rise or fall velocity in water and are simply transported [2]
Interfacial transfer is the transport of a chemical across an interface. The most studied
form of interfacial transfer is absorption and volatilization, or condensation and
evaporation, which is the transport of a chemical across the air–water interface.
Another form of interfacial transfer would be adsorption and desorption, generally
from water or air to the surface of a particle of soil, sediment, or dust. Illustration of
both of these forms of interfacial transfer will be given later in this section.
Finally, there is multiphase transport, which is the transport of more than one
phase, usually partially mixed in some fashion. The settling of particles in water or
air, the fall of drops, and the rise of bubbles in water are all examples of multiphase
transport. Figure 1.3 illustrates three flow fields that represent multiphase transport.
Mass transport problems are solved with the diffusion equation, often
represented as
2
@C @C @C @C @ C @2C @2C @2C
þu þv þw ¼D þ 2 þ 2 þ 2 þS
@t @x @y @z @x 2 @y @y @z (1.1)
1j 2 !j j 3 ! j 4
where C is the concentration of a chemical; t is time; u, v, and w represent the temporal

mean velocity in the x, y, and z directions, respectively; and D represents a diffusion
coefficient. The first term (1) on the far left of Eq. 1.1 represents the rate of accumula-
tion of chemical concentration. The second terms (2) represent the mean convection of
the chemical. The third terms (3) to the right of the equal sign represent either diffusion
or turbulent diffusion of chemical. The fourth term (4) represents the multitude of
reactions that are possible in a fluid in environmental media.
Table 1.1 Examples of short-term and long-term fate

Chemical Media Short-term fate Long-term fate
PCB Soil and water
Adsorbed to soil Bioremediated degradation
PCB Atmosphere Adsorbed to aerosols Photocatalyzed degradation
CO2 Water Reactions to carbonate Photosynthesis to oxygen
and bicarbonate and biomass
Benzene Water Adsorbed to suspended particles Bioremediated degradation
Ammonia Soil and water Reaction to ammonium Bioremediated degradation to N2
Chemical Fate
Chemical fate is the eventual short-term or long-term disposition of chemicals,

usually to another chemical or storage. Some examples that fit the concept of
short-term and long-term fate are given in Table 1.1. If a polychlorinated biphenyl
(PCB) compound is in groundwater, the media is soil and water. The “short-term”
fate will be that the PCB will primarily adsorb to the soil. The “long-term” fate is
that the chemical will desorb, when the PCB-laden water has left, and eventually
be bioremediated by microbacteria looking for carbon sources. If this PCB is in
the atmosphere, it will be adsorbed primarily to aerosols and particles in the short-
term, while its long-term fate will probably be photocatalyzed degradation.
There are as many examples of short-term and long-term fate as there are
chemical–media combinations. An important consideration for this topic is whether
one is interested in short-term or long-term fate. This is often a question to be
answered by toxicologists because it is the most toxic forms of a chemical that are
of most interest.
The Importance of Mixing
Mixing is a rate-related parameter in that most rates of reaction or transport are

dependent upon mixing in environmental systems. When mixing is dominant
(the slowest process), the first-order rate equation can be described as
Rate of process = Mixing parameter

Difference from equilibrium (1.2)
Thus, two items are needed to compute the rate of the process: the equilibrium
concentrations for all species involved and the mixing rate parameter. A common
example would be dissolved oxygen concentration in aquatic ecosystems.
One of the most common chemicals of concern in water bodies is oxygen.
Without sufficient oxygen, the biota would be changed because the “desirable”
organisms in the water body require oxygen to live. The rate of oxygen transfer
6 J.S. Gulliver
between the atmosphere and a water body is therefore important to the health of the
aquatic biota. For air-water oxygen transfer, Eq. 1.2 can be formulated as

dM Ca
¼ KL A C (1.3)
dt H
where dM/dt is the rate of mass transfer into the water, KL is a bulk oxygen
transfer coefficient, A is the surface area for transfer, Ca is the concentration of
oxygen in the air, H is a coefficient that partitions oxygen between the air and
water at equilibrium (called Henry’s Law constant for liquids and gas equilib-
rium), and C is the concentration of oxygen in the water. Air is 20.8% oxygen, so
the concentration of oxygen in the atmosphere is determined primarily by atmo-
spheric pressure. Henry’s Law constant for oxygen is a function of pressure as
well as temperature. Thus, the equilibrium concentration of oxygen is influenced
by the thermodynamic variables pressure and temperature. The rate parameter is
KL A, which has units of volume/second. The difference from equilibrium
partitioning is represented by Ca/H – C. It is C that is typically needed to bring
as close to equilibrium with the atmosphere as possible, and the means to do it is
by having a large dM/dt. This usually means a large KLA because it would be
difficult to alter either Ca or H in the atmosphere. While the surface area is often
established by the boundary conditions, KL is determined by the turbulence and
diffusion coefficient (i.e., mixing) close to the water surface and represents the
rate of mixing per unit surface area. Thus, the primary variable that can be
changed in order to increase dM/dt is the mixing parameter represented by KL.
Some further examples of mixing rate and equilibrium parameters in environmen-
tal processes are given in Table 1.2.
Resistance to Transport
An important concept for environmental transport is resistance. That is, the inverse
of a rate parameter is a resistance to chemical transport, or in equation form:
1
¼ Resistance to chemical transport = R (1.4)
Rate parameter
Figure 1.4 gives an example of the adsorption of a compound to suspended

sediment, modeled as two resistances in series. At first, the compound is dissolved
in water. For successful adsorption, the compound must be transported to the
sorption sites on the surface of the sediment. The inverse of this transport rate can
also be considered as a resistance to transport, R1. Then, the compound, upon
reaching the surface of the suspended sediment, must find a sorption site for
adsorption. This second rate parameter is more related to surface chemistry than
Table 1.2 Examples of important mixing rate and equilibrium parameters in environmental processes
Process Mixing-rate-related parameter Equilibrium parameter
1. Treatment processes
Coagulation/flocculation Size of coagulation and flocculation basins Dose of coagulants (alum)
and proper mixing (residence time)
Softening Design of softening tank to increase mixing Dose of softening agent (lime)
Settling Design of settling basin to reduce mixing !0
Chlorination Design of chlorination and dechlorination chambers Dose of chlorine
for proper mixing and res. time
Filtration Size of filter bed Length of time before backflushing
2. Surface waters
Oxygen transfer Diffusion and turbulent mixing Atmospheric O2 conc.
Henry’s Law constant
Volatilization of pollutants Diffusion and turbulent mixing !0
Toxic spills Diffusion and turbulent mixing Spill–water equilibrium
Internal loading of nutrients Hypolimnetic mixing O2 concentration in hypolimnion
Sorption onto suspended sediments Turbulent mixing exposes chemicals to sediment Sediment–water partitioning
3. Atmosphere
Greenhouse gases (CO2, CH4) Turbulent mixing Atmospheric concentrations
1 Transport and Fate of Chemicals in the Environment, Introduction
Volatilization of spills Turbulent mixing and dispersion !0

Aerosols Turbulent mixing, dispersion, and settling None
4. Groundwater and sediments
Spills Advection and dispersion !0
O2 Diffusion and advection Atmospheric conc. of O2
Sorption of soil Advection and diffusion Equilibrium soil–water partitioning
7
8 J.S. Gulliver
Fig. 1.4 Adsorption analogy Compound R1 R2 Sorbed

to two resistors in a series: in water compound
adsorption of an organic compound compound on sediment
compound to sediment [2] “approaches” “decides” to
sorption site adsorb
on sediment
surface
to diffusive transport, and is considered as a second resistance, R2, that acts in

series to the first resistance. The second resistance cannot occur without crossing
the first resistance of transport to the sorption site, so they must occur in series.
Now, if R1 is much greater than R2, it can be assumed that R2 is zero without
compromising the accuracy of the rate calculation. In electric circuits, two
resistances applied in series are simply added together in calculating the line
resistance. The same is true for resistance to chemical transport. If R1 is 1,000
resistance units and R2 is 1 resistance unit, R2 can be ignored and still be within
99.9% of the correct answer.
Another example is the air–water transfer of a compound, illustrated in Fig. 1.5.
This example will be used to explain volatile and nonvolatile compounds. There is
resistance to transport on both sides of the interface, regardless of whether the
compound is classified as volatile or nonvolatile. The resistance to transport in
the liquid phase is given as RL = 1/KL. If chemical transfer is being described
through an equation like Eq. 1.3, the resistance to transfer in the gas phase is given
as RG = 1/(HKG). The equilibrium constant is in the RG equation because the
equivalent waterside concentrations are being used to represent the concentration
difference from equilibrium, and the gas phase resistance needs to be a resistance to
an equivalent water concentration.
The gas phase and the liquid phase resistances are applied in series. In general,
gas film coefficients are roughly two orders of magnitude greater than liquid film
coefficients. It is also true that Henry’s Law constant, H, varies over many orders of
magnitude as the transported compounds are varied. Nitrogen gas, for example, has
a Henry’s Law constant of approximately 15, using mass concentrations. The
herbicide, atrazine, has a Henry’s Law constant of 310–6. Thus, the ratio RG/RL
would vary by seven orders of magnitude between nitrogen gas and atrazine.
If these orders of magnitude are put into a series resistance equation,
1 1
R ¼ RL þ R G ¼ þ (1.5)
KL HKG
Due to the Henry’s Law constants, it can be seen that for nitrogen gas, R ffi RL,
and for atrazine, R ffi RG. If a typical ratio of KG/KL 100 is applied, RG = RL when
H = 0.01.
Now, the mass transfer between phases is given as

dM A Ca
¼ C (1.6)
dt R H
Fig. 1.5 Air–water transfer Volatile Nonvolatile

analogy to two resistors in compounds compounds
a series [2]
RG RG
RL RL
Volatile compounds: Nonvolatile compounds:

RL >> RG (HKG >> KL) RG >> RL (KL >> HKG)
Liquid side resistance Gas side resistance
dominates dominates
or

dM A Ca
¼ C (1.7)
dt 1=KL þ 1=HKG H
Nitrogen gas would be a volatile compound because the equilibrium is strongly

to the gas phase, and there is little gas phase resistance to its transfer, i.e., 1/KL
1/(HKG). For that reason, N2 is generally called a gas, as are many other volatile
compounds such as methane, oxygen, and propane.
Atrazine, on the other hand, would be a nonvolatile compound, 1/(HKG) 1/KL,
because equilibrium is strongly to the liquid phase due to the small Henry’s Law
constant. There is also a strong gas phase resistance to the transfer. Atrazine was
manufactured to remain in the liquid phase, where it will act as a herbicide, rather
than in the gas phase, where the farm personnel will be breathing this toxic
chemical. If you were going to pick a nonionic compound that is not made by
humans from the list of those that are gas or liquid in our environment, a good guess
is that it would be a volatile or semivolatile compound. There are only a few
environmental compounds that are nonvolatile. Remarkably, one of them is
water. While the atmosphere may be as much as 3% water, the water bodies in
the world are very close to 100% water. The equilibrium is strongly to the liquid
side, due to a large equilibrium-partitioning constant.
One theme of this discussion can now be stated as follows: when transport
processes occur in series, it is the slower transport processes that are important
10 J.S. Gulliver
Fig. 1.6 Transport to Primary

a sorption site and the resistor transport
analogy [2] RS path
sorption
Turbulent diffusion, Compound
Solid
Diffusion, RT in solution
RD2
Diffusion, RD1
1 ∼ R +R +R
R= + RD2 + RS = T D2 S
1/RT+1/RD1
for chemical transport calculations because the resistance to transport is large, just
as the large resistors of a series in an electronic circuit are the most important.
Now is the time for the second theme: when transport processes occur in
parallel, the fast transport process with the low resistance dominates. The result
is the opposite of resistances in series. Figure 1.6 illustrates this concept with the
transport of a compound from a water body to a sorption site on a solid. In the bulk
solution, there is diffusion and turbulent diffusion occurring simultaneously. Trans-
port can occur due to either process, so there are two different paths that may be
followed, without the need of the other path. These transport processes are
operating in parallel, and the faster transport process will transport most of the
compound. The analogy to electronic circuits applies in this case as well. Beginning
with a compound in solution in Fig. 1.6, there are two parallel transport paths,
each with a resistance to transfer. Most of the compound will be transported
through the path with the least resistance. Many times, the path with the greater
resistance can be ignored because the quantity of compound transported through
this path is very small. When the compound comes close to the solid, however,
the turbulent diffusion dissipates because eddies become so small that they are
dissipated by the viscous action of the water. Now, one is back to one transport
path, with the act of sorption and diffusion acting in series. Thus, the slowest
transport path once again becomes the important process.
The overall resistance to the sorption process illustrated in Fig. 1.6 can be written
as follows:
1
R¼ þ RD2 þ RS ffi RT þ RD2 þ RS (1.8)
1=RT þ1=RD1
where RT, RD1, RD2, and RS are the resistances to turbulent transport, diffusive
transport in the bulk of the fluid, diffusive transport near the solid surface, and
adsorption, respectively. It can be seen that, in Fig. 1.6 and Eq. 1.8, the resistance
due to diffusion in the bulk of the fluid can be neglected because turbulent diffusion
is a parallel path. The resistance due to diffusion only needs to be considered when
there is no parallel path for turbulent diffusion, such as very near the surface of the
solid. Thus, RD1 can be ignored, but not RD2.
Conclusion
In this chapter, some of the topics that will be covered and applied in the other
chapters of the section have been introduced, where the physics of mass transport
are essential. These and similar engineering concepts will be revisited throughout
the first half of this volume, in an attempt to develop models in the environmental
fate and transport of chemicals that are close to realistic, but can be solved, even if
that solution is approximate.
Future Directions
Global climate change is a topic for the present and future because society has only
begun to assess and tackle the causes, implications, and remediation of the new
Anthropocene [1]. There are global warming gases (CO2 and CH4) and global
cooling gases (refrigerants) that are transported between the oceans and the atmo-
sphere and could be transported between the atmosphere and the earth (carbon
sequestration). These global transport rates are not quick, so we humans will be
dealing with transport to a greater extent in the future. On a more local scale,
urbanization is creating nonsource pollution of water and the atmosphere that is
increasing. Roads, parking lots, and rooftops are often directly connected to lakes
and rivers through storm sewers, and the pollution of urban activities is not
insubstantial. This will be a bigger concern as discovery about this source of
pollution occurs. Finally, the nitrogen of various forms released from agricultural
activities continues to create hypoxic (dead) zones when the rivers have transported
this nutrient to the oceans. The nitrification–denitrification processes as organic
nitrogen is decomposed to ammonia, utilized by aerobic bacteria to make nitrites
and nitrates, and finally converted by anaerobic bacteria to nitrogen gas occurring in
streams and rivers add an interesting complexity to the fate and transport
calculations.
Bibliography
1. Ellis E (2011) A man-made world. The Economist London, UK, 26 May 2011
2. Gulliver JS (2007) Introduction to chemical transport in the environment. Cambridge University
Press, Cambridge, UK
Chapter 2
Chemicals in the Environment,
Diffusive Transport
Edward Cussler
Glossary
Convection Mass transfer effected by flow due to applied forces

like pressure (forced convection) or to density
differences (free convection).
Diffusion Mixing caused by molecular motion.
Diffusion coefficient The negative of the flux per concentration gradient.
Diffusivity Another name for the diffusion coefficient.
Dispersion Mixing caused by diffusion and simultaneous flow.
Flux Mass or moles transferred per area per time.
Mass transfer Diffusion and dispersion, especially across interfaces.
Mass transfer coefficient The flux per concentration difference, especially near
an interface.
Overall mass transfer The flux per virtual concentration difference from one
coefficient phase across an interface into a second phase.
E. Cussler (*)
Department of Chemical Engineering, University of Minnesota, 55455 Minneapolis, MN, USA
(612)-625-1596(612)-626-7246
e-mail: [email protected]; [email protected]

14 E. Cussler
Definition of the Subject
Diffusion is mixing without stirring. It is mixing caused by Brownian motion, that

is, by thermally induced random motion of molecules or small particles. Because
diffusion is often slow, it frequently limits the overall rate of the process. Diffusion
has the reputation of being a difficult subject, which it can be; however, the
difficulty most often comes from complicated units, from interfaces, or from the
combination of diffusion and convection. By itself, diffusion is not hard. It is easier
than viscous flow and much easier than ideas like entropy or chemical potential.
Two other phenomena, closely related to diffusion, are also reviewed in this
entry. Dispersion is mixing caused by the interaction of flow and diffusion. Often, it
is described using mathematics similar to those which describe diffusion. In
environmental problems, these two phenomena are sometimes treated without
distinction and without penalty. Mass transfer, an alternative description of diffu-
sion, assumes that all concentration changes occur near interfaces. While it is used
largely to describe chemical processing, it has considerable value in environmental
problems.
Introduction
This entry is organized as four sections. The first section gives the mathematical
description of diffusion itself. The second reviews dispersion, a different phenome-
non that is mathematically similar to diffusion but which is caused by different
physical effects. The third section, “Diffusion Coefficients,” reviews values of the
diffusion coefficients themselves; and the fourth section “Diffusion Across
Interfaces,” explains mass transfer across interfaces, especially air–water interfaces.
Finally, the section “Important Special Cases” reports some common situations
which have important features.
Basic Diffusion
The basic mathematical description of diffusion is Fick’s Law, suggested by

Adolph Fick (1829–1910) when he was just 26 years old. For a dilute solution,
Fick’s Law is:
dc1
j1 ¼ D (2.1)
dz
where ji is the flux of a solute “1,” and dc1/dz is the concentration gradient, that is,
the change of the solute concentration with position. The diffusion coefficient D is
a proportionality constant which is nearly constant in almost all situations. The flux
is the amount of solute moving per cross-sectional area per time and so has the
2 Chemicals in the Environment, Diffusive Transport 15
Fig. 2.1 Concentration

profile across a thin film c10
c1l
z
l
Δz
dimensions of mass (M) per area (length squared L2) per time (t). The concentration
gradient has dimensions of concentration (M/L3) per distance (L). Thus the diffu-
sion coefficient has dimensions of (L2/t).
The form of Fick’s Law in Eq. 2.1 has some hidden implications. First, it is a
one-dimensional equation of what is actually a more general vector relation. Because at
least four out of five diffusion problems are one-dimensional, this is often not a major
issue. Second, the concentration can be expressed in different units. If it were expressed
in moles per volume, then the flux would be in moles per area per time. If it were
expressed as a mole fraction or a mass fraction or a partial pressure, then unit
conversions would be necessary and annoying, but not difficult. Third, the minus
sign in Eq. 2.1 is arbitrary, stuck in to make the diffusion coefficient positive. The
only difficult implication of Eq. 2.1 is the restriction to dilute solutions, explored in
more detail at the end of this section. The restriction is rarely important because
solutions in the environment are so often dilute. For example, liquid water contains
55 mol/l, so almost every aqueous solution is dilute.
The most important case of Fick’s law is diffusion across a thin film, described
next. Other important cases and concentrated solutions are reviewed later.
Diffusion Across a Thin Film
The simplest case, steady diffusion across a film, is also the most important.
Imagine a thin film separating two well-stirred solutions, as shown in Fig. 2.1. On
the left, the solution has a concentration c10; on the right, the concentration is c1‘.
The key parts of this case are the variation of concentration across the film c1(z) and
the flux j1. Finding these requires a mass balance on a differential volume Dz thick
and located at an arbitrary position z within the film:
ðmass accumulationÞ ¼ ðmass diffusing in outÞ

16 E. Cussler
@
ðc1 ADzÞ ¼ ðj1 AÞz ðj1 AÞzþDz (2.2)
@t
where A is the constant cross-sectional area of the film. Because diffusion is steady,
the concentration does not change with time, the left-hand side of Eq. 2.2 is zero,
and
j1 jz j1 jzþDz
0¼
ðz þ DzÞ z
dj1
0¼ (2.3)
dz
This restates the assumption of steady-state diffusion, independent of time.

Combining this relation with Fick’s Law (Eq. 2.1) yields:
d 2 c1
0¼D (2.4)
dz2
This is subject to two boundary conditions:
z¼0 c1 ¼ c10 (2.5)
z ¼ ‘ c1 ¼ c1‘ (2.6)
This is enough to solve this important problem.

Equation 2.4 may be integrated once to find:
dc1
¼A (2.7)
dz
where A is an integration constant. Integrating a second time gives:
c1 ¼ Az þ B (2.8)
where B is a second integration constant. Evaluating A and B from Eqs. 2.5 and 2.6
gives:
c1 c10 z
¼ (2.9)
c1‘ c10 ‘
The flux can now be found by combining this result with Fick’s Law:
dc1
j1 ¼ D
dz (2.10)
D
¼ ðc10 c1‘ Þ
‘
If the concentration difference across the film is doubled, the flux doubles. If the
diffusion coefficient is twice as big, the flux will be twice as big, too. If the film
thickness increases two times, the flux will be cut in half.
This important example is so simple mathematically that many novices tend to
skip over it. This is a mistake. Its nuances are explored by the following questions:
1. How does the flux change if the film is chemically different than the adjacent
solutions?
In this case, the boundary conditions in Eqs. 2.5 and 2.6 change to:
z ¼ 0 c1 ¼ Hc10 (2.11)
z ¼ ‘ c1 ¼ Hc1‘ (2.12)
where H is a partition coefficient, the ratio at equilibrium of the concentration inside

the film to that in the adjacent solution. Paralleling the arguments above,
ðDH Þ
j1 ¼ ðc10 c1‘ Þ (2.13)
‘
The diffusion coefficient D in Eq. 2.10 is replaced with the product (DH), which
is called the permeability. (The term (DH/‘) is called the permeance.) As will be
shown later, diffusion coefficients in gases and liquids do not vary much, but
partition coefficients vary a lot. Thus partition is often the key to permeability.
2. How is the flux changed by a fast reversible reaction giving an immobile
product?
This case occurs surprisingly frequently in, for example, adsorption in soil or
dyeing of wool. The answer is that at steady state, the flux does not change. Every
point crossing the film has a different concentration which is in equilibrium with
a different absorbed amount. Still, at steady state, the reaction is at local equilibrium
and does not affect the flux. This is not the case for unsteady state, for irreversible
reactions, or for mobile reaction products.
Other Important Cases
Many diffusion problems are not thin films. Surprisingly, many do behave as if they
were thin films. For example, for a sphere of radius R slowly dissolving in
a stagnant fluid with a concentration of c11, the flux is:
D
j1 ¼ ðc10 c11 Þ (2.14)
R
18 E. Cussler
Fig. 2.2 Concentration

profile into a semi-infinite
c10 Δz
slab
Time
c1∞
Position z
where c10 is the concentration in solution at the surface of the sphere. For diffusion
from a solution of c10 through a very thin, impermeable film with a cylindrical orifice
of radius R, and into a solution at c11, the flux is:
D
j1 ¼ p ðc10 c11 Þ (2.15)
2R
The fluxes in these cases are strong mathematical parallels to that in Eq. 2.10.
The mathematics in these cases is different, but the final result is remarkably
similar. The thin film limit is a good guide about 80% of the time.
For unsteady state diffusion, this is not true. Fluxes and concentration profiles for
a wide variety of unsteady cases have been calculated and are tabulated in a few clear
texts. One of these cases, useful in perhaps 10% of all cases, is unsteady diffusion into
a semi-infinite slab, shown in Fig. 2.2. In this case, a mass balance gives:
@c1 @ 2 c1
¼D 2 (2.16)
@t @z
where t is the time. For this semi-infinite slab, the initial and boundary conditions
are:
t¼0 all z c1 ¼ c11 (2.17)
t¼0 z¼0 c1 ¼ c10 (2.18)
t ¼ 0 z ¼ 1 c1 ¼ c11 (2.19)
Table 2.1 Flux across a film or into a slab. These two cases are important because they
bracket almost all diffusion problems. In this table, K is the equilibrium constant of the rapid
chemical reaction
Thin film Semi-infinite slab
Concentration difference Dc1 Dc
pffiffiffi1ffi
Diffusion coefficient D D
Thickness ‘1 –
Time – t1
qffiffiffi
Flux without reaction j1 ¼ D‘ Dc1 j1 ¼ pt D
Dc1
qffiffiffiffiffiffiffiffiffiffiffiffi
Flux with fast reversible reaction j1 ¼ D‘ Dc1 j1 ¼ Dð1þK Þ
pt Dc1
The concentration profile in this case is:

c1 c11 z
¼ 1 erf pffiffiffiffiffiffiffiffi (2.20)
c10 c11 4Dt
where erf is the error function. The flux at the edge of the slab, that is, at z = 0, is:
rffiffiffiffiffi!
D
j1 ¼ ðc1 c11 Þ (2.21)
pt
This is the key result for this case.

The cases of a thin film and a semi-infinite slab are especially important because
they bracket observed behavior. More specifically, the result in Eq. 2.21 for a semi-
infinite slab is compared with the flux across a thin film in Table 2.1. In both cases,
the flux will double if the concentration difference doubles. If the diffusion coeffi-
cient doubles, the flux across a thin film doubles, but the flux into the semi-infinite
pffiffiffi
slab increases 2 times. If the film’s thickness doubles, the flux drops two times for
the film, but is unchanged for the slab. If diffusion occurs for twice as long, the flux
for the thin film keeps its steady value, but that for the slab drops by a factor of
pffiffiffi
1= 2 . These limits will usually bracket all diffusion behavior because all shapes
will be between the film and the slab. These two cases are key to understanding the
mathematics of diffusion.
A third special case is especially important for environmental engineering. This
is the decay of a pulse. In this case, a large amount of solute is released at
a particular plane at z = 0. Solute diffuses away from this position in only one
dimension. The solute concentration as a function of position z and time t gives the
details of any environmental impact.
The mathematics follows the same route as the slab: a mass balance is subject to
initial and boundary conditions, which are combined with Fick’s Law and solved to
give the concentration profile. The mass balance is:
@c1 @ 2 c1
¼D 2 (2.22)
@t @z
20 E. Cussler
This mass balance, identical with the mass balance for the slab (Eq. 2.16), occurs
so frequently that some just call it “the diffusion equation.” The initial and
boundary conditions are different from Eqs. 2.17–2.19 for the slab:
M
t¼0 all z c1 ¼ dðzÞ (2.23)
A
dc1
t¼0 z¼0 ¼0 (2.24)
dz
z¼1 c1 ¼ 0 (2.25)
In these conditions, M is the total solute injected, A is the cross-sectional area,

and d(z) is the Dirac function, equal to zero everywhere except at z = 0, where it is
infinity. While the mathematical solution of Eqs. 2.22–2.25 is tricky, the answer is
simple:

M =A z2
c1 ¼ pffiffiffiffiffiffiffiffiffiffi e 4Dt (2.26)
4pDt
In this Gaussian concentration profile, the quantity in square brackets is the

maximum concentration, which drops as time grows. The mathematical form of this
important result is also observed for other environmental problems which do not
depend only on diffusion, as described in the section “Dispersion.”
Concentrated Diffusion
One reason that diffusion has the reputation of being difficult comes from the major
complexities of concentrated solutions. These complexities are rarely important in
environmental engineering and so should be ignored unless there are good experi-
mental reasons not to do so. These complexities are mentioned here only to
illustrate when they are important.
To explore this, imagine putting a pot of room temperature water of 25 C on
a stove that is turned off. The flux of any evaporating water will be given by
Eq. 2.10. If the air above the stove were dry, c1‘ would be zero. Because the water at
25 C has a vapor pressure of about 25 mmHg, c10 is about:

25 mm Hg 1 mol
c10 ¼
750 mm Hg 22:4 103 m3
(2.27)
mol
¼ 1:5 3
m
If the liquid water in the pot is 0.1 m below the rim and the diffusion coefficient
of water vapor in air is about 2.8 105 m2/s, the flux is:
D
j1 ¼ ðc10 c1‘ Þ
‘

2:8 105 m2 /s mol
¼ 1:5 3 0 (2.28)
0:1m m
mol
¼ 4 104 2
m s
Now imagine heating the liquid water in the pot to boiling. If the heat flux q is 20
kJ/m2 s, then the molar flux caused by boiling is:
q
n1 ¼
DH~vap
20 kJ /m2 s
¼ (2.29)
48 kJ/mol
mol
¼ 0:4 2
m s
The boiling flux n1 is 1,000 times greater than the diffusion flux at room tempera-
ture, but it is not a function of D. Thus slow dilute evaporation is a function of the
diffusion coefficient; but boiling depends not on diffusion but on heating rate.
But what about intermediate cases? For example, how fast will evaporation take
place at 50 C?
Answering this question requires a more complete form of Fick’s Law. Unfortu-
nately, there is no single way to do this. One choice is the following:
½Total flux ¼ ½Diffusion flux þ ½Convective flux
n1 ¼ c1 ðv1 vÞ þ c1 v ¼ j1 þ c1 v
dc1 (2.30)
¼ D þ c1 v
dz
where v is most often a volume average velocity. The most common alternative
form of Fick’s law, strongly advocated by a few zealots, may be approximated as:
c1 c2
rc1 ¼ ðv2 v1 Þ (2.31)
cD
22 E. Cussler
where c2 and v2 are the concentration and velocity of the solvent. This form avoids
choosing a convective velocity, but it can cloud the physical significance of the
problem.
Fortunately, the result for a problem like the water evaporation given above is
the same for both forms of Fick’s Law given in Eqs. 2.30 and 2.31. If water at 50 C
is evaporating from the pot above into dry stagnant air, the total flux is:
Dc c10
n1 ¼ ln 1
‘ c
2:8 105 m2 =s 1 mol
22:4103 m3 92:5 (2.32)
¼ ln 1
0:1 m 760
3

¼ 1:62 10 mol m2 s
This is about 7% greater than the result would be if calculated from Eq. 2.10 for
a dilute solution at 50 C. The moral is clear: the effects of concentrated diffusion
will only rarely be important in the atmospheric and aquatic environments.
This completes our basic description of the three cases key to understanding
diffusion. These are the thin film (80% of the cases), the semi-infinite slab (10% of
the cases), and the decay of a pulse (5% of the cases). Other cases with different
boundary conditions do occur, and solutions for these are tabulated in the literature.
However, these other cases are not as common in practice. Similarly, diffusion in
concentrated solutions is complicated but infrequently important. The simple form
of Fick’s Law in these three cases is the best way to get started.
Dispersion
We now turn to an environmentally important problem mathematically similar to

diffusion but with a different physical origin. To make this problem specific,
imagine dealing with the spill of a single toxin on the ground. Imagine the
concentration of the toxin spreads in one dimension with time and groundwater
flow, producing a roughly Gaussian concentration profile. We want to know how
the spread of this toxin varies with the diffusion coefficient of the toxin.
The answer is surprising: if the diffusion coefficient increases, the spread of the
toxin may be bigger, smaller, or unchanged. The toxin’s concentration profile is:
M =A ðzvtÞ2
c1 ¼ pffiffiffiffiffiffiffiffiffiffi e 4Et (2.33)
4pEt
where M is the total amount of toxin, A is the cross-sectional area across which the
dispersion occurs, and v is the velocity of any flow through the soil. The dispersion
coefficient E has the same units as the diffusion coefficient but will often be much
larger. This result, a complete analogue to Eq. 2.26, can be derived from parallels to
Eqs. 2.22–2.25 by replacing the diffusion coefficient D with the dispersion coeffi-
cient E. However, while this mathematical parallel is complete, it does not explain
the physics responsible for dispersion.
To explore the physics involved, imagine the toxin is injected as a pulse into
a small tube of diameter d. The toxin’s dispersion will be a strong function of how
much flow is in the tube. If there is absolutely no flow, then the dispersion coefficient
equals the diffusion coefficient:
E¼D (2.34)
Increasing diffusion increases dispersion. If there is a small, laminar flow of

velocity v, then:
v2 d2
E¼ (2.35)
192D
Increasing diffusion decreases dispersion. The velocity where Eq. 2.35 becomes
dominant is when:
v2 d 2
1 (2.36)
192D2
For non-absorbing soil with the equivalent of 500 mm particles and diffusion in
water liquid, D is about 109 m2/s, so v must be much greater than 30 mm/s or
2 m/day for Eq. 2.35 to swamp Eq. 2.34. If there is a large, turbulent flow (dv/n >
2,000), then:
dv
E¼ (2.37)
2
Dispersion, now independent of diffusion, is due to the coupled turbulent

fluctuations of concentration and velocity. The physical basis of Eqs. 2.33–2.37 is
associated with G.I. Taylor.
Those with a more practical bent may correctly be skeptical of modeling flow
through a soil as occurring in a straight tube. Others sharing this skepticism have
extended this analysis to flow in packed beds. The key results involve two new
quantities:
t ¼ tð1 þ k0 Þ (2.38)

0 Soil concentration 1e
k ¼ (2.39)
Solution concentration e
24 E. Cussler
where e is the void fraction available for flow. In physical terms, t is the time
corrected for any absorption by the soil, including material that diffuses into the
soil’s pores. The quantity k 0 is a type of equilibrium constant between the soil and
the solution, lumping together adsorption and absorption. In this case, a pulse of
toxin may still be dispersed to give the Gaussian concentration profile in Eq. 2.24,
but with time t replaced by t. The dispersion coefficient E is now given by:
!
0 d2 v2 1 þ 6k0 þ 11ðk0 Þ2
E ¼ Dð1 þ k Þ þ
192D 1 þ k0
0 (2.40)
d2 v 2 k
þ
3D 1 þ k0
0
where d is an equivalent thickness of an absorbent and D0 is the diffusion coefficient

in the absorbent, not in the solution. The first term on the right-hand side of
Eq. 2.40, the parallel of Eq. 2.34, is due to diffusion in the direction of flow. The
second term, the analogue of Eq. 2.37, comes from Taylor dispersion and is often
the most important. The third term is new, the result of the rate of absorption.
This overview of dispersion is intended as a caution and a starting point. The
caution is that many Gaussian concentration profiles are due to diffusion coupled
with other phenomena. The starting point in understanding these profiles is
recognizing that their spread can depend inversely on diffusion. In other words,
slow diffusion may result in wide dispersion.
Diffusion Coefficients
Diffusion is an important process because it is slow, and diffusion coefficients thus

often control the overall rate of processes involving diffusion, flow, and chemical
reaction. Typical values of diffusion coefficients, shown in Table 2.2, are chosen
from the wide number of references in the literature but corrected to a temperature
of 25 C. This wide literature is much less extensive than studies of other physical
properties like viscosity or Young’s modulus, because diffusion coefficients are
relatively difficult to measure.
The values in Table 2.2 show diffusion in gases is about 10,000 times faster than
diffusion in liquids, which is in turn over a billion times faster than diffusion in
solids. Diffusion coefficients in gases fall around 105 m2/s. Diffusion coefficients
in liquids fall around 109 m2/s. Diffusion coefficients in solids are much more
variable, but are so slow that most of the mass transport occurs in fluid-filled gaps
and pores within the solid. For example, in a bed of sand, most transport occurs in
the spaces between sand grains and relatively little within the bulk of the grains
themselves. In environmental problems, diffusion in gases and liquids is more
important.
Table 2.2 Diffusion coefficients. Values given are in m2/s and at 298 K and 1 atm
Gases
Gas pair Diffusion coefficient
Air-H2O 2.6 105
CO2-O2 1.6 105
H2-N2 7.8 105
H2-O2 8.9 105
N2-O2 2.2 105
N2-H2O 2.9 105
O2-H2O 2.8 105
O2-octane 0.7 105
Solids
Diffusion coefficient
C in Fe (BCC) 6 1025
Fe in Fe (BCC) 3 1052
B in Si 7 1033
He in SiO2 4 1014
Na+ in NaCl 1 1036
Ag+ in AgCl 1 1019
Liquids
Solute-solvent Diffusion coefficient
O2-H2O 2.10 109
CO2-H2O 1.92 109
H2S-H2O 1.41 109
HCl-H2O 3.33 109
NaCl-H2O 1.61 109
CaCl-H2O 1.33 109
NH3-H2O 1.64 109
Urea-H2O 1.38 109
Sucrose-H2O 0.52 109
Albumin-H2O 0.08 109
H2O-C2H5OH 1.24 109
Benzene-butanol 0.99 109
Hexane-heptane 4.21 109
The diffusion coefficients, given in Table 2.2 for 1 atm and 25 C, do change with
process variables, as outlined in Table 2.3. The variation with temperature in gases
and liquids is small. For example, the temperature must be increased from 25 C
(= 298 K) to 200 C (= 473 K) to double the diffusion coefficient in gases. Because
the viscosity of a liquid drops as the temperature rises, the diffusion coefficient in
a liquid changes faster with temperature, but the change is still modest. In contrast,
diffusion coefficients in solids usually change more rapidly with temperature,
doubling every 10 C or so.
Other process variables also have relatively small effects. The diffusion coeffi-
cient in gases does vary inversely with pressure; but the gas concentration varies
26 E. Cussler
Table 2.3 Variations of diffusion coefficients

Gases Liquids Solids
Typical value, m2/s 105 109 Much smaller
vs. T T 3/2 T Large
vs. p p1 – –
vs. solute diameter size2 size1 size2
vs. viscosity m m1 m1 –
directly with pressure; so the flux, related to the diffusion coefficient times the
concentration, may remain more constant. The diffusion coefficient does vary
inversely with the size of the diffusing species. While these effects are usually
modest for gases and liquids, they can be much larger for solids. These
generalizations are justified by the approximate physical arguments given next.
Gases. The diffusion coefficients in gases can be predicted with reasonable accu-
racy from kinetic theory. This theory assumes that a gas contains individual
molecules moving with thermal motion and colliding with each other only as
pairs. Under these cases, the diffusion coefficient is given by:
1
D ¼ lv (2.41)
3
where l is the distance between collisions and v is the molecular velocity. For
a monatomic gas, this velocity is kinetic, related to the thermal energy:
1 2
mv ¼ kB T (2.42)
2
where m is the molecular mass and kB is Boltzmann’s constant. Keep in mind that
here v is a molecular velocity. It is the sonic velocity; it is much greater than the
average velocity v used in Eqs. 2.30 and 2.31.
We must now estimate the distance between collisions l. There are two cases.
First, for the bulk gas, l is found from the volume occupied by one molecule:
½Volume of one molecule ¼

½Distance between collisions
½Area swept out between collisions
kB T hp i
¼ l s2 (2.43)
p 4
where s is the molecular diameter. Equations 2.41–2.43 can be combined to find:
pffiffiffi
3
4 2 kB T 2
D¼ pffiffiffiffi (2.44)
3p ps2 m
This approximate relation is close to that found from more complex theories:
pffiffiffiffi
D does vary inversely with p, s2, and m; it does vary with T to a power greater
than one and less than two.
The second important limit of Eq. 2.41 occurs only for a gas diffusing in small
pores of diameter d. In this case, the diffusing species is much more likely to collide
with the pore walls than with other molecules. Thus the combination of Eqs. 2.41
and 2.34 becomes:
pffiffiffi rffiffiffiffiffiffiffiffiffi
2 k T
D¼ d B (2.45)
3 m
This case, called Knudsen diffusion, has a diffusion coefficient which depends
on the pore diameter d, but not on the molecular diameter s. Now, the diffusion
coefficient is independent of pressure, though it does vary inversely with the square
root of solute mass. Under ambient temperature and pressure, Knudsen diffusion is
important when the pores are much less than 0.1 mm.
Liquids. Diffusion in liquids is normally not described by a kinetic theory but as the
motion of a rigid, spherical solute diffusing in a continuum of solvent. Despite the
major approximations obviously made by this simple model, it gives remarkably
good results. It is the standard against which new predictions are always judged.
The model begins by describing the friction on a solute sphere:
Force = ½Coefficient of friction f Velocity v1 (2.46)
The velocity v1 now is the average and not the sonic value v used for gases. The
coefficient of friction f is given by Stokes Law:
f ¼ 6pmR (2.47)
where m is the solvent velocity and R is the solute radius. The force was suggested
by Einstein to be the negative of the gradient of the chemical potential m1. Thus:
dm1
¼ ½6pmRv1
dz

(2.48)
k T dc1
¼ r kB T ln c1 ¼ B
c1 dz
Rearranging:

kB T dc1
c1 v1 ¼ (2.49)
6pmR dz
But (c1v1) is the total flux n1, equal in dilute solution to the diffusion flux j1.
Comparing this with Eq. 2.33 gives:
28 E. Cussler
k T
D¼ B (2.50)
6pmR
The diffusion coefficient in liquids varies inversely with solute size and with
solvent velocity. This simple relation is called the Stokes–Einstein equation.
Equation 2.50 often gives good estimates of diffusion in liquids. Its simplicity is an
invitation to attempt improvements. These include assuming the sphere is not solid but
gas, replacing the sphere with an ellipsoid, allowing the sphere to spin, and putting it in
a small pore. Other attempts at improvement allow for nonideal solutions, assigning
different friction coefficients to solute and solvent, and considering changes close to
the spinoidal. While none of these efforts is definitive, each can clarify the perspective
of a particular chemical system. Still, the simple Stokes–Einstein equation is the best
place to start for understanding diffusion in liquids.
Solids. As explained above, the diffusion in solids is so slow that most transport
usually occurs in any fluid-filled flows and voids within the solid. Some solid
processes are certainly dramatically affected by diffusion – metallic welds and
doped semiconductors are two good examples – but the diffusion of chemicals in
the environment is usually through fluids.
The relative unimportance of diffusion in solids is fortunate, because diffusion
coefficients in solids scatter. These coefficients do not cluster around a single value,
and they depend strongly on crystal structures. For example, the diffusion of carbon
in body-centered cubic iron is 1010 times faster than the diffusion of carbon in face-
centered cubic iron. Sometimes, an anomalously high coefficient reflects different
types of vacancies in the solid crystals. For example, silver ion diffuses 1017 times
faster in AgCl than sodium ion diffuses in NaCl.
Estimates of diffusion in solids, which normally begin with a face-centered
cubic lattice, assume a coefficient given by:
D ¼ R2 No (2.51)
where R is now the distance between atoms or ions in the crystal; N is the
dimensionless fraction of vacant sites; and o is the jump frequency, the number
of atomic or ionic movements per time. The size of R is estimated from crystal
structure, and the fraction N from the free energy of mixing. The jump frequency o
is often felt to have an Arrhenius temperature dependence. Arguments like this are
not predictions but are rationales to organize data.
Diffusion Across Interfaces
Diffusion from one phase to another is an important and complex limit, a source of
confusion for many. In this case, there are two limits that are close parallels to the
cases of a thin film and a semi-infinite slab discussed above in the sections
“Diffusion Across a Thin Film” and “Other Important Cases.” One case is
exemplified by the so-called infinite couple, when two alloy bars of different but
homogeneous composition are closely joined together. In this case, each atomic
species can diffuse between the two bars, giving concentration profiles that are
known. This limit is rarely important in environmental science and engineering.
The second, much more important limit occurs when solutes diffuse from one
relatively well-mixed phase across a phase boundary to a second relatively well-
mixed phase. This limit approximates what happens when sulfur dioxide in the air
diffuses into a lake. In this case, bulk air is often well-mixed, and the bulk water in
the lake is, too. However, this good mixing does not extend all the way to the
air–water interface. About the last one millimeter of air and about the last one-tenth
millimeter of the water are not well-mixed. Diffusion across these two films, one in
air and the other in the water, is what governs the rate of sulfur dioxide dissolution
in the lake.
We develop these ideas below. “The Mathematics of Mass Transfer” derives the
mathematical framework. “Concentration Units” details transport across interfaces.
“The Meaning of c1∗” uses this framework to calculate the mass transfer in several
environmentally relevant situations.
The Mathematics of Mass Transfer
To begin our study of mass transfer, imagine a small volume of air containing
hydrogen sulfide at concentration c10 that is suddenly contacted with a large
volume of water. The sulfide dissolves in the water so that its concentration c1
drops with time. Predicting this concentration change with the diffusion equations
given above is possible, but difficult. Often, an easier prediction is to use an
alternative tool, a mass transfer analysis, which is more suitable for engineering
applications.
This mass transfer analysis begins by writing a mass balance on the H2S in the air:
½Accumulation in the air ¼

½Amount dissolved in water
dc1
V ¼ AK ðc1 c1 Þ (2.52)
dt
where V is the air volume, A is the interfacial area between air and water, and c1∗ is
proportional to the concentration of the H2S in the water. When there is a lot of
pure, well-mixed water present, this concentration is zero. The rate constant K in
Eq. 2.52 is an overall mass transfer coefficient, a function of H2S diffusion in both
30 E. Cussler
the water and the air. It has the units of velocity, that is, of length L per time t. This
mass balance is subject to the initial condition:
t¼0 c1 ¼ c10 (2.53)
Integrating, Eq. 2.52 becomes:
c1
¼ eKðVÞt
A
(2.54)
c10
The H2S concentration in the air decays exponentially with time, as if it were
undergoing a first-order chemical reaction. The rate constant of this reaction (KA/V)
has units of reciprocal time. However, the concentration in air is not dropping
because of a chemical reaction but because of diffusion of H2S from the air into the
water.
As a second example, imagine absorbing carbon dioxide from flue gas. The flue
gas is steadily flowing upward in a small absorption tower. Excess strong base is
steadily flowing downward through the tower. A mass balance on the carbon
dioxide in a small differential volume dV in the tower results in:
½Accumulation in dV ¼ ½CO2 Flow in out

þ ½CO2 Absorbed by base
dc1
0¼Q Kaðc1 c1 Þ (2.55)
dV
where Q is the volumetric flow rate of flue gas, a is the interfacial area per volume
in the tower, and c1∗ is about zero because the base is strong and there is a lot of it.
As before, K is an overall mass transfer coefficient describing the rate of reaction.
This mass balance is subject to a boundary condition:
V ¼ 0 c1 ¼ c10 (2.56)
Integration gives:
c1
¼ eKaðQÞ
V
(2.57)
c10
The CO2 concentration exiting the absorption column decreases exponentially as

the column volume V is increased or as the column flow Q is decreased. Note that
Eqs. 2.54 and 2.57 are complete mathematical parallels, even though the former
describes unsteady dissolution without flow, and the latter describes steady absorp-
tion with flow.
Interfacial mass transfer is not hard. It is just an alternative description of
diffusion which complements that given by Fick’s Law. The three features do
make interfacial mass transfer complicated. These three complications are the units
of concentration, the detailed meaning of c1∗, and the values of the mass transfer
coefficient K. Details of these features follow.
Concentration Units
The first issue, concentration units, results because the units used for clearly
explaining the ideas are not always those easiest to use in practice. The concentra-
tion units implied in this essay are of the amount per volume, for example, moles
per liter or grams per cubic meter. The concentration units used in practice are
different. In gases, the units are sometimes partial pressures; in liquids, the units are
often mole fractions.
Expressing concentrations as partial pressures or mole fractions leads to differ-
ent definitions of mass transfer coefficients. In particular, the total flux across the
interface N1 from one dilute gaseous solution into another dilute liquid solution may
be defined as:
N1 ¼ n1 jinterface ¼ j1 jinterface
(2.58)
¼ K ðc1 c1 Þ
where c1 is the concentration of species “1” in the gas. The restriction to dilute
solution is not a major constraint. Alternatively, the interfacial flux can be defined
as:
N 1 ¼ K p ð p1 p1 Þ (2.59)
where p1 is the partial pressure of solute “1” in the gas, and Kp is a new, different
overall mass transfer coefficient. But from the ideal gas law:
n1 RT
p1 ¼ ¼ c1 RT (2.60)
V
Comparing the two equations shows:
K
Kp ¼ (2.61)
RT
If K has units of meters per second, then Kp may have units of moles per square
meter per second per pascal.
Similarly, for mass transfer from a liquid into a gas, an alternative definition is:
N1 ¼ K ðc1 c1 Þ (2.62)
32 E. Cussler
where c1 is now the concentration of species “1”in the liquid, and K is an overall
mass transfer coefficient different from that in Eq. 2.58. Alternatively,
N1 ¼ Kx ðx1 x1 Þ (2.63)
where x1 is the mole fraction of species “1” in the liquid, and Kx is still another
overall mass transfer coefficient. Because
c1 ¼ cx1 (2.64)
where c is the total concentration in the liquid, the two overall coefficients are
related:
Kx ¼ cK (2.65)
For example, if K is in meters per second, and c is in moles per meter cubed, then
Kx will have units of moles per square meters per second. Other definitions of
coefficients are also possible, but are no harder.
The Meaning of c1∗
The meaning of the concentration c1∗ appearing in Eqs. 2.52, 2.55, 2.58, 2.59, and
2.62 is the hardest step in this description. These flux equations all assert that the
flux is proportional to a concentration difference. The flux will be zero when the
concentration is zero. Thus, c1∗ must be the hypothetical gaseous concentration of
species “1” that is in equilibrium with species “1” dissolved in the liquid. This is
harder than interfacial heat transfer: there, the heat flux is proportional to the
temperature on one side of the interface minus that on the other side. Here, the
mass flux is proportional to a concentration difference which equals to one real
concentration that does exist minus a second one which is hypothetical.
To be more specific, imagine the case in Eq. 2.58, where c1 is the actual
concentration of species “1” in the well-mixed, bulk gas on one side of the
interface. The concentration c1∗ is equal to the concentration in the well-mixed,
bulk liquid times some type of Henry’s Law constant, which describes equilibrium
between gas and liquid. Sometimes, those studying this point for the first time can
be helped by silently chanting
▾c * is the concentration that would be in the gas if it were in equilibrium with the liquid
1
(which it isn’t).
Remembering this chant may help mastering this difficult point.

To try to make this point clearer, imagine in calculating the flux of oxygen in air
into wastewater with a concentration of 1 104 mol/‘. At equilibrium,
c1 ðgasÞ ¼ 30c1 ðliquidÞ (2.66)
Thus,
c1 ðgasÞ ¼ 30 104 mol/‘ (2.67)
As a result,
0:21 mol mol

c1 c1 ¼ 30 104
22:4‘ ‘ (2.68)
4
¼ 64 10 mol/‘
Understanding problems like these is often helped by always checking what

happens when the system is at equilibrium.
Values of Mass Transfer Coefficients
We now turn to the variations of the overall mass transfer coefficient with quantities
like the diffusion coefficient in the adjacent phases. The most common case is that
of transfer from a gas into a liquid. The concentration in the gas is expressed as
a partial pressure, and the concentration in the liquid is expressed as a mole fraction.
The flux N1 across the interface is then:
N 1 ¼ K p ð p1 p1 Þ
¼ kp ðp1 p1i Þ
¼ kx ðx1i x1 Þ (2.69)
where p1 and x1 are the average concentrations in the gas and liquid, respectively;
and p1i and x1i are the corresponding but unknown gas and liquid concentrations at
the interface. The mass transfer coefficients kp and kx describe transport in the gas
and in the liquid. Sensibly, the individual mass transfer coefficient kp is a function
of diffusion in the gas, but not of diffusion in the liquid; and the individual mass
transfer coefficient kx is the reverse.
The concentrations across the interface will normally be in equilibrium, so that:
p1i ¼ Hx1i (2.70)
where H is a Henry’s Law constant. Combining this constraint with Eq. 2.69 gives:
" #
1
N1 ¼ 1 H ðp1 Hx1 Þ (2.71)
k p þ kx
34 E. Cussler
By comparing this with the overall mass transfer coefficient Kp yields:
1
Kp ¼ (2.72)
1
kp þ kHx
p1 ¼ Hx1 (2.73)
These are the results sought. Obviously, similar equations are possible for other
concentrations and other equilibria analogous to Henry’s Law.
The way in which kp and kx vary with the diffusion coefficients can be estimated
either from experiments or from theories. The experiments are summarized as
correlations, most often in terms of dimensionless numbers. For example, for
mass transfer into a liquid flowing through a packed tower with packing of size d,
the most widely accepted correlation is:
kx ¼ kðliquidÞcðliquidÞ (2.74)
13 v 0:67 D0:50

1
kðliquidÞ ¼ 0:0051 ðadÞ0:4 (2.75)
ng an n
where n is the kinematic viscosity of the liquid, g is the acceleration due to gravity,
v is the superficial liquid velocity, a is the surface area per volume of the packing,
and D is the diffusion coefficient in the liquid. The quantity (v/an) is one form of the
dimensionless Reynolds number; the quotient (n/D) is the dimensionless Schmidt
number. Correlations like this, which are based on extensive experiments, should be
used for estimates whenever possible.
In many cases, however, appropriate correlations may not be reliable, or may not
be available at all. In these cases, estimates for liquids can be made by assuming that:
D D
kðliquidÞ ¼ ¼ (2.76)
‘ 104 m
where D is the diffusion coefficient in the liquid and ‘ is often called the film
thickness or the boundary layer. This casual description can be confusing, because
these terms are more specifically defined in theories of mass transfer.
A corresponding estimate for gases is:
D D
kðgasÞ ¼ ¼ 3 (2.77)
‘ 10 m
where D is now the diffusion coefficient in gases, typically 104 times larger than
that in liquids. Equations 2.76 and 2.77 are major approximations to be used only in
desperation.
Important Special Cases
The sections above describe the mathematics of diffusion and dispersion. They
have summarized characteristics of diffusion coefficients and listed some typical
values. They have discussed mass transfer coefficients as an alternative description
of interfacial diffusion frequently valuable in environmental engineering. None of
the ideas presented are especially difficult to understand.
However, actually putting these ideas into practice can be complicated, largely
because of difficult units and subtle definitions. This final section considers specific
chemical examples that illustrate the ideas involved. These examples are approxi-
mate but can serve as a warning of where trouble can occur.
Overall Mass Transfer Coefficient of Oxygen (Case #1)
Imagine wanting to estimate the mass transfer coefficient Kp of oxygen from air into
water. From Table 2.2 and Eq. 2.77,
D 2 104 m2 =s
kðgasÞ ¼ ¼ ¼ 0:02 m=s (2.78)
‘ 103 m
The coefficient kp is found from this by a unit conversion:
kðgasÞ 0:02 m=s mol

kp ¼ ¼ ¼8 2 (2.79)
RT 6 m 2
atm m s atm
8:2 10 molK 298 K
Similarly, from Table 2.3 and Eq. 2.76,
2 109 m2 =s
kðliquidÞ ¼ ¼ 2 105 m=s (2.80)
104 m
The coefficient kx has a different conversion:

1 mol
kx ¼ kðliquidÞc ¼ 2 105 m=s
18 106 m3 (2.81)
mol
¼ 1:1 2
m s
Henry’s Law for this system is (cf. Eq. 2.70):

p1 ¼ 4:3 104 atm x1 (2.82)
36 E. Cussler
Thus from Eq. 2.72,
1
Kp ¼
m2 s atm þ 4:3104 atm m2 s
8 mol 1.1 mol (2.83)
5 mol
¼ 2:6 10
m2 s atm
This coefficient describes oxygen transport between air and water when the
concentration difference is expressed as partial pressures of oxygen.
Overall Mass Transfer Coefficient of Oxygen (Case #2)
The example above is straightforward because it matches the detailed equations

given earlier. The same problem can be solved in different units. Thus, the flux N1 is:
N1 ¼ Kc ðc1 c1 Þ (2.84)
where c1 is the actual oxygen concentration in water, c1∗ is the oxygen concentra-
tion in water that is in equilibrium with air, Kc is a different overall mass transfer
coefficient given by:
1
Kc ¼ (2.85)
m
kðgasÞ
þ k liquid
1
ð Þ
and m is a different form of Henry’s Law constant, defined by the equilibrium:
c1 ðliquidÞ ¼ mc1 ðgasÞ (2.86)
Comparing Eqs. 2.81 and 2.82 gives:
cðliquidÞRT
m¼
H 3

mol 6 m atm
¼ 8:2 10
18 106 m3 mol K (2.87)

298 K
4:3 104 atm
¼ 0:03
The numerical value of Henry’s Law constant is completely different. The

combination of Eqs. 2.78, 2.80, 2.85, and 2.87 gives:
1
Kc ¼ 0:03
0:02 m=s þ 21053 m=s (2.88)
5
¼ 2 10 m=s
In both this formulation and that in Eq. 2.83, diffusion in the liquid dominates the
mass transfer. This is often taken as a consequence of the slower diffusion in the
liquid. This is not completely true, as the next example shows.
Overall Mass Transfer Coefficient of Ammonia from Air into Water
This example illustrates how the rate at which ammonia is dissolved in water can be
estimated. The individual mass transfer coefficients of ammonia are easily found:
DðgasÞ 2:3 105 m2 =s

kðgasÞ ¼ ¼ ¼ 0:023 m=s (2.89)
‘ 103 m
and
DðliquidÞ 1:6 109 m2 =s

kðliquidÞ ¼ ¼
‘ 104 m (2.90)
4
¼ 1:6 10 m=s
One Henry’s Law constant for dilute acid is given in the literature as:

atm ‘
p1 ðatmÞ ¼ 7000 c1 ðliquid, molarÞ (2.91)
mol
Converting the pressure into a molar concentration:

atm ‘ mol K
c1 ðgas, molarÞ ¼ 7000
mol 0.082 atm ‘

c1 ðliquid, molarÞ (2.92)
298 K
¼ 290c1 ðliquid, molarÞ
Combining these results with Eq. 2.85 gives:
1
Kc ¼ 290
0:023 m=s þ 1:61015 m=s (2.93)
4
¼ 1:3 10 m =s
38 E. Cussler
The much higher solubility of ammonia in dilute acid means the mass transfer is
now more affected by diffusion in air.
Toxin Diffusion in a Biofilm
The final example imagines a dilute toxin dissolved in water and metabolized
irreversibly by microorganisms immobilized in a biofilm. This example, which is
not as chemically specific as the first three, also assumes that the concentration of
dissolved oxygen is much greater than the concentration of the toxin. Thus, the rate
per biofilm area N1 is given by the overall rate of diffusion of the toxin to the
biofilm, followed by the diffusion and reaction of the toxin within the biofilm.
This overall rate is mathematically equivalent to mass transfer across an inter-
face, where the solute diffused through the gas to reach the interface, quickly
crossed the interface, and then diffused into the liquid. In fact, the biofilm case is
often easier because most biofilms are largely water and hence their partition
coefficient m is one. Thus,
c1
N1 ¼ (2.94)
1
kðliquidÞ
þ k biofilm
1
ð Þ
The coefficient k (liquid) can often be found from mass transfer correlations;
Equation 2.76 provides a first guess. The value in the biofilm depends on the details
of the reaction. However, delightfully, most theories give the same result:
12
DðbiofilmÞ
kðbiofilmÞ ¼ (2.95)
t
where t is the half-life of the reaction. While beyond the scope of this entry, this
result is carefully derived in most books on diffusion and reaction. This result
underscores the value of the simple ideas of diffusion and reaction presented here.
Future Directions
After 150 years of concentrated effort, diffusion is an established subject. Active

research does continue on, for example, semiconductors and polymer membranes,
but this does not have major environmental application.
Diffusion is an important tool for describing environmentally significant mass
transfer. In many cases, this transfer can be described in terms of diffusion
coefficients. In many air pollution problems, mass transfer can be described in
terms of dispersion, which is mathematically similar to diffusion but due to coupled

diffusion and flow.
The underused description of mass transfer, especially across interfaces, is in
terms of mass transfer coefficients. These are functions of diffusion coefficients and
of other parameters, like velocity and viscosity. Exploiting this topic offers poten-
tial gain for environmental engineering.
Bibliography
Books and Reviews
Aris R (1999) Mathematical modeling. Academic, San Diego. ISBN 0-12-604585-1

Astarita G, Savage DW, Bisio A (1983) Gas treating with chemical solvents. Wiley, New York.
ISBN 0-471-057681
Berg HC (1993) Random walks in biology. Princeton University Press, Princeton. ISBN 0-691-
000646
Bird RB, Stewart WS, Lightfoot EN (2002) Transport phenomena, 2nd edn. Wiley, New York.
ISBN 0-471-41077-2
Carslaw HS, Jaeger JC (1976) Conduction of heat in solids. Oxford University Press, New York.
ISBN 0-19-853303-9
Crank J (1980) The mathematics of diffusion. Oxford University Press, New York. ISBN 0-19-
853411-6
Cussler EL (2009) Diffusion: mass transfer in fluid systems, 3rd edn. Cambridge University Press,
Cambridge. ISBN 0-521-56477-8
Dutta BK (2009) Mass transfer and separation processes. PHI Learning, New Delhi. ISBN 978-81-
203-2990-4
Seader JD, Henley EJ (2006) Separation processes principles, 2nd edn. Wiley, New York. ISBN
0-471-46480-5
Seinfeld JH, Pandis S (2006) Atmospheric chemistry and physics, 2nd edn. Wiley, New York.
ISBN 0-471-72018-6
Sherwood TK, Pigford RL, Wilke CR (1975) Mass transfer. McGraw Hill, New York. ISBN
0-070-566929
Taylor GI (1953) Dispersion of soluble matter in solvent flowing slowly through a tube. Proc
R Soc London Ser A 219:186–203
Taylor GI (1954) The dispersion of matter in turbulent flow through a pipe. Proc R Soc Lond Ser
A 223:446–468
Taylor R, Krishna R (1993) Multicomponent mass transfer. Wiley, New York. ISBN
0-47-157417-1
Treybal RE (1980) Mass transfer operations, 3rd edn. McGraw-Hill, New York. ISBN 0-070-
651760
Chapter 3
Toxic Organic Chemicals
Simanga Gama, Jon A. Arnot, and Don Mackay
Glossary
Bioaccumulation The phenomenon similar to bioconcentration but including

uptake from food as well as uptake from the ambient environ-
ment. This is expressed as a bioaccumulation factor (BAF)
and generally applies to organisms in the environment. The
BAF is the steady-state ratio of the chemical concentration in
the organism to that in the environment.
Bioconcentration The phenomenon by which an organism, such as a fish,
absorbs chemical from its ambient environment of water or
air by respiratory uptake and/or dermal absorption. The
steady-state ratio of the chemical concentration in the organ-
ism to the chemical concentration in its ambient environment
of water or air is the bioconcentration factor, (BCF) and the
BCF is usually measured in a laboratory test.
Biomagnification The ratio of the chemical concentration in the predator to
that of the prey.
S. Gama (*) • D. Mackay
Canadian Centre for Environmental Modelling and Chemistry, Trent University,
1600 West Bank Drive, Peterborough, ON K9J 7B8, Canada
e-mail: [email protected]; [email protected]
J.A. Arnot
Department of Physical and Environmental Sciences, University of Toronto Scarborough,
1265 Military Trail, Toronto, ON M1C 1A4, Canada

42 S. Gama et al.
Hazard The inherent toxic potency of a chemical, usually

expressed as the quantity or concentration of the substance
necessary to elicit a defined adverse effect in the organism.
Mass Balance Model A mathematical description of the fate and transport of
a chemical in the environment, usually in the form of
a computer program. The model provides a complete
accounting of all processes experienced by the chemical
and is used to estimate environmental concentrations, per-
sistence, and exposures. Models may also be used to fore-
cast future changes in concentrations as a result of actions
to reduce contamination.
Persistence The average time that a discharged chemical survives in
the environment before it is degraded into another sub-
stance or substances. It may be expressed as a “half-life”
by analogy to radioactive substances or as a residence time.
Risk The likelihood that there will be an adverse effect as
a result of exposure to the chemical. Risk thus depends
both on toxic potency (hazard) and the prevailing
exposure.
Toxicity The phenomenon by which a chemical substance elicits an
adverse effect on an exposed organism. The effect may be
death (lethality) or a less severe effect such as a failure to
reproduce, an increased vulnerability to predation or sig-
nificant behavioral changes.
Organic chemicals play an invaluable role in the modern lifestyle. They include
pharmaceuticals, pesticides, plastics, fuels, solvents, explosives, surface coatings,
adhesives, disinfectants, and fire retardants. From the perspective of conservation
and sustainability, the preferred strategy is to use chemicals such that they perform
their desired function, cause no unintended adverse effects, and hence leave no legacy
of contamination. This is a fundamental component in the move toward “green
chemistry” which also strives to reduce resource depletion, energy use, ozone deple-
tion, and interference with natural biogeochemical cycles. Of the over 50 million
chemicals that have been characterized, most are organic compounds. Some 100,000
are commercially produced in quantities large enough to raise concerns that they may
become present in the environment in sufficient quantities and at sufficient
concentrations to cause risks to the well-being of humans or other organisms. As
a response to this concern, most regulatory agencies have listed chemicals of national
concern, a typical number being in the hundreds to thousands. These “toxic organic
chemicals” merit regulatory scrutiny and possible controls over synthesis and use.
3 Toxic Organic Chemicals 43
At the international level, the Stockholm Convention has listed over 20 substances as
worthy of regulation or even bans. Most of these are synthetic chlorinated organic
substances, but some naturally occurring and inadvertently produced substances are
also of concern. In this entry, the chemical attributes or properties of these substances
are described and criteria that dictate the level of concern about hazard and risk to
environmental and human health are outlined. Selected classes of organic chemicals
are then discussed in more detail.
Introduction
A consensus has emerged among the scientific and regulatory communities that it is
a combination of properties of chemical substances that dictates their designation as
“toxic organic substances” and thus the need for their regulation. The result has been
the identification of classes of chemicals as persistent organic pollutants (POPs) and
persistent bioaccumulative and toxics (PBTs) and those that undergo long-range
transport (LRT) on a global scale. Although numerical criteria have been developed
for certain key chemical properties such as toxicity and persistence, it is apparent that
there is a continuum of properties, thus there is no clear demarcation between for
example, PBTs and non-PBTs. There are varying degrees of PBT-like character among
“toxic organic chemicals.” It can be justifiably argued that all chemical substances are
toxic if the administered dose or exposure is sufficient, that is, as stated over 500 years
ago by Paracelsus, “it is the dose that makes the poison.” Accordingly, there is no clear
demarcation between toxic and nontoxic substances: the demarcation must be on the
basis of both potency as a toxic agent and the exposure experienced in the environment.
The regulatory focus must be on ensuring the principles of sustainability and conserva-
tion by reducing exposure.
It is important to discriminate between the terms hazard and risk, since these
terms are often wrongly used interchangeably. Hazard reflects the inherent
properties of the chemical such as its toxic potency, regardless of the actual
exposure. Risk reflects the probability of adverse effects as a function of both
hazard and the actual exposure. A highly hazardous substance may pose a low
risk and vice versa.
Properties and Characteristics of Toxic Organic Chemicals
Organic chemicals tend to fall into distinct classes based on their molecular
structure, an example being the alcohols of increasing carbon chain length, metha-
nol, ethanol, propanol, butanol, etc. There is often a systematic change in properties
such as boiling point or vapor pressure in such homologous series. This feature is
invaluable in that interpolation and modest extrapolation of properties is often
possible. This is formalized in linear free energy relationships (LFER) and is
exploited in the development of quantitative structure activity (or property)
44 S. Gama et al.
relationships (QSARs and QSPRs) that are widely used for property estimation
purposes. As quantum chemical molecular modeling computation methods become
more reliable and accessible, it is clear that the key properties of chemical
substances are amenable to a priori estimation from molecular structure.
A key LFER arises from a systematic increase in halogenation of a parent
organic substance. For example, the series benzene, chlorobenzene, dichloroben-
zene to hexachlorobenzene displays consistent changes in properties such as
a decrease in vapor pressure and aqueous solubilities. This also occurs with bromine
and to an extent with fluorine substitution and with increased methylation, for
example, benzene, toluene, xylene, etc. These systematic property changes are
exploited when later reviewing the properties of groups of chemicals such as the
polychlorinated biphenyls (PCBs). The key properties as discussed by Mackay et al.
[1] are addressed below.
Molecular structure and molar mass (g/mol) are obvious fundamental properties
that reflect the constituent elements (molar mass) and their location in the chemical
(molecular structure). It is necessary to discriminate between isomers – chemicals
comprised of the same elements but with different structural or spatial patterns. For
example, the four stereoisomers of hexachlorocyclohexane (HCH) are comprised of
the exact same elements (six carbon, six chlorine, and six hydrogen atoms)
contained in the same general structure (a cyclic ring of six carbon atoms, each
with a bond to a chlorine and a hydrogen atom), but with different three-
dimensional spatial patterns of these bonds. The seemingly minor differences in
HCH isomer structure result in significant differences in toxicity.
Melting point and boiling point dictate the state of the pure chemical at atmospheric
temperatures and pressure as either solids, liquids, or gases.
Chemical solubility and partitioning: the tendency of a substance to partition from
its pure state into the atmosphere, water, and octanol, is expressed, respectively, as
the saturation vapor pressure, aqueous solubility, and solubility in octanol. These
properties are useful but are not always measurable. For example, the solubility of
ethanol in water is not measurable because of miscibility. Octanol is widely
accepted as a surrogate phase for partitioning into lipids (fats) in biota and for
natural organic matter (OM) and organic carbon (OC) in soils and sediments. Since
saturation conditions for organic chemicals rarely exist in the environment, it is
more convenient to use ratios of these three solubilities, that is, the three partition
coefficients or ratios of concentrations, KAW (air–water), KOW (octanol–water), and
KOA (octanol–air). Clearly only two of these are independent since KOA is KOW/
KAW. The air–water partition coefficient KAW is H/SSW or (PS/RT)/SSW where PS is
the saturation vapor pressure (Pa), R is the gas constant (8.314 Pa m3/mol K), T is
absolute temperature (K), SSW is the solubility in water (mol/m3), H is the Henry’s
law constant (Pa m3/mol) and is PS/SSW. For most environmental purposes, KOW is
assumed to be equivalent to the lipid–water partition coefficient. KOW is also used
to estimate partitioning to organic carbon from water (KOC) as approximately 0.35
KOW plus or minus a factor of three [2].
The KAW, KOW, KOA properties thus largely dictate the equilibrium partitioning
characteristics of the chemicals between air, water, soils, and sediments that contain
organic carbon as well as biota in terms of their lipid content. A chemical space
diagram provides a convenient depiction of these properties as plots of either log
KAW versus log KOW, (in which case the constant log KOA values lie on a diagonal),
or as log KAW versus log KOA, with log KOW values lying on a diagonal. Figure 3.1
is such a plot and shows the points corresponding to the series of chemicals
representing particular chemical classes. For example, points for 1,10 -(2,2,2-
trichloroethylidene)bis 4-chloro-benzene (DDT), benzo(a)pyrene, and PCBs are
plotted showing the wide variation in partitioning properties which translate into
differences in environmental fate. On this plot are lines corresponding to various
estimated mass percentage partitioning between air, water, and octanol phases for
an assumed volumetric proportion of these phases.
These simple relationships apply only if the molecule does not dissociate or
ionize as applies to organic acids such as phenols at high pH and organic bases such
as amines at low pH. For these substances, knowledge of the dissociation constant
pKA and the prevailing pH is essential.
Reactivity, expressed as a rate constant or half-life, is an essential property
because it determines the persistence, that is, the residence time of the chemical
in the environment. Persistence is important because to a first approximation the
quantity of the chemical residing in the environment (kg) and hence the
concentrations that exert toxicity (kg/m3) are proportional to the discharge rate
of chemical to the environment (kg/h) and to the chemical’s residence time (h).
Chemicals usually react following second-order kinetics
Rate ¼ k2 CC CR ¼ k1 CC ¼ 0:693CC =t1=2
where k2 is a second-order rate constant, CC is chemical concentration, and CR is

the concentration of a reacting species such as hydroxyl radicals in the atmosphere
[3]. It is often convenient to lump k2 and CR as a first-order rate constant k1 (h1)
but this assumes that CR and k2 are constant. In reality, both vary spatially and
temporally, thus the half-life t1/2 is also expected to vary. Despite this variability,
persistence is such an important property that it is necessary to provide estimates of
half-lives of the chemical for all relevant media to which the chemical partitions.
Some literature and databases exist on the half-lives of well-studied priority
chemicals in air, water, soils, sediments, and in biota by metabolic conversions or
biotransformation [1, 4–6]. There is, however, a paucity of data for other substances
including “emerging” contaminants. The most common mechanisms are reaction
with oxidizing species such as hydroxyl radicals, hydrolysis, photolysis (direct and
indirect), biodegradation by microorganisms, and biotransformation in animals and
plants.
Regulatory agencies have set half-life criteria for specific environmental media
(for example, 60 days in water and 2 days in the atmosphere) [7]; however, in
addition to medium-specific half-lives, the overall or average persistence of
–4 –3 –2 –1 0 1 2 3 4
4
46
3 1% octanol
n–Hexane 5
n–Heptane Air and Octanol
2 Air 6
1
2–Methylpropene 7
Cyclohexene X y lene
0 1% air
Dimethyl ether Ethylnaphtahlene 8
−1 Epichlohydrin
Dibenzofuran Monochlorobiphenyl 9
Dichlorobiphenyl
−2 Ethylbenzoate
Diphenyl ether Tetrachlorobiphenyl 10
Air and Water
−3 A&W&O 2,3,7,8-TCDD
2,4–Dichlorophenol dibenzo–p–dioxin 2,3,7,8-TCDF 11
−4
Log KOA
Log KAW
2,4–Dimethylpehenol 4–Bromophenyl ether 2,7–DCD
DDT 12
−5 2,8–DCDF
Permethrin
Catechol
γ–HCH 13
−6 Diethyl phathalate BaP
14
−7
Water Water and Octanol
−8 15
Octanol
−9 1% water 16
−10 BPA
17
−11
0 1 2 3 4 5 6 7 8 9
Log KOW
Fig. 3.1 Chemical space plot of the air–water partition coefficient (log KAW) versus the octanol–water partition coefficient (log KOW) showing positions of
chemical classes on the three media of air, water, and octanol. Octanol serves as a surrogate for organic matter in soils and sediments. Constant log KOA values
S. Gama et al.
form the dashed diagonals on this plot of 1, 6, and 10 (from top to bottom) since log KOA is log KOWlog KAW. The heavy lines represent 1% partitioning into
each medium
10,00,000
Air Soil
Water Sediment
1,00,000
Reactivity (Half-life)/h
10,000
1,000
100
10
1
he e
Xy e
yl e
a I e
tra zo ha n
di lor ]py e
nz ip ne
2, iox l
2, - in
D ,8- DD
zo DD
et e ,3,7 -DC n
l p ich -T F
D ala ohy F
4- hy (D in
hl th )
ph te
at ol
et l
in
γ- T
he 4
lA
H
p- ny
rm ho
ic h P
lo an
o- xan
p- len
en
Te Ben pht nda
ch [a len
ra
sp D
hy p ,8 D
h r D
D
C
2, et te dr
C en
hr
o- la
D lp M
be ob re
no
o- he
ht lo C
Pe ec
D
7 C
en C
fu
H
or a
yc x
d
C -He
3, D
ib T
X
8
7
,
n
Bi
l-N
2
hy
i
Et
1-
im
D
Chemicals
Fig. 3.2 Estimated half-lives (hours) of selected chemicals in air, water, soil, and sediment [6]
a chemical in the environment is also important and is controlled by the relative

quantities of chemical in each medium [8]. The relative quantities in each environ-
mental medium, often referred to as the environmental fate and distribution, are
largely a function of the partitioning properties of the chemical and the composition
of the environment. In principle, the overall reaction rate constant (kOV) is the
weighted mean of the rate constants, the weighting being done according to the
proportions in each medium. Computer programs such as the OECD Tool [9] are
used to perform such calculations and yield an “overall persistence.” This quantity is
essential in any assessment of sustainability because the year-to-year carryover of
a chemical is controlled by overall persistence. In principle, an overall persistence
(POV = 0.693/kOV) of say 2 years implies that an appreciable fraction of the chemical
discharged in year 1 will remain after 5 or more years. Indeed, this is the most
important single property of an organic chemical, hence their designation as persis-
tent organic pollutants (POPs) [10]. See also the discussion by Gouin et al. [11],
Mackay [12] and Scheringer [13]. Figure 3.2 illustrates the variation in reactivity of
a selection of chemicals in each medium.
Potential for long-range transport (LRT) in air or oceans is increasingly recognized
as a cause for concern because the adverse effects may be experienced at locations
remote from the source [13]. Obvious regions of concern are the Arctic and
48 S. Gama et al.
10,000
Tetrachlorobiphenyl 2,4-Dichloro-phenol
n -Hexane
o-Xylene
1,000 Epichlorohydrin
Diethyl phthalate γ-HCH D4
100
DDT
LRT/km
10
Benzo[a]pyrene
1
Permethrin
0.1
2,3,7,8-TCDD
0.01
Alkanes/Alkenes Monoaromatics PAHs PCBs Dioxins/ Ethers Esters Phenols Pesticides Siloxanes
Furans
Organic chemical classes
Fig. 3.3 Long-range transport (LRT; km) in air estimated by Level II fate calculations using the
RAIDAR 1.00 model [16]. The chemicals are categorized into general classes
Antarctic where local biota may be unacceptably contaminated and used as food by
mammals including humans. Multimedia mass balance models have been devel-
oped that predict the influence of a chemical’s environmental phase distribution on
its ability to be transported over long distances [14]. Persistence, and hence
reactivity, of the chemical in air influences its potential to be transported over
long distances in air. Multimedia models currently used to estimate LRT parameter
include Globo-POP [15], RAIDAR [16], and the OECD Tool [9]. The models
simulate LRT by incorporating the chemical partitioning properties discussed
above. More complex global atmospheric and oceanic circulation models have
also been used to simulate and forecast the global cycling of chemicals, for
example, Guglielmo et al. [17].
Figure 3.3 shows the variation in long-range transport potential as calculated by
the RAIDAR model for a selection of substances.
Bioaccumulation/Bioconcentration and Toxicity
Bioaccumulation refers to the tendency for an organic chemical to partition into biota
from their surrounding environment. The bioconcentration factor (BCF) is a metric of
bioaccumulation and is the steady-state ratio of the chemical concentration in an
organism, such as a fish, to the chemical concentration in an environmental medium,
such as water. The BCF is usually measured in a laboratory test and does not include
exposures to chemical from dietary sources. Bioaccumulation includes chemical
6 2
5 1
0
4
log 1/LC50
−1
log BCF
y = 0.63x - 3.3
3
r 2 = 0.63
−2
2
−3
1 −4
y = 0.60x - 0.02
r 2 = 0.88 2,3,7,8-TCDD
0 −5
0 2 4 6 8
log KOW
Fig. 3.4 Plot of bioconcentration factors (BCF; L/kg, blue circles) in fish and 14-day aquatic
toxicity estimates for fish (1/LC50; L/mg, red squares) of selected organic chemicals as a function
of chemical hydrophobicity (KOW) [6]
uptake from the environment, that is, bioconcentration, and from food. Biomagni-
fication refers to the increase in chemical concentration from prey to predator [18].
Chemical hydrophobicity, as characterized by KOW, plays a large role in deter-
mining the BCF and the toxicity of the chemical as measured in aquatic tests for fish.
Hydrophobic organic chemicals bioconcentrate in aquatic organisms as a result of the
chemical partitioning from the water into the lipids of the organism. Prolonged
bioconcentration of the chemical may induce toxicity by, for example, disrupting the
integrity of the lipid bilayer of cell membranes, that is, baseline narcotic mode-of-toxic
action. Indeed, the bioaccumulation process results in the transport of chemical from
external media, where there are generally no sites for toxic action, into the organism,
where there are sites for toxic action [19]. Figure 3.4 illustrates the fundamental role of
partitioning (or bioconcentration) on aquatic toxicity when the parameters for the BCF
(L/kg) and toxicity in fish (LC50; mg/L) are plotted as functions of KOW.
Bioconcentration and aquatic toxicity (expressed as 1/LC50) both increase with KOW
and the slopes of the two relationships are almost identical. Chemicals with log KOW
greater than 5 include polycyclic aromatic hydrocarbons (PAHs), many pesticides,
chlorinated furans and dioxins, and polychlorinated biphenyls (PCBs) which are
implicated as being both toxic and persistent. Chemicals that are highly toxic, such
as 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD), and that have modes of action
that are more specific than baseline narcosis will typically fall well below the 1/LC50
regression line.
50 S. Gama et al.
The Stockholm Convention on Persistent Organic

Pollutants (POPs)
The Stockholm Convention is an international agreement designed to limit the use of

chemicals that are persistent, bioaccumulative, and subject to long-range transport on
a continental or global scale such that they cause significant adverse effects (toxicity) in
regions distant from where they are used or manufactured [10]. The initial group
comprised of 12 substances or classes was colloquially designated the “dirty dozen.”
In recent years, other substances have been added. Table 3.1 lists these substances. It is
Table 3.1 Chemicals listed under the Stockholm Convention or under review as of 2010;
these chemicals are categorized under Annex A (Elimination), Annex B (Restriction), or
Annex C (Unintentional production)
Initial 12 chemicals (i.e., “Dirty Dozen”)
Aldrin (Annex A) Pesticide
Chlordane (Annex A) Pesticide
DDT (Annex B) Pesticide
Dieldrin (Annex A) Pesticide
Endrin (Annex A) Pesticide
Heptachlor (Annex A) Pesticide
Hexachlorobenzene (Annex A, C) Industrial chemical
Mirex (Annex A) Pesticide
Toxaphene (Annex A) Pesticide
Polychlorinated biphenyls (Annex A, C) Industrial chemical and flame retardant
Polychlorinated dibenzo-p-dioxins (Annex C) By-product
Polychlorinated dibenzofurans (Annex C) By-product
Recently listed chemicals
a-hexachlorocyclohexane (Annex A) Pesticide, by-product
b-hexachlorocyclohexane (Annex A) Pesticide, by-product
g-hexachlorocyclohexane (Annex A) Pesticide
Chlordecone (Annex A) Pesticide
Hexabromobiphenyl (Annex A) Industrial chemical and flame retardant
Pentachlorobenzene (Annex A, C) Pesticide, industrial chemical
Perfluorooctane sulfonic acid, its salts and
perfluorooctane sulfonyl fluoride
(Annex B)
Polybrominated diphenyl ethers (Annex A) Industrial chemical
Tetra, pentabromodiphenyl ether (Commercial
penta)
hexa- and heptabromodiphenyl ether
(Commercial octa)
Chemicals under review as of 2010
Endosulfan (Annex F) Insecticide
Hexabromocyclododecane (Annex E) Flame retardant
Short-chained chlorinated paraffins (Annex E) Used in pressure lubricants in metal industry.
Also used in sealers, glue coatings in building
industry, and in leather and rubber treatments.
noteworthy that many are chlorinated hydrocarbons, the reason being that chlorination
impacts stability and hence persistence to the molecule.
Chemical Classes
A brief description of some chemical classes is given and provides references for
further reading. Molecular structures for selected chemicals in each class are also
provided.
Alkanes and Alkenes
The alkanes are present in fuels and generally exhibit low solubilities in water, which
when coupled with their relatively high vapor pressures results in relatively high
air–water partition coefficients (KAW). Values of KOW are high and increase with
carbon number, for example, n-hexane has a log KOW value of 4.11 while the
higher molecular weight n-heptane has a log KOW value of 5.0 [1]. Half-lives in air
and water are relatively short (approximately 40 and 200 h in air and water) [1, 3].
The corresponding unsaturated hydrocarbons or alkenes are more soluble in water, have
lower KAW and KOW values and are more reactive, especially in air in which half-lives
may be only a few hours. They are thus implicated in the formation of photochemical
smog due to the presence of the double bond which renders them susceptible to
oxidation by hydroxyl radical. This class of chemicals predominantly partitions to air
as indicated in Fig. 3.1.
The halogenated (primarily chlorinated) alkanes are less flammable and are thus
valuable solvents but they have longer half-lives and are more environmentally
persistent, especially in soils and ground water. They are more hydrophobic than
the parent alkanes as a result of their lower solubility in water. Of particular
environmental concern are the chlorinated C2 alkanes and alkenes such as trichlo-
roethylene that are widely used as solvents and have become frequent groundwater
contaminants. The chlorinated long- and short-chain paraffin compounds have
many uses, notably as cutting oils.
The cycloalkanes are similar in properties to the alkanes. The fully chlorinated
cyclohexanes are of concern because of their persistence, volatility, and potential
for long-range transport. The pesticide lindane (g-hexachlorocyclohexane or g-
HCH) and its isomers a-HCH and b-HCH have been distributed globally by
transport in air and ocean currents. Uses of these chemicals are now limited.
Brominated cyclohexanes that are used as fire retardants are also persistent and
bioaccumulative.
52 S. Gama et al.
n-hexane
hexene
cyclohexane
cyclohexene
Mono-aromatics
The mono-aromatics, for example, benzene, toluene, and the isomers of xylene,
are significant components of fuels and are important chemical intermediates.
They have higher solubilities in water than the corresponding cycloalkanes and
they are relatively volatile and reactive in air contributing to photochemical smog
[3]. Mono-aromatics such as xylene partition mostly to air as indicated in Fig. 3.1,
while higher molecular weight aromatics such as polyaromatic hydrocarbons
(e.g., 1-ethyl naphthalene) have lower volatilities and water solubilities. Chlori-
nation results in increased persistence and hydrophobicity (KOW). Hexachlor-
obenzene is very persistent and has become globally distributed by atmospheric
transport and is banned under the Stockholm Convention.
o-xylene
m-xylene
p-xylene
Polyaromatic Hydrocarbons (PAHs)
Fusion of a number of benzene rings results in an important series of PAHs such

as the 2-ring naphthalene, 3-ring phenanthrene, 4-ring pyrene, and 5-ring benzo-
a-pyrene (BaP). These substances are formed along with smoke and soot during
incomplete combustion. They are hydrophobic, relatively involatile, and are regarded
as particularly toxic. BaP is a well-studied human carcinogen. The primary envi-
ronmental concern is human inhalation in the vicinity of combustion sources.
Chlorinated naphthalenes are used as PCB substitutes and are hydrophobic and
persistent. Chlorination of PAHs increases their hydrophobicity and stability, and
decreases their water solubility and volatility. For further reading, see Haritash and
Kaushik [20], Mastral and Callen [21], and Wilson and Jones [22].
1-ethyl-naphthalene
Polychlorinated Biphenyls (PCBs)
Polychlorinated biphenyls (PCBs) are a class of stable synthetic organic

chemicals consisting of 209 possible congeners. PCBs are used industrially as
dielectric fluids in capacitors and transformers but also as flame retardants, inks,
sealants, and plasticizers. The stability of the C–Cl covalent bond makes these
chemicals more persistent in the environment. Most biodegradation reactions
involve oxidation or reduction of substrate chemicals. PCBs are also resistant to
oxidation, making them resistant to biodegradation. As one of the most notorious
toxic organics they have been banned under the Stockholm Convention, but large
54 S. Gama et al.
quantities remain in use and they are widespread in distribution. PCBs are
associated with adverse health effects in many fish, marine mammal, and avian
species. For instance, levels of PCBs above 20 mg/kg in the blubber of whales and
seals have been linked to thyroid, immune, and reproductive disruption [23, 24].
PCBs are predicted to partition to octanol more than air and water, but they
are sufficiently volatile to be subject to atmospheric transport. For further
reading, see also Fox et al. [25], Hoffman et al. [26], Borga et al. [27], Aken
et al. [28], Van den Berg et al. [29], and Domingo and Bocio [30].
Cl
2-chlorobiphenyl
Cl Cl
2,2’-dichlorobiphenyl
Dioxins and Furans
Non-chlorinated dioxins and furans such as dibenzo-dioxin and dibenzo-furan,

respectively, are relatively water soluble and volatile as indicated in Fig. 3.1 and
they can partition to all three media. Chlorinated dioxins and furans are by-products
of industrial processes and naturally occurring, and are categorized as being
unintentionally produced under Annex C of the Stockholm Convention [10].
Chlorination of dioxins and furans results in increased hydrophobicity, lower
water solubilities, and lower volatilities, as demonstrated in Fig. 3.1 where
chloro-furans and chloro-dioxins such as 2,8-dichlorodibenzo-furan (2,8-DCDF),
2,7-dichlorodibenzo-dioxin (2,7-DCDD), respectively, are regarded as partitioning
largely to octanol, that is, organic phases. Increased chlorination further increases
their hydrophobicity as indicated with 2,3,7,8-TCDD and 2,3,7,8-TCDF. A notable
congener is 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD), which is a very
potent and persistent toxicant, indeed one of the most toxic organic chemicals
known. See also Lohmann and Jones [31], Aberg et al. [32], Van den Berg et al.
[29], and Domingo and Bocio [30].
O dibenzo-p-dioxin
Cl O
O Cl 2,7-dichlorodibenzo-p-dioxin
Cl O Cl
Cl O Cl 2,3,7,8-tetrachlorodibenzo-p-dioxin
Cl
Cl
Cl
Cl
O 2,3,7,8-tetrachlorodibenzofuran
Ethers
Of more recent environmental concern are certain halogenated ethers such as

polybrominated diphenyl ethers (PBDEs). PBDEs are used as fire retardants and
are added to a wide variety of commercial and household products. For exam-
ple, PBDEs are added to polyurethane foam and used in upholstered furniture to
reduce their flammability [33]. Commercial pentabromodiphenyl ether, which is
mainly composed of tetrabromodiphenyl ether and pentabromodiphenyl ether, is
categorized as a persistent organic pollutant and has recently been banned under
the Stockholm Convention [10]. Because of their low water solubility, rela-
tively high log KOW, low log KAW, and low vapor pressure, in the environ-
ment they partition primarily to sediments and soils. Their persistence in soil
and sediments can result in accumulation in the environment and in food
webs. Short non-brominated ethers such as dimethyl ether (a common labo-
ratory solvent) are highly soluble, have a high vapor pressure and low log
KOW (see Fig. 3.1), and are less persistent. See also Vonderheide et al. [34],
Wang et al. [35], and de Wit et al. [36].
56 S. Gama et al.
O
dimethyl ether
O diphenyl ether
Br 4-bromodiphenyl ether
Esters
Esters are a product of a condensation reaction between alcohols and carboxylic

acids. Esters of low molecular weight are relatively volatile and are components
of fragrances. Notably among the esters of environmental concern are the phthal-
ate acid esters (PAEs). These commercial chemicals have been used in a wide
range of products namely; household electronics, building materials, insecticides
and pharmaceutical products. These compounds are found worldwide in oceans,
groundwater, sediment, and the atmosphere [37, 38]. Examples of PAEs include
dimethyl phthalate (DMP) and diethyl phthalate (DEP).
O diethyl phthalate
Phenols
Phenols are derivatives of benzene possessing a hydroxyl (OH) group. They are
used as reagents in the manufacture of dyes, drugs, fungicides, and pesticides and
in chemical and paper industries. Sources of emissions of phenolic waste to the
environment are from petroleum refineries, industrial plants, mine discharges,
and through general uses in products containing resins and paints. The hydroxyl
group can dissociate into a phenolate and hydrogen ions imparting acidity and
relatively high solubility in water [39]. Substituted phenols such as chlorinated
and methylated (2, 4-dichlorophenol, pentachlorophenol (PCP) and 2, 4-
dimethylphenol) are more susceptible to dissociation and are more acidic and
toxic. PCP has been widely used as a wood preservative, while phenol was one of
the first medical disinfectants. Catechol is a unique phenol with two hydroxyl
groups and is implicated in affecting the human nerve center system, inhibits
DNA replication, and leads to chromosomal aberration [40]. Derivatives of
catechol include the toxic agents in poison ivy and poison oak.
OH
phenol
HO
HO
catechol
OH
Cl Cl 2,4-dichlorophenol
Pesticides
Pesticides are designed to have properties that enable them to reach and impact the
target organism or pest including insects, rodents, and vegetation. Effectiveness is
increased by persistence, but this can result in nonselective toxicity to other non-
targeted organisms including birds and humans. They may have a greater poten-
tial to contaminate a wide range of environmental media. Many of the original
58 S. Gama et al.
pesticides were organochlorine (OC) compounds which are now either banned or
highly restricted under the Stockholm Convention [10]. Many OCs are relatively
volatile; thus, evaporation is an important loss process for pesticides from the
areas where they are applied initially, they then move to the atmosphere and
eventually get transported through air to remote areas. Notable among the OCs are
DDT and its related compounds DDE and DDD, and lindane (g-HCH). Permeth-
rin, a synthetic pyrethroid insecticide used worldwide to control a broad range of
insect pests such as mosquitoes, is also highly hydrophobic and partitions in the
environment in a manner similar to DDT. OC pesticides have been largely
replaced by organophosphates, pyrethroids, and other substances that are less
persistent, but often more potent toxicants. For further reading on pesticides,
especially on their global distribution, see Weber et al. [41], Hoferkamp et al.
[42], Li and Macdonald [43], and Muir and de Wit [44].
Cl
Cl Cl
Cl Cl DDT
Personal Care, Indoor, and Household Products
Most organic substances in personal care and household products can impact
humans directly in the home and they may enter the environment through sewage
wastes (down the drain releases) or by direct releases to air. Included in this group
are detergents, fragrances, flame retardants, additives to plastics, and indoor
pesticides and solvents. An example is the class of cyclic siloxanes whose
backbone is formed of repeated oxygen and silicon atoms with methyl groups
attached. These compounds are components of toiletries such as antiperspirants,
soaps, and shampoos and are used in many other “household” products including
textiles, paints, sealants, lubricants, and non-medicinal ingredients in
pharmaceuticals. They are relatively nontoxic but they can be quite persistent
[45]. They are highly volatile with large KAW values and hence predominantly
partition to air as illustrated in Fig. 3.1. Bisphenol A (BPA) (2,2-bis(4-
hydroxyphenyl) propane) has been used often in the production of hard plastic
(polycarbonate) bottles and metal-based food and beverage can liners in the past
few decades. BPA is regarded as an endocrine-disrupting chemical (EDC) [46].
Regulatory agencies have recently expressed serious concerns about the potential
effects of BPA on the brain and prostate gland in fetuses, infants, and young
children and some have banned the sale and importation of baby bottles
containing BPA [47]. A variety of other organic substances are emitted from
various indoor sources, including office equipment such as printers [48].
Si O
O Si
Si O
O Si
octamethylcyclotetrasiloxane (D4)
Si O
O Si
Si O
O Si
Si O
decamethylcyclopentasiloxane (D5)
Conclusions
There is a wide range of organic chemicals that have the potential to be toxic
with a multitude of uses and with properties that vary greatly resulting in
a diversity of environmental behavior characteristics and exposure routes.
Management of these substances presents a considerable challenge to
governments and industries. Mistakes have been made in the past by
synthesizing and dispersing substances such as DDT and PCBs that have
proven to be persistent, bioaccumulative, toxic and have become globally
distributed. There is optimism that such mistakes can be avoided in the future
if there is an adequate understanding of the factors that dictate environmental
partitioning, fate, exposure, and toxicity. Effort is now being devoted to
searching for new toxic organics by advance knowledge of their structural
and property characteristics [49–51]. It is increasingly possible to estimate
the critical chemical properties from molecular structure and assess environ-
mental fate and transport quantitatively using mass balance models that can be
validated using monitoring data. If this information can be exploited in advance
60 S. Gama et al.
of commercial distribution, it should be possible to enjoy the many benefits of

synthetic organic chemicals without the adverse effects that have been, and are
being experienced, from uses that do not satisfy the principles of conservation
and sustainability.
Future Directions
The effective future management of toxic organic chemicals in the environment can
be improved by a number of scientific and regulatory actions. There is a need for
better information on the production, uses and rates of discharge of chemicals, both
nationally and internationally. Improved methods of estimating chemical properties
from molecular structure are desirable including partitioning, degradability, and
toxicity. These improvements could enhance the assessment of chemical fate and
possible effects by industry, thus avoiding repetition of past mistakes. Undoubtedly,
there are as yet unidentified toxic organic substances, thus there is an incentive to
improve techniques and sensitivity of chemical analyses and monitoring of envi-
ronmental media. As a complement to these monitoring efforts there is a need to
further develop, apply, and validate the use of computer models that predict
environmental fate and effects. Finally there is a strong incentive to increase the
effectiveness of regulatory programs, both nationally and internationally by
incorporating the emerging scientific advances and providing a holistic and trans-
parent evaluation of risk to the public and the ecosystem. Ultimately it is an
informed public that is the best guarantee that society will maintain and enhance
actions to improve sustainability and conservation.
Bibliography
1. Mackay D, Shiu WY, Ma KC (2000) Physical-chemical Properties and Environmental Fate

Handbook. Chapman and Hall CRCnetBASE, CRC Press LLC, Boca Raton
2. Seth R, Mackay D, Muncke J (1999) Estimating the organic carbon partition coefficient
and its variability for hydrophobic chemicals. Environ Sci Technol 33(14):2390–2394
3. Atkinson R, Arey J, Aschmann SM (2008) Atmospheric chemistry of alkanes: review and
recent developments. Atmos Environ 42(23):5859–5871
4. Howard PH, Boethling RS, Jarvis WF, Meylan WM, Michalenko EM (1991) Handbook of
Environmental Degradation Rates. Lewis Publishers, Chelsea
5. Arnot JA, Mackay D, Parkerton TF, Bonnell M (2008) A database of fish biotransformation
rates for organic chemicals. Environ Toxicol Chem 27(11):2263–2270
6. U.S. EPA (2009) Exposure assessment tools and models, estimation programs interface (EPI)
suite, ver. 4.0. U.S. Environmental Protection Agency, Exposure Assessment Branch,
Washington, DC
7. Government of Canada, Persistence and Bioaccumulation Regulations. Canada Gazette Part II
March 29, 2000
8. Webster E, Mackay D, Wania F (1998) Evaluating environmental persistence. Environ

Toxicol Chem 17:2148–2158
9. Wegmann F, Cavin L, MacLeod M, Scheringer M, Hungerbühler K (2009) The OECD
software tool for screening chemicals for persistence and long-range transport potential.
Environ Modell Softw 24(2):228–237
10. UNEP (2001) Final Act of the Conference of Plenipotentiaries on the Stockholm Convention
on Persistent Organic Pollutants. In: United Nations Environment Program, Geneva,
Switzerland, p 44
11. Gouin T, Mackay D, Webster E, Wania F (2000) Screening chemicals for persistence in the
environment. Environ Sci Technol 34(5):881–884
12. Mackay D, McCarthy LS, MacLeod M (2001) On the validity of classifying chemicals for
persistence, bioaccumulation, toxicity and potential for long-range transport. Environ Toxicol
Chem 20(7):1491–1498
13. Scheringer M (2002) Persistence and Spatial Range of Environmental Chemicals. Wiley-VCH,
Weinheim
14. Wania F (2003) Assessing the potential of persistent organic chemicals for long-range
transport and accumulation in polar regions. Environ Sci Technol 37(7):1344–1351
15. Wania F, Mackay D (1995) A global distribution model for persistent organic-
chemicals. Sci Total Environ 161:211–232
16. Arnot JA, Mackay D, Webster E, Southwood JM (2006) Screening level risk assessment
model for chemical fate and effects in the environment. Environ Sci Technol
40(7):2316–2323
17. Guglielmo F, Lammel G, Maier-Reimer E (2009) Global environmental cycling of g-HCH and
DDT in the 1980s – a study using a coupled atmosphere and ocean general circulation model.
Chemosphere 76(11):1509–1517
18. Gobas FAPC, Morrison HA (2000) Bioconcentration and Biomagnification in the Aquatic
Environment. In: Boethling RS, Mackay D (eds) Handbook of Property Estimation Methods
for Chemicals: Environmental and Health Sciences. CRC Press, Boca Raton
19. McCarty LS, Mackay D (1993) Enhancing ecotoxicological modeling and assessment: body
residues and modes of toxic action. Environ Sci Technol 27(9):1719–1728
20. Haritash AK, Kaushik CP (2009) Biodegradation aspects of polycyclic aromatic hydrocarbons
(PAHs): a review. J Hazard Mater 169(1–3):1–15
21. Mastral AM, Callen MS (2000) A review on polycyclic aromatic hydrocarbon (PAH)
emissions from energy generation. Environ Sci Technol 34(15):3051–3057
22. Wilson SC, Jones KC (1993) Bioremediation of soil contaminated with polynuclear aromatic
hydrocarbons (PAHs): a review. Environ Pollut 81(3):229–249
23. Kannan K, Blankenship AL, Jones PD, Giesy JP (2000) Toxicity reference values for the
toxic effects of polychlorinated biphenyls to aquatic mammals. Hum Ecol Risk Assess
6:181–201
24. Hornbuckle K, Robertson L (2010) Polychlorinated Biphenyls (PCBs): sources, exposures,
toxicities. Environ Sci Technol 44(8):2749–2751
25. Fox GA, Weseloh DV, Kubiak TJ, Erdman TC (1991) Reproductive outcomes in colonial fish-
eating birds – a biomarker for developmental toxicants in Great-Lakes food-chains .1. Histori-
cal and ecotoxicological perspectives. J Great Lakes Res 17(2):153–157
26. Hoffman DJ, Rice CP, Kubiak TJ (1996) PCBs and dioxins in birds. In: Beyer WN, Heinz
GH, Redmon-Norwood AW (eds) Environmental Contaminants in Wildlife: Interpreting
Tissue Concentrations. Lewis Publishers, Boca Raton, pp 165–207
27. Borga K, Fisk AT, Hargrave B, Hoekstra PF, Swackhamer D, Muir DCG
(2005) Bioaccumulation factors for PCBs revisited. Environ Sci Technol 39(12):4523–4532
28. Aken BV, Correa PA, Schnoor JL (2009) Phytoremediation of polychlorinated biphenyls: new
trends and promises. Environ Sci Technol 44(8):2767–2776
62 S. Gama et al.
29. Van den Berg M, Birnbaum LS, Bosveld ATC, Brunstrom B, Cook P, Feeley M, Giesy JP,
Hanberg A, Hasegawa R, Kennedy SW, Kubiak T, Larsen JC, van Leeuwen FXR, Liem AKD,
Nolt C, Peterson RE, Poellinger L, Safe S, Schrenk D, Tillitt DE, Tysklind M, Younes M,
Waern F, Zacharewski T (1998) Toxic equivalency factors (TEFs) for PCBs, PCDDs PCDFs
for humans and wildlife. Environ Health Perspect 106(12):775–792
30. Domingo JL, Bocio A (2007) Levels of PCDD/PCDFs and PCBs in edible marine species and
human intake: a literature review. Environ Int 33(3):397–405
31. Lohmann R, Jones KC (1998) Dioxins and furans in air and deposition – a review of levels,
behavior and process. Sci Total Environ 219:53–81
32. Aberg A, MacLeod M, Wiberg K (2008) Physical-chemical property data for dibenzo-p-dioxin
(DD), dibenzofuran (DF), and chlorinated DD/Fs: a critical review and recommended values.
J Phys Chem Ref Data 37(4):1997–2008
33. Hites RA (2004) Polybrominated diphenyl ethers in the environment and in people: a meta-
analysis of concentrations. Environ Sci Technol 38(4):945–956
34. Vonderheide AP, Mueller KE, Meija J, Welsh GL (2008) Polybrominated diphenyl ethers:
causes for concern and knowledge gaps regarding environmental distribution, fate and toxic-
ity. Sci Total Environ 400(1–3):425–436
35. Wang Y, Jiang G, Lam PKS, Li A (2007) Polybrominated diphenyl ethers in the East Asian
environment: a critical review. Environ Int 33(7):963–973
36. de Wit CA, Herzke D, Vorkamp K (2010) Brominated flame retardants in the Arctic environ-
ment – trends and new candidates. Sci Total Environ 408(15):2885–2918
37. Staples CA, Peterson DR, Parkerton TF, Adams WJ (1997) The environmental fate of
phthalate esters: a literature review. Chemosphere 35(4):667–749
38. Peijenburg WJGM, Struis J (2006) Occurrence of phthalate esters in the environment of the
Netherlands. Ecotoxicol Environ Saf 63:204–215
39. Shiu W, Ma K-C, Varhanickova D, Mackay D (1994) Chlorophenols and alkylphenols:
a review and correlation of environmentally relevant properties and fate in an evaluative
environment. Chemosphere 29(6):1155–1224
40. Topping DC, Bernard LG, O’Donoghue JL, English JC (2007) Hydroquinone: acute and
subchronic toxicity studies with emphasis on neurobehavioral and nephrotoxic effects. Food
Chem Toxicol 45:70–78
41. Weber J, Halsall CJ, Muir D, Teixeira C, Small J, Solomon K, Hermanson M,
Bidleman T (2010) Endosulfan, a global pesticide: a review of its fate in the environ-
ment and occurrence in the Arctic. Sci Total Environ 408:2966–2984
42. Hoferkamp L, Hermanson MH, Muir DCG (2010) Current use pesticides in Arctic media;
2000–2007. Sci Total Environ 408(15):2985–2994
43. Li YF, Macdonald RW (2005) Sources and pathways of selected organochlorine pesticides to
the Arctic and the effect of pathway divergence on trends in biota: a review. Sci Total Environ
342(1–3):87–106
44. Muir DCG, de Wit CA (2010) Trends of legacy and new persistent organic pollutants in the
circumpolar arctic: overview, conclusions, and recommendations. Sci Total Environ
408:3044–3051
45. Graiver D, Farminer KW, Narayan R (2003) A review of the fate and effects of silicones in the
environment. J Polym Environ 11(4):129–136
46. Hiroi T, Okada K, Imaoka S, Osada M, Funae Y (2006) Bisphenol A binds to protein disulfide
isomerase and inhibits its enzymatic and hormone-binding activities. Endocrinology
147:2773–2780
47. Meeker JD, Calafat AM, Hauser R (2010) Urinary bisphenol A concentrations in relation to
serum thyroid and reproductive hormone levels in men from an infertility clinic. Environ Sci
Technol 44(4):1458–1463
48. Destaillats H, Maddalena RL, Singer BC, Hodgson AT, McKone TE (2008) Indoor pollutants
emitted by office equipment: a review of reported data and information needs. Atmos Environ
42(7):1371–1388
49. Brown TN, Wania F (2008) Screening chemicals for the potential to be persistent organic
pollutants: a case study of arctic contaminants. Environ Sci Technol 42(14):5202–5209
50. Howard PH, Muir DCG (2010) Identifying new persistent and bioaccumulative organics
among chemicals in commerce. Environ Sci Technol 44(7):2277–2285
51. Muir DCG, Howard PH (2006) Are there other persistent organic pollutants? A challenge for
environmental chemists. Environ Sci Technol 40(23):7157–7166
Chapter 4
Transport in the Environment
John S. Gulliver
Glossary
Adsorption The process of dissolved chemicals sticking to a solid.

Convection The movement of a constituent with movement of the fluid.
Desorption The detachment of a chemical from a solid.
Diffusion The spreading of fluid constituents through the motion
inherent to atoms and molecules.
Diffusion coefficient A coefficient that describes the tendency of molecules to
spread a constituent mass.
Dirac delta An impulse of a given quantity (mass) that occurs over an
infinitely short time or space.
Kinematic viscosity The fluid viscosity divided by the fluid density, resulting in
units that are similar to a diffusion coefficient, or length
squared per time.
Laminar flow Flow that has no turbulent eddies, where the fluid flows in
laminas and diffusion creates the mixing of the fluid.
Retardation factor A divisor that indicates the slowing of chemical movement
through a media due to adsorption.
Reynolds number The ratio of inertial to viscous forces, resulting in
a meaningful velocity times a meaningful distance divided
by kinematic viscosity.
J.S. Gulliver (*)
St. Anthony Falls Laboratory, Department of Civil Engineering, University of Minnesota,
500 Pillsbury Drive S.E., Minneapolis, MN 55454, USA

66 J.S. Gulliver
Definition of Transport in the Environment
In this section various solution techniques for the convection-diffusion equation are
reviewed, which is generally defined as the mass transport equation with diffusive
terms. These techniques will be applied to chemical transport solutions in
sediments. There are also a number of applications to chemical transport in
biofilms. There are many other applications of the convection-diffusion equation,
but they require more background with regard to the physics of mixing
processes, which will be addressed in later sections of the volume.
Introduction
What is mass (or chemical) transport? It is the transport of a solute (the dissolved
chemical) in a solvent (everything else). The solute is the dissolvee and the solvent
is the dissolver. There are liquids that are generally classified as solvents because
they typically play that role in industry. Some examples would be degreasing and
dry-cleaning solvents, such as trichloroethylene (TCE). In environmental
applications, these “solvents” are the solutes, and water or air is usually the solvent.
In fact, when neither water nor air are the solvents, a general term “nonaqueous
phase liquid,” or NAPL, is applied. NAPL is defined as a liquid that is not water,
which could be composed of any number of compounds.
The substance being transported can either be dissolved (part of the same phase
as the solvent) or particulate substances. The diffusion equation will also be
discussed by considering mass conservation in a fixed control volume. The mass
conservation equation can be written as:
Flux rate IN Flux rate OUT þ Rate of

(4.1)
ðSources SinksÞ ¼ Rate of Accumulation
Now that there is our mass conservation equation, it must be decided which control
volume would be the most convenient for our applications. The control volumes used
most for this type of mass balance are given in Fig. 4.1. The general control volume,
given in Fig. 4.1a, is used for descriptive purposes, to maintain generality. It is rare that
one works with something that approximates such a contorted control volume. The
control volumes that are used in practice are given in Fig. 4.1b, c, and d. For the
environmental applications of chemical transport, the rectangular control volume,
Fig. 4.1b, has proven to be the most useful. The cylindrical control volume, Fig. 4.1c,
is used to make pipe or tube flow problems easier to solve, and the spherical control
volume, Fig. 4.1d, is often helpful when dealing with transport in and around particles
4 Transport in the Environment 67
?
θ
?
dx
?
r
x dr
r dθ
General control volume Cylindrical control volume

a c
z ϕ
dz y
r
θ
x
dy
dx
Rectangular control volume Spherical control volume
b d
Fig. 4.1 Common control volumes found in engineering texts and (for the latter three) used in
solving the diffusion equation. (From [1])
or drops. For this control volume, it is convenient to imagine a light being shined along
the axis, which casts a shadow of the vector on to a plane normal to the light. The j
angle is measured from the reference axis to the shadow in this plane.
A rectangular control volume will be used for the development of our mass
conservation (diffusion) equation.
Development of the Diffusion Equation
The diffusion equation will be developed by considering each term in Eq. 4.1
separately. In addition, the flux terms will be divided into diffusive and convective
flux rates.
Diffusive Flux Rate
The molecules of a fluid “at rest” are still moving because of their internal energy.
They are vibrating. In a solid, the molecules are held in a lattice. In a gas or liquid,
they are not, so they move around because of this vibration. Since the molecules are
vibrating in all directions, the movement appears to be random. Diffusive fluxes are
68 J.S. Gulliver
t=0 C
0 < t < infinity t ⇒ infinity
C
x x
a b
Fig. 4.2 Illustration of net diffusive flux through one side of the rectangular control volume.
(From [1])
described by Fick’s law [9], given in the section by Dr. Cussler on diffusion. For
this purpose, let us consider one side of our control volume, normal to the x-axis,
with an area Ax, shown in Fig. 4.2. Fick’s law describes the diffusive flux rate as:
Diffusive flux rate (g/s)

@C (4.2)
¼ D ðm2 /s) ðg/m4 Þ Ax ðm2 Þ
@x
where C is concentration of the solute (tracer), D is the diffusion coefficient of the

solute in the solvent (water), which relates to how fast how and far the tracer
molecules are moving to and fro, and ∂C/∂x is the gradient of concentration with
respect to x, or the slope of C with x, as shown in Fig. 4.2. Thus, the diffusive flux
rate depends upon the diffusion coefficient and the gradient of concentration with
distance.
Convective Flux
The convective flux rate into our control volume is simply the chemical mass
carried in by convection. If the same box of Fig. 4.2 is considered, except with
a velocity component u in the x-direction, the convective flux rate into the box from
the left-hand side is:
Convective flux rate ðg/sÞ ¼ Velocity component

normal to surface ðm/sÞ Surface area ðm2 Þ (4.3)
Concentration (g/m Þ 3
or
Convective flux rate ¼ uAx C (4.4)
where u is the component of velocity in the x-direction and Ax is the surface area
normal to the x-axis on that side of the box. All six sides of our box would have
a convective flux rate through them, just as they would have a diffusive flux.
Rate of Accumulation
The rate of accumulation is the change of chemical mass per unit time, or:
@C
V ðm3 Þ
Rate of accumulation ðg=sÞ ¼ ðg=m3 =sÞ (4.5)
@t
where
V is the volume of our box.
Source and Sink Rates
The solute chemical can appear or disappear through chemical reaction. In

addition, interfacial transfer is often integrated over the control volume and
considered as a source or sink throughout the control volume. This type of
pseudo-reaction can be of significant help in solving chemical transport problems
when averages over a larger control volume, such as cross-sectional mean
concentrations, are being computed. For both cases (chemical reactions and
pseudo-reactions), the source and sink rates are given as:
Source sink rate ðg=sÞ ¼ S (g=m3 =s)

V ðm3 Þ (4.6)
where S is the net source/sink rate per unit volume. The particular reactions that
a given chemical is likely to undergo will determine the form of S used in Eq. 4.6.
These are listed in Table 4.1. The source/sink term could be a combination of two
or more of these reactions. For convenience in determining analytical solutions to
70 J.S. Gulliver
Table 4.1 Common source and sink terms used in the convection-diffusion equation
Source/sink name Equation Units of constant
Zero order S = ko ko – g/m3-s
First order S = k1C k1 = S1
2
Second order S = k2C k2-m3/g-s
Independent variable S = k1i P a
k1i – s1
S = k2i PCb k2i – m3/g-s
Monod kineticsc mm C mm = maximum growth rate (s1)
S¼ P
kc þ C kc = half-saturation coefficient (g/m3)
a
If P is nearly constant, then k1i can be provided as a zero-order term
b
Often called second order
c
Common for biologically mediated reactions
Δz
Δy
Fig. 4.3 Dimension of the

rectangular control volume Δx
the diffusion equation, most source/sink terms are approximated as either a first-
order or zero-order reaction.
Mass Balance on Control Volume
A mass balance on one compound in our box is based upon the principle that
whatever comes in must do one of three things: be accumulated in the box, flux out
of another side, or react in the source/sink terms. If it seems simple, it is.
We will begin by assigning lengths to the sides of our box of dx, dy, and dz, as
shown in Fig. 4.3. Then, for simplicity in this mass balance, we will arbitrarily
designate the flux as positive in the +x-direction, +y-direction, and +z-direction.
The x-direction flux, so designated, is illustrated in Fig. 4.4. Then, the two flux
terms in Eq. 4.1 become:
Flux rate in þ Difference in flux rate ¼ Flux rate out (4.7)

Fig. 4.4 Illustration of the x-

component of mass flux rate
into and out of the rectangular
Flux rate Flux rate
control volume. (From [1])
in = U0C0 out = U1C1
dx
∂(UC)
= Slope
Flux ∂x
rate in
UC
Flux rate
out
dx
0 x 1
or, because a difference can be equated to a gradient times the distance over which
the gradient is applied:
Flux rate out Flux rate in ¼ Gradient in flux rate

(4.8)
Distance
Equation 4.8 can thus be applied along each spatial component as:
@
Flux rate (out inÞx ¼ ðflux rateÞdx (4.9a)
@x
@
Flux rate ðout inÞy ¼ ðflux rateÞdy (4.9b)
@y
@
Flux rate ðout inÞz ¼ ðflux rateÞdz (4.9c)
@z
Convective flux rates. The convective and diffusive flux rates are dealt with
separately. They will eventually be separated in the final diffusion equation, and
it is convenient to make that break now. The x-component of the convective flux
rate is equal to the x-component of velocity times the concentration times the area
of our box normal to the x-axis. Therefore, in terms of convective flux rates,
Eq. 4.9a becomes:
@
Convective flux rateðout inÞx ¼ ðu C Ax Þdx
@x (4.10a)
@
¼ ðu CÞdx dy dz
@x
72 J.S. Gulliver
Because the normal area, Ax = dy dz, of our box does not change with x, it can be
pulled out of the partial with respect to x. This is done in the second part of Eq. 4.10a.
The same can be done with the y- and z-components of the convective flux rate:
@
Convective flux rate (out in)y ¼ ðv C Ay Þdy
@y
(4.10b)
@
¼ ðv CÞdx dy dz
@y
@
Convective flux rate (out in) z ¼ ðw C Az Þdz
@y
(4.10c)
@
¼ ðw CÞdx dy dz
@z
Finally, adding Eqs. 4.10a, 4.10b, and 4.10c results in the total net convective
flux rate.
Net convective flux rate

@ @ @ (4.11)
¼ ðu CÞ þ ðvCÞ þ ðwCÞ dx dy dz
@x @y @z
Diffusive flux rates. For net diffusive flux rate in the x-direction, Eq. 4.9a
becomes:
Diffusive flux rate(out inÞx

@ @C
¼ D Ax dx
@x @x (4.12a)

@ @C
¼ D dx dy dz
@x @x
The y- and z-directions give a result similar to Eq. 4.12a:
Diffusive flux rate (out in)y

@ @C
¼ D Ay dy
@y @y (4.12b)

@ @C
¼ D dx dy dz
@y @y
Diffusive flux rate (out in)z

@ @C
¼ D Az dz
@z @z (4.12c)

@ @C
¼ D dx dy dz
@z @z
Finally, Eqs. 4.12a, 4.12b, and 4.12c can be added to write an equation describ-
ing the net diffusive flux rate (out–in) out of the control volume:
Net diffusive flux rate

@ @C @ @C @ @C
¼ D þ D þ D (4.13)
@x @x @y @y @z @z
dx dy dz
The diffusion coefficient is often not a function of distance, such that Eq. 4.13
can be further simplified by putting the constant value diffusion coefficient in front
of the partial derivative. However, we will also be substituting turbulent diffusion
and dispersion coefficients for D when appropriate to certain applications, and they
are not always constant in all directions. We will therefore leave the diffusion
coefficient inside the brackets for now.
Control volume mass balance. Now Eqs. 4.1, 4.5, 4.6, 4.11, and 4.13 can be
combined into a mass balance on our box for Cartesian coordinates. After dividing
by
V = dx dy dz and moving the diffusive flux terms to the right-hand side, this mass
balance is:
@C @ @ @
þ ðuCÞ þ ðvCÞ þ ðwCÞ
@t @x @y @z
(4.14)
@ @C @ @C @ @C
¼ D þ D þ D þS
@x @x @y @y @z @z
When working with a computational transport code, there is little reason to

further simplify Eq. 4.14. One primary objective of this section, however, is to
develop approximate analytical solutions to environmental transport problems, and
we will normally be assuming that diffusivity is not a function of position, or x, y,
and z. The convective transport terms can be expanded with the chain rule of partial
differentiation:
@ @C @u
ðu CÞ ¼ u þC (4.15a)
@x @x @x
@ @C @v
ðv CÞ ¼ v þC (4.15b)
@y @y @y
@ @C @w
ðw CÞ ¼ w þC (4.15c)
@z @z @z
This may not seem like much help, because we have expanded three terms into
six. However, if the flow is assumed to be incompressible, a derivation given in
fluid mechanics texts (the continuity equation) is:
74 J.S. Gulliver

@u @v @w
r þ þ ¼0 (4.16)
@x @y @z
where r is the density of the fluid. Since Eqs. 4.15a, b, and c are added together in
the mass balance equation, the incompressible assumption means that the terms on
the far right-hand side of these equations will sum to zero, or:
@ @ @ @C @C @C
ðuCÞ þ ðvCÞ þ ðwCÞ ¼ u þ v þ w (4.17)
@x @y @z @x @y @z
The incompressible flow assumption is most always accurate for water in

environmental applications, and is often a good assumption for air. Air flow is
close to incompressible as long as the Mach number (flow velocity/speed of sound)
is below 0.3. A Mach number of 0.3 corresponds to an air flow velocity of
approximately 110 m/s.
Equation 4.14 then becomes
@C @C @C @C
þu þv þw
@t @x @y @z
2 (4.18)
@ C @2C @2C
¼D þ þ 2 þS
@x2 @y2 @z
The only assumptions made in developing Eq. 4.18 are (1) that diffusivity does
not change with spatial coordinate and (2) incompressible flow. Equation 4.18 will
be further simplified in order to develop analytical solutions for mass transport
problems. In some cases, all that needs to be done is orient the flow direction so that
it corresponds with one of the coordinate axes. There would then be only one
convection term.
Adsorption and Desorption in Sediments and Soils
Sorption relates to a compound sticking to the surface of a particle. Adsorption

relates to the process of compound attachment to a particle surface, and desorption
relates to the process of detachment. Sorption processes will now be reviewed
because there are many compounds that are sorptive and subject to spills. Then the
solutions of the diffusion equation can be examined as they apply to highly sorptive
compounds.
Environmental chemicals are generally classified by the Greek terms hydro-
philic (likes water) and hydrophobic (hates water). Water is a polar molecule, in
that it has two hydrogen atoms on one side, and an oxygen atom on the other.
Solutes with a polarity or charge, therefore, will have water molecules
surrounding them with the tendency to have the proper charge of atom adjacent to
the solute. Most amides and alcohols are strongly polar, and also soluble in water.
These are generally hydrophilic compounds. Other organic compounds with
larger molecular weights, especially with aromatic rings, are generally nonpolar
and are classified as hydrophobic compounds. It makes sense that these hydro-
phobic compounds would adsorb to the nonpolar organic material in the sediments
or soils. There are handbooks [7] that can be used to estimate the chemical
thermodynamics of a water-particle system.
How can sorption be handled in our transport equation? For particles that are not
transported with the flow field, like sediments and groundwater flow, we are
interested in the water concentrations. The sorbed portion of the compound is not
in the solute phase, and should not be considered in the transport equation except
when transfer of the compound between the water and particles occur. Adsorption
would then be a sink of the compound and desorption would be a source.
Let us assign Sp to be the mass of chemical sorbed to particles per mass of solids
contained in our control volume, and C to be the concentration of the compound in
solution. Then, the source term in the diffusion equation is equal to the rate of
change of mass due to adsorption and desorption per unit volume, or:
rb @Sp
S¼ (4.19)
e @t
where rb is the bulk density of the solid (mass of solid/volume of fluid and solid), e
is the porosity of the media (volume of fluid/volume of fluid and solid), and ∂Sp/∂t
is the rate of sorption relative to the mass of solid (mass adsorbed/mass of solid/
time). If the sorption rate is negative, desorption is occurring. The units of S in
Eq. 4.19 are mass adsorbed/volume of fluid/time. This is similar to the units for
the ∂C/∂t term, which are a change of mass/volume of fluid/time.
The source term in Eq. 4.19 requires a separate differential equation for Sp,
which would incorporate the concentration of the compound in solution. There
would thus be two equations that need to be solved simultaneously. However,
most sorption rates are high, relative to the transport rates in sediments and soil.
Thus, local equilibrium in adsorption and desorption is often a good assumption.
It also simplifies the solution to a transport problem considerably. If that assump-
tion is made, Sp changes in proportion to C alone, or:
Sp ¼ Sp ðCÞ (4.20)
and
@Sp @Sp @C
¼ (4.21)
@t @ C @t
Now, if Eq. 4.21 is substituted into 4.19, we get:

76 J.S. Gulliver
rb @Sp @C
S¼ (4.22)
e @C @t
The ∂Sp/∂C term can be found from the equilibrium relationship of Freundlich
isotherms, expressed as:
Sp ¼ K d C b (4.23)
where Kd is an equilibrium-partitioning coefficient between the fluid and sorption to

the solid and b is a coefficient fit to measured data. Then,
@Sp
¼ b Kd Cb1 (4.24)
@C
At the lower concentrations normally found in the environment, b = 1 is a valid
assumption. Then Eq. 4.24 becomes
@Sp
¼ Kd ðb ¼ 1Þ (4.25)
@C
Substituting Eq. 4.25 into 4.22 now results in a source term that no longer
contains the variable Sp, and keeps the partial differential equation (PDE) of our
mass balance linear:
rb @C
S¼ Kd (4.26)
e @t
Now, if Eq. 4.26 is substituted into our mass transport Eqs. 4.15a–c for the
source term, the result is a PDE where the only dependent variable is C:
@C @C @C @C
þu þv þw
@t @x @y @z
2 (4.27)
@ C @ C @2C
2
r @C
¼D þ þ 2 b Kd
@x2 @y2 @z e @t
or
r @C @C @C @C
1 þ b Kd þu þv þw
e @t @x @y @z
2 (4.28)
@ C @2C @2C
¼D þ þ 2
@x2 @y2 @z
If we divide Eq. 4.28 by the term (1 + Kdrb/e), we can see that all convective and
diffusive transport is retarded by equilibrium adsorption and desorption. Thus,
a retardation factor is defined:
R ¼ retardation factor ¼ 1 þ Kd rb =e (4.29)

and Eq. 4.28 becomes:
@C u @C v @C w @C
þ þ þ
@t R @x R @y R @z
(4.30)
D @2C @2C @2C
¼ þ þ
R @x2 @ y2 @ z2
Equation 4.30 indicates that as long as it can be assumed that the sorption rates
are fast compared to our transport rates and the equilibrium partitioning is linearly
related to concentration, the retardation factor can utilize and simply convert all of
the transport terms through dividing by R. Thus, if there is a spill into the ground-
water table that is highly hydrophobic, it would transport through the soil more
slowly than one which is hydrophilic. Both the convective and the diffusive flux
would be “retarded” for the hydrophobic compound. If both hydrophilic and
hydrophobic compounds are contained in the spill, the hydrophilic compound
would show up first at a downstream location. The similarity to the manner in
which a chromatographic column separates compounds is not fortuitous, because
the column is separating compounds through their sorption to the column’s media.
Determination of Kd from octanol-water partitioning coefficient. There have been
a number of empirical equations developed to determine the water-solid
partitioning coefficient, Kd [7]. These are primarily for the many organic chemicals
that exist in the environment, usually due to human impacts. Many of them use the
octanol-water partitioning coefficient for the compound as an indicator of
hydrophobicity. Octanol is a relatively insoluble organic compound. Since most
organic compounds tend to adsorb to the organic portion of the particles,
a hydrophobic organic compound placed in an octanol-water solution will tend
toward the octanol. The ratio of concentration in the octanol over concentration in
the water will indicate the degree of the hydrophobicity. It is a straightforward and
relatively easy measurement to make, so most organic compounds of interest in the
environment have an octanol-water partitioning coefficient that has been measured.
Karikhoff et al. [2] developed a simple empirical equation for equilibrium
partitioning of organic compounds that will be used in this text (other equations
are given in Lehman et al. [7]):
Kd ¼ bf Kow (4.31)
where Kow is the dimensionless octanol-water partitioning coefficient, f is the

fraction of soil that is organic matter (usually from zero in sand to 0.01 in sandy
soil to 0.10 in muck), and b is an empirical coefficient, estimated by Karikhoff to be
0.41 cm3/g. It is generally the organic matter in the medium to which organic
compounds adsorb, hence the use of organic fraction.
The other parameters required to compute a retardation coefficient are the bulk
density, rb, and the porosity of the media, e. The bulk density of the water and soil is
typically 1.6–2.1 g/cm3. The porosity of the soil or sediments is typically 0.2–0.4.
Thus rb/e is typically between 4 and 10 g/cm3.
78 J.S. Gulliver
Fig. 4.5 Illustration of

dissolved oxygen profile in
lake sediments. (From [1])
C0
Water
Sediment
z z = 1 mm
Example Applications of the Diffusion Equation
The first application of the diffusive is transport of oxygen into lake sediments and
the use of oxygen by the bacteria to result in a steady-state oxygen concentration
profile.
Example 1: Steady O2 concentration profile in lake sediments (steady-state
solution with a first-order sink) Given a concentration, Co, in the overlying
water, and a first-order sink of oxygen in the sediments, develop an equation to
describe the dissolved oxygen concentration profile in the sediments (Fig. 4.5).
Assume:
– Steady: @@t ) 0
– No flow: u; v; w ) 0
Small horizontal variation: @@zC2 @@xC2 ; @@yC2
2 2 2
–
– No sorption: R = 1 (accurate for O2 in sediments)
– First-order sink: S = kC, where k is a rate constant
Then, the diffusive mass transport Eqs. 4.15a–c becomes:
@2 C
0¼D kC
@z2
or, since C = C(z)
d2 C
0¼D kC
@z2
A solution to this equation requires two boundary conditions because it is

a second-order equation. These two are:
B.C.#1: @ z = 0, C = C;0
B.C.#2: @ z ! 1, C ) 0
This solution may be achieved by: (1) separating variables and integrating or
(2) solving the equation as a second-order, linear ordinary differential equation
(ODE). The latter will be used since the solution technique is more general.
d
1. Assign l to be the dz operator. Then, the equation becomes

k
l2 C¼0
D
2. Solve for l
pffiffiffiffiffiffiffiffiffi
l¼ k=D
3. The solution, developed in texts on solving ordinary differential equations [6], is

C ¼ b1 el1 z þ b2 el1 z l1 ¼ þ k=D
l2 ¼ k=D
4. b1 and b2 are determined from boundary conditions

Apply B.C. #2:
pffiffiffiffiffiffi pffiffiffiffiffiffi
k=D1
C ¼ 0 ¼ b1 e þ b2 e k=D 1
This is only possible if b1 = 0. Apply B.C. #1:
C0 ¼ 0 þ b2 e0 ¼ b2
Thus, the solution is:

pffiffiffiffiffiffi
C ¼ Co e k=Dz
which is plotted in Fig. 4.6.

At steady state, the oxygen profile is a balance between diffusion from the sediment
surface and bacterial use of oxygen in the sediments. If the sediments are mostly sand,
the depth of the layer with oxygen can be 10 cm or more. If the sediments have
a substantial organic content (like a mud), the aerobic layer (>0.1 g/m3 oxygen
concentration) can be less than 1 mm.
Example 2: Unsteady dissolution of a highly soluble pollutant (Herbicides,
Pesticides, Ammonia, Alcohols, etc.) into groundwater (unsteady, one-dimensional
solution with pulse boundary conditions) A tanker truck carrying a highly
soluble compound in Mississippi tried to avoid an armadillo at night, ran
80 J.S. Gulliver
Fig. 4.6 Solution to Concentration (C/Co)

Example 1 0.0 0.2 0.4 0.6 0.8 1.0
0
1
2
3
Depth (mm)
4
k = 0.1 / hr
5 k = 1 / hr
k = 10 / hr
6 k = 100 / hr
7
8
9
10
Tanker truck spill

Initial conditions C = C0
fast h Spill
Z Water
Water table slow table
C= 0
Mathematical representation
C
∇
Dirac
t= 0
Z
t
Fig. 4.7 Illustration of the tanker truck spill. (From [1])
off the interstate at a high speed, turned over in the drainage ditch, and
spilled a soluble compound. The compound has infiltrated into the ground,
and much of it has reached and temporarily spread out over the groundwater
table, as illustrated in Fig. 4.7. As part of a spill response team, you need to
estimate the groundwater contamination. Predict concentrations over time in
the groundwater table.
The mass transport equation for this example is:
@C @C @C @C
þu þv þw
@t @x dy @z
2
@ C @ C @2C
2
¼D þ þ þS
@x2 @y2 @z2
Assume:
1. Minimal horizontal variations
@C @ 2 C @C @2C
0 ffi ¼ 2 ffi ffi
@x @x @y @y2
2. No flow in the vertical direction, w = 0

3. No reactions, including adsorption and desorption, such that S = 0.
Then with these three assumptions, the governing equation becomes:
@C @2C
¼D 2
@t @z
The initial conditions will be simulated with these boundary conditions:

1. The mass of chemical is assumed to be spread instantaneously across a very thin
layer at t = 0 (a Dirac delta in z and t). At z = 0+, t = 0, the total mass = M and the
total surface area is A.
2. At z ) 1, C ) 0.
The above equation, with boundary conditions (1) and (2), has the solution:
2 M=A
C ¼ pffiffiffiffiffiffiffiffiffiffiffiffi ez =4 Dt
2
4 p Dt
What does the solution look like? The solution can be made dimensionless by
assuming that the initial thickness of the spill layer is Dh. Then, a new variable z = k
Dh will be used in assigning:
Dh
¼ pffiffiffiffiffiffiffiffi
4Dt
with
CA Dh
C ¼
2M
82 J.S. Gulliver
Fig. 4.8 Solution to the Concentration (C*)

tanker truck spill illustrating 0.00 0.02 0.04 0.06 0.08 0.10
groundwater concentration 0
versus distance at various
times. Dh = 2 mm, D = 6
104 mm2/s 20 t = 1 day
t = 2 days
t = 3 days
Depth (mm)
40 t = 7 days
t = 14 days
60
80
100
Substituting these equations into the solution gives:
2
C ¼ pffiffiffi e ðkÞ
p
which is plotted versus depth at various times in Fig. 4.8. The concentration at z = 0
decreases as the initial mass is diffused. At low values of time, the concentration at
and close to z = 0 is strongly dependent upon the Dh chosen. At larger times and
deeper depths, however, this dependency decreases, and the solution becomes
independent of Dh.
It is interesting to note that the solution is very similar to a Gaussian probability
distribution, with the following relationship for P(z):
1 2
PðzÞ ¼ pffiffiffiffiffiffi eðzzm Þ =2s
2
s 2p
where zm is the depth of the maximum concentration (or the center of concentration
mass).
Comparing the probability distribution and the solution to this problem, we can
see that:
2 s2 , 4 Dt
or:
D ¼ s2 =2t
Note that if we measure s, we can determine D.

1,2, Di Cl Benzene Volatilization

(solvent) Cl
[DCB] Cl
Diffusion
Kd = 15 cm3/g
[DCB] rapidly reaches a steady concentration in lake waters, due to volatilization:

Accum. → 0 Inflow + Accum. = Volatilization + Diff into sediments
Inflow = volatilization
Assumed Actual
C
t 1yr 2yr
Co
Water
Sediment
C
z
Fig. 4.9 Illustration of concentration front moving down into the sediments of a lake. (From [1])
Example 3: Dichlorobenzene concentration in lake sediments due to a plating

facility discharge (solution to a concentration front) Sometimes the boundary
conditions can be approximated as a step in concentration. This difference in
boundary conditions changes the solution from one which is related to pulse
boundaries (known mass release) to one resulting from a concentration front with
a known concentration at one boundary.
For many years, a plating facility for a telecommunications company let their
rinse waters flow into an adjacent lake. The compounds used in their rinse
included dichlorobenzene, which is a semi-volatile compound that also has
a fairly high tendency to adsorb to organic compounds in the sediments. Within
a few years of the plating facility opening, the dichlorobenzene concentration
reached a steady-state value in the lake waters as illustrated in Fig. 4.9. Estimate
the buildup of dichlorobenzene in the sediments during the 50 years since the
facility opened until it stopped discharging its untreated waste water.
Assumptions:
1. Biodegradation is small. ∴ S ) 0 except for sorption.
2. Variation in x and y are small
@2C @2C @2C

;
@x2 @y2 @z2
84 J.S. Gulliver
3. No flow in sediments under lake: u = v = w = 0

4. D ffi 6 1010 m2/s
5. rb r
¼ 6:3 ) R ¼ 1 þ b Kd ¼ 96
e e
Then the diffusion equation for the sediments becomes:
@C D @ 2 C
¼
@t R @z2
with boundary conditions:

(a) t > 0, z = 0; C = C0
(b) t = 0, z 6¼ 0; C = 0
There are three known techniques to solve this governing equation:
Laplace transforms, Fourier transforms, and change of variables, which
incorporates both luck and skill. We will use change of variables:
Assign ¼ pffiffiffiffiffiffiffiffiffi
z ffi
4 Dt=R
@C @C @ 1 z @C @C
¼ ¼ pffiffiffiffiffiffiffiffiffiffiffiffi ¼
@t @ @t 4 Dt=Rt @ 2t @
@C @C @ 1 @C
¼ ¼ pffiffiffiffiffiffiffiffiffiffiffi
@z @ @z 2 Dt=R @

@2C @ @C @ R @2C
¼ ¼
@z2 @ @z @z 4Dt @2
Then the governing equation becomes:

@C D R @ 2 C
þ ¼0
2t @ R 4Dt @2
or
d2 C dC
2
þ 2 ¼0 (4.32)
d d
0
0
d þ 2C ¼ 0
This equation may be written as: dC
where C0 ¼ dC or
d
1
dC0 ¼ 2 d
C0
We can integrate this:
‘n C0 ¼ 2 þ bo
or
C0 ¼ ebo e ¼ b1 e

2 2
Now integrate again:
Z Z
2
ef df þ b2
2
C ¼ b1 e d þ b2 ) b1
0 0
Now, note that the error function is given as:
Z
2
ef df
2
erf ðÞ ¼ pffiffiffi
p
0
and the complementary error function is erfc() = 1erf(). Values of the error
function and complimentary error function for various values of Z may be found in
an Internet search. The error function is designed such that erf(1) = 1, erfc(1) = 0,
erf(0) = 0, and erfc(0) = 1. The solution may therefore be written as:
C ¼ b1 erf ðÞ þ b2
Now we need to determine our boundary conditions in terms of :

1. t > 0, z = 0, = 0, C = C0
2. t = 0, z = 0, = 1, C = 0
Checking other boundary conditions:
t ! 1, ) 0, C = C0
z ! 1, ) 1, C = 0
Now, at = 0, C = C0, thus:
C0 ¼ b1 0 þ b2
or
b2 ¼ C0
At = 1, C = 0
86 J.S. Gulliver
Fig. 4.10 Illustration of the Concentration (C/Co)

effect of the change of 0.0 0.2 0.4 0.6 0.8 1.0
variables used in Example 3. 0
Eta = 0.2
0.4
0.6
0.8
Eta
1
1.2
1.4
1.6
1.8
2
Table 4.2 Penetration of dichlorobenzene into the sediment over time

Time, z 1 year, C/C0 4 years, C/C0 10 years, C/C0 50 years, C/C0
1 mm 0.96 0.98 0.988 0.994
1 cm 0.62 0.803 0.87 0.94
10 cm 0 0.015 0.11 0.48
20 cm 0 0 0.01 0.16
30 cm 0 0 0 0.03
100 cm 0 0 0 0
0 ¼ b1 1 þ C0
or
b1 ¼ C0
Then, our solution is:

!!
z
C ¼ C0 1 erf pffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ C0 erfcðÞ (4.33)
4 Dt=R
which is illustrated in Fig. 4.10.

We will now apply Eq. 4.33 to estimate the dichlorobenzene penetration versus
time from spillage. The results are given in Table 4.2, which gives the interstitial
dichlorobenzene concentrations.
The total concentration (TDCB(z)) includes compound adsorbed to the
sediments.
TDCBðzÞ ¼ CðzÞ þ ð1 eÞrs S

where e = porosity
1e = % by volume sediment ffi 0.6
rs = density of sediment ffi 2.5 g/cm3
S = concentration of sorbed compound (g DCB/g sediment)
Since S = Kd C1, the above equation becomes:
TDCBðzÞ ¼ CðzÞð1 þ ð1 eÞ rs Kd Þ
¼ CðzÞð1 þ 0:6ð2:5 g=cm3 Þ(15 cm3 =g)
or
TDCB ¼ 23:5 CðzÞ
Thus, the total dichlorobenzene per volume of sediment and water would be 23.5
times the concentrations given in Table 4.2.
Conclusion
The purpose of this section of the volume is to introduce the reader to the
equations and mathematics used in developing approximate solutions (due to
simplified boundary conditions) to convection-diffusion processes. One may say
that, with computational capabilities, there is no longer any need to develop these
approximate solutions. However, these approximate solutions are useful in the
following manners:
1. A quick, back of the envelope solution is always much quicker and more reliable
than a computational solution. Computational solutions require substantial time
to develop and are often wrong until they are fully vetted.
2. A computational solution always requires vetting, which means that
a computational solution is compared to an analytical solution, hopefully in
a similar condition with simplified boundary conditions. This means that some
analytical solution is always needed, and as close to the real simulation as
possible, to make sure that the computational solution is doing what the user
desires.
3. Developing analytical solutions are an excellent means of getting a feel for
solutions to the convection-diffusion equations. It is a knowledge-building
practice that is difficult to surpass.
Note that the examples given in this section do not include any with convection.
That is because convection in the environment most always includes either turbu-
lence (surface waters and the atmosphere) or dispersion (groundwater). These will
be dealt with in other sections.
88 J.S. Gulliver
Future Directions
Most of the future directions with regard to transport in the environment (without
turbulent transport) will involve transport across interfaces, such as the air–water
and solid–fluid interfaces. While research has been conducted on describing the
predominant transport mechanisms for these two cases (McCready et al. [8], [3–5]),
there is more to be done. An especially vexing problem is transport in the vadose
zone of soils (unsaturated zone). The multiplicity of three interfaces, air, water, and
soil, and the heterogeneities in the soil make this a complex problem to handle in
a deterministic manner. However, meaningful relationships for an effective diffu-
sion coefficient still need to be developed.
Bibliography
Primary Literature
1. Gulliver JS (2007) An introduction to chemical transport in the environment. Cambridge

University Press, Cambridge
2. Karikhoff SW, Brown DS, Scott TA (1979) Sorption of hydrophobic pollutants on natural
sediments. Water Res 13:241
3. Brunley BH, Jirka GH (1987) Near-surface turbulence in a grid-stirred tank. J Fluid Mech
183:235–263
4. Campbell JA, Hanratty TJ (1982) Mass transfer between a turbulent fluid and a solid boundary:
linear theory. AICHE J 28:988
5. Janzen J, Jirka H, Jirka G, Schulz H, Gulliver JS (2010) Estimation of mass transfer velocity
based on measured turbulence parameters. Am Inst Chem Eng J 56(8):2005–2017
6. Kreyszig E (1982) Advanced engineering mathematics, 4th edn. Wiley, New York
7. Lehman WJ, Reehl WF, Rosenblatt DH (1990) Handbook of chemical property estimation.
American Chemical Society, Washington, DC
8. McCready MA, Vassiliadou E, Hanratty T (1986) Computer simulation of turbulent mass
transfer at a mobile interface. AICHEJ 32(7):1108
9. Fick AE (1855) Ueber Diffusion, Annelen der Physik 170(1):59–86
Books and Reviews
Crank J (1975) The mathematics of diffusion, 2nd edn. Oxford University Press, Oxford
Cussler EL (1997) Diffusion: mass transfer in fluid systems, 2nd edn. Cambridge University Press,
Cambridge
Chapter 5
Turbulent Transport
John S. Gulliver
Glossary
Diffusion The spreading of fluid constituents through the motion

inherent to atoms and molecules.
Diffusion coefficient A coefficient that describes the tendency of molecules to
spread a constituent mass
Dirac delta An impulse of a given quantity (mass) that occurs over an
infinitely short time or space.
Kinematic viscosity The fluid viscosity divided by the fluid density, resulting in
units that are similar to a diffusion coefficient, or length
squared per time.
Laminar flow Flow that has no turbulent eddies, where the fluid flows in
laminas and diffusion creates the mixing of the fluid.
Prandtl’s mixing The mean length that the turbulence in the flow will trans-
length port mass, momentum, or energy.
Reynolds number The ratio of inertial to viscous forces, resulting in
a meaningful velocity times a meaningful distance divided
by kinematic viscosity.
Turbulent diffusion The mixing of fluids through turbulent eddies created by
convection.
J.S. Gulliver
Department of Civil Engineering, University of Minnesota,
500 Pillsbury Drive S.E., Minneapolis, MN 55454, USA

90 J.S. Gulliver
Turbulent diffusion A coefficient that comes from the multiplication of two

coefficient turbulent velocities of the flow, divided by density of the
fluid. The coefficient’s location in the mass transport equa-
tion is similar to diffusion coefficients, and the units are
similar; so it is called a “turbulent diffusion coefficient.”
Definition of Turbulent Transport in the Environment
It is fairly safe to state that, except for flow through porous media, the environment
experiences turbulent flow. To emphasize this point, the constriction of a water flow
or airflow that would be required will be considered to have the other option,
laminar flow.
An experimentally based rule of thumb is that laminar flow typically occurs
when the pipe Reynolds number, Vd/n, is less than roughly 2,000, or when an open-
channel Reynolds number, Vh/n, is less than roughly 500, where V is the cross-
sectional mean velocity, d is the pipe diameter, n is the kinematic viscosity of the
fluid, and h is the channel depth. The diameter or depth that would not be exceeded
to have laminar flow by these experimental criteria is given in Table 5.1.
Table 5.1 shows that with the boundary conditions present in most environmen-
tal flows, that is, the earth’s surface, ocean top and bottom, river or lake bottom,
etc., turbulent flow would be the predominant condition. One exception that is
important for interfacial mass transfer would be very close to an interface, such as
air–solid, solid–liquid, or air–water interfaces, where the distance from the inter-
face is too small for turbulence to occur due to the high viscous dissipation. Because
turbulence is an important source of mass transfer, the lack of turbulence very near
the interface is also significant for mass transfer, where diffusion once again
becomes the predominant transport mechanism.
Table 5.1 Maximum diameter or depth to have laminar flow, with the transition Reynolds
number for a pipe at 2,000
Water (n = 106 m2/s) Air (n 2 105 m2/s)
V (m/s) D (m) h (m) d (m)
10 2 104 5 105 0.004
3 7 104 1.5 104 0.014
1 0.002 0.0005 0.04
0.3 0.007 0.0015 0.14
0.1 0.02 0.005 0.4
0.03 0.07 0.015 1.4
0.01 0.2 0.05 4.0
5 Chemicals in the Environment, Turbulent Transport 91
Fig. 5.1 Turbulent eddies superimposed on a temporal-mean velocity and temporal-mean

concentration profiles (From Gulliver [1])
C
u C
u
Time Time
Fig. 5.2 Time traces of typical measurements of velocity and concentration in a turbulent flow
Introduction
What is turbulent flow? The simple illustration of a free-surface flow given in

Fig. 5.1 is used to describe the essential points of the turbulence phenomena.
Turbulent open-channel flow can be described with a temporal mean velocity
profile which reaches a steady value with turbulent eddies superimposed upon it.
These turbulent eddies are continually moving about in three dimensions, only
restricted by the boundaries of the flow, such that they are eliminated from the
temporal mean velocity profile, u in Fig. 2.1. It is this temporal mean velocity profile
that is normally sketched in turbulent flows.
There will also be a temporal mean concentration. If there is a source or sink in
the flow, or transport across the boundaries as in Fig. 5.1, then the temporal mean
concentration profile will eventually reach a value such as that given in Fig. 5.1.
This flux of compound seems to be from the bottom toward the top of the flow.
Superimposed upon this temporal mean concentration profile will be short-term
variations in concentration caused by turbulent transport. The concentration profile
is “flatter” in the middle of the flow because the large turbulent eddies that transport
mass quickly are not as constrained by the flow boundaries in this region. Now,
if a concentration-velocity probe is placed into the flow at one location, the two
traces of velocity and concentration versus time would look something like that
shown in Fig. 5.2.
Turbulent diffusion is thus not really diffusion, but the mixing of chemicals
through turbulent eddies created by convection. Turbulent diffusion is thus a form
of convection. Although it has the appearance of diffusion in the end, that is,
92 J.S. Gulliver
random mixing similar to diffusion, the causes of diffusion and turbulent diffusion
are very different. Since the end products are similar, diffusion coefficients and
turbulent diffusion coefficients are often simply added together.
It is convenient to divide the velocity and concentration traces into temporal
mean values and fluctuating components:
u ¼ u þ u0 (5.1)
and
C ¼ C þ C0 (5.2)
where u is the temporal mean velocity at a point location, u 0 is the fluctuating

component of velocity (variable over time),C is the temporal mean concentration at
a point location, and C 0 is the fluctuating concentration component of concentration
which is also variable over time. Formal definitions of u and C are as follows:
ZDt
1
u ¼ u dt (5.3)
Dt
0
and
ZDt
1
C ¼ C dt (5.4)
Dt
0
where Dt is long compared to the time period of the oscillating components.
Mass Transport Equation with Turbulent Diffusion Coefficients
In this section the most common equations for dealing with mass transport in
a turbulent flow will be derived. Beginning with the mass transport equation
developed in the entry “▾Transport in the Environment,”
@C @ðuCÞ @ðvCÞ @ðwCÞ

þ þ þ
@t @x @y @z
(5.5)
@ @C @ @C @ @C
¼ D þ D þ D þS
@x @x @y @y @z @z
the temporal mean of the entire equation will be taken and eventually one will end
up with an equation that incorporates turbulent diffusion coefficients.
In a turbulent flow field, Eq. 5.5 is difficult to apply because C, u, v, and w are all
highly variable functions of time and space. Osborne Reynolds [2] reduced the
complexities of applying Eq. 5.5 to a turbulent flow by taking the temporal mean of
each term (e.g., the entire equation). Then, the mean value of a fluctuating compo-
nent will be equal to zero, or
@C @ðC þ C0 Þ @ C @C0 @ C
¼ ¼ þ ¼ þ0 (5.6)
@t @t @t @t @t
Equation 5.6, the change of a temporal mean over time, may seem like a misno-
mer, but it will be left in to identify changes in C over a longer time period than Dt.
Continuing,
@C @ðC þ C0 Þ @ C @C0 @ C
¼ ¼ þ ¼ (5.7)
@x @x @x @x @x
@C @ðC þ C0 Þ @ C @C0 @ C
¼ ¼ þ ¼ (5.8)
@y @y @y @y @y
@C @ðC þ C0 Þ @ C @C0 @ C
¼ ¼ þ ¼ (5.9)
@z @z @z @z @z
However, the temporal mean value of two fluctuating components, multiplied by

each other, will not necessarily be zero:
u0 C0 6¼ u0C0 (5.10)
This is similar to a least-square regression, where the mean error is zero, but the
sum of square error is not. The x-component of our convective transport terms will
be dealt with first:
u þ u0 ÞðC þ c0 Þ ¼ uC þ uC
uC ¼ ð 0 þ u0C þ u0 C0 (5.11)
Three of the four terms in Eq. 5.11 may be reduced to something known:
uC ¼ uC (5.12)
0¼0
uC (5.13)
u0C ¼ 0 (5.14)
but, the fourth term will take some additional consideration, because it is not equal
to zero:
u0 C0 6¼ 0 (5.15)
94 J.S. Gulliver
By inference, the following can be written for all three convective transport
terms:
uC ¼ uC þ u0 C0 (5.16)
vC ¼ vC þ v0 C0 (5.17)
and
wC ¼ wC þ w0 C0 (5.18)
Finally, applying continuity ðu þ v þ w ¼ 0Þ to Eq. 5.5 and taking the temporal
mean results of Eqs. 5.6, 5.7, 5.8, 5.9, 5.16, 5.17, and 5.18
@ C @ C @ C @ C
þ u þ v þw
@t @x @y @z

@ @ @ @ @ C
¼ u0 C0 v0 C0 w0 C0 þ D (5.19)
@x @y @z @x @x

@ @C @ @C
þ D þ D þ S
@y @y @z @z
where the turbulent convective transport term can be moved to the right-hand side,
because the concentration distribution that results from these terms looks similar to
diffusion.
With this temporal mean process, we have reduced the terms for which we will
have difficulty defining boundary conditions in turbulent flow fields from seven in
Eq. 5.5 to three in Eq. 5.19. We will now deal with these three terms.
The diffusion equation is a useful and convenient equation to describe mixing in
environmental flows, where the boundaries are often not easily defined. It also lends
itself to analytical solutions and is fairly straightforward in numerical solutions.
Although an alternative technique for solutions to mixing problems is the mixed
cell method described in the entry “▾Chemicals in the Environment, Dispersive
Transport,” there are complications when applied to multiple dimensions and to
flows that vary with space and time. Finally, we are comfortable with the diffusion
equation, so we would prefer to use that to describe turbulent mixing if possible.
Therefore, let us consider the following thought process: if the end result of
turbulence, when visualized from sufficient distance, looks like diffusion with
seemingly random fluctuations, then we should be able to identify the terms causing
these fluctuations in Eq. 5.19. Once identified, they can be related to a “turbulent
diffusion coefficient” that describes the diffusion caused by turbulent eddies.
Looking over the terms in Eq. 5.19 from left to right, we see an unsteady term,
three mean convective terms, the three “unknown” terms, the diffusive terms and
the source/sink rate terms. The “unknown” terms are the only possibility to describe
turbulent diffusion.
In the late nineteenth century, Boussinesq [3] probably went through something
similar to the thought process described above. The end result was the Boussinesq
eddy diffusion coefficient:
@ C
u0 C 0 ¼ e x (5.20a)
@x
@ C
v0 C0 ¼ ey (5.20b)
@y
@ C
w0 C0 ¼ ez (5.20c)
@z
where ex, ey , and ez are the turbulent (or eddy) diffusion coefficients, with units of
m2/s similar to the (molecular) diffusion coefficients.
Then Eq. 5.19 with Eqs. 5.20a, 5.20b, and 5.20c becomes

@ C @ C @ C @ C @ @ C
þ u þ v þw ¼ ðD þ ex Þ
@t @x @y @z @x @x
(5.21)
@ @ C @ @ C
þ ðD þ ey Þ þ ðD þ ez Þ þS
@y @y @z @z
Turbulent diffusion is created by the flow field, which can vary with distance.
Hence, turbulent diffusion coefficient cannot be assumed constant with distance.
Removing that assumption leaves turbulent diffusion coefficient inside of the
brackets.
Character of Turbulent Diffusion Coefficients
A turbulent eddy can be visualized as a large number of differently sized rotating

spheres or ellipsoids. Each sphere has sub-spheres, and so on until the smallest eddy
size is reached. The smallest eddies are dissipated by viscosity, which explains why
turbulence does not occur in narrow passages: there is simply no room for eddies
that will not be dissipated by viscosity.
The cause of the rotation is shear forces created by solid boundaries or variations
in velocity lateral to the primary flow direction. A buoyant plume of smoke or
steam, for example, will have a temporal mean velocity profile develop laterally to
the plume, as the rising plume mixes with the ambient air. Turbulent eddies are
formed by this velocity gradient, and can be seen at the edge of the smoke or steam
plume. The magnitude of turbulent diffusion coefficients is primarily dependent
96 J.S. Gulliver
Fig. 5.3 Character

of turbulent diffusion Distance an eddy
coefficients encompasses
(From Gulliver [1]) increases ε
ε large > ε small
Speed at which
it spins also
increases ε
ε fast > ε slow
upon the scale of turbulent eddies and the speed of the eddy rotation. As illustrated
in Fig. 5.3, a large eddy will have greater eddy diffusion coefficient than a small
eddy because it will transport a compound (or solute) farther in one rotation.
Likewise, a faster spinning eddy will have a larger eddy diffusion coefficient than
one which is the same size but spinning more slowly because the solute simply gets
there faster. These two facts provide meaning to the following observations:
1. The largest scale of turbulence is roughly equal to the smallest overall scale of the
flow field. This may be seen in comparing the size of eddies at the edge of the smoke
or steam plume to the width of the plume.
2. The rotational eddy velocity is roughly proportional to the velocity gradient
times the eddy scale.
3. Eddy size decreases near boundaries to the flow field. Since the eddy size is zero
at a solid boundary, and often close to zero at a fluid density interface (like an
air–water interface), the turbulent eddy size has to decrease as one approaches
the boundaries. In addition, since the flow cannot go through a boundary, the
largest eddy size cannot be greater than the distance from the center of the eddy
to the boundary.
4. Turbulent diffusion occurs because turbulent eddies are transporting mass,
momentum, and energy over the eddy scale at the rotational velocity. This
transport rate is generally orders of magnitude greater than the transport rate
due to molecular motion. Thus, when a flow is turbulent, diffusion is normally
ignored because e D. The exception is very near the flow boundaries, where
the eddy size (and turbulent diffusion coefficient) decreases to zero.
Thus, what influences the velocity and scale of eddies? For the most part, it is the
velocity gradients and scale of the flow. Velocity gradients are the change in
velocity over distance. If we have a high velocity, we typically have a large velocity
gradient somewhere in the flow field. At solid walls, for example, the velocity must
go to zero. Thus, the large velocity difference results in large velocity gradients,
which results in faster spinning eddies and a larger turbulent diffusion coefficient.
This process is illustrated in Fig. 5.4.
Fig. 5.4 Eddy formation at

the edge of a jet issuing into
a tank illustrates the
importance of velocity
gradients in eddy diffusion Large velocity
coefficient (From Gulliver gradient
[1])
Velocity profile
Fig. 5.5 Large and small

eddies in an open-channel
flow. The large eddies
perform most of the top-to-
bottom transport u
(From Gulliver [1])
The scale of the flow field is also important because the larger eddies perform
most of the transport. The small eddies are always there in a turbulent flow, and
their existence is important for local mixing. It is the large eddies, however, that are
the most responsible for transport, as illustrated in Fig. 5.5.
The four observations, listed above, were enough for Ludwig Prandtl [4] to
hypothesize a simple model for describing turbulent transport that works surpris-
ingly well, considering the complexity of turbulent flow.
Prandtl’s Mixing Length Hypothesis for Turbulent Flow
Prandtl’s mixing length hypothesis was developed for momentum transport, instead
of mass transport. The end result was a turbulent viscosity, instead of a turbulent
diffusivity. However, since both turbulent viscosity and turbulent diffusion coeffi-
cient are properties of the flow field, they are related. Turbulent viscosity describes
the transport of momentum by turbulence, and turbulent diffusivity describes the
transport of mass by the same turbulence. Thus,
98 J.S. Gulliver
ex ¼ mtx =r; ey ¼ mty =r; and ez ¼ mtz =r
where mtx, mty , and mtz are the turbulent viscosity in the x, y, and z directions. Now,
for the x-component of momentum (ru), the Boussinesq approximation is
@ u
r u0 u0 ¼ mtx (5.22)
@x
@ u
r v0 u0 ¼ mty (5.23)
@y
@ u
r w0 u0 ¼ mtz (5.24)
@z
Let us consider the fully developed velocity profile in the middle of a wide-open
channel, with x-, y-, and z-components in the longitudinal, lateral, and vertical
directions, respectively. It is fully developed because @ u=@x is close to zero. The
fact that it is a wide channel means that @ u=@y also is very small in the middle. From
Eqs. 5.22 and 5.23, we can see that the turbulent transport of momentum in the x-
and y-directions will be small because the gradients are small. Equation 5.24
indicates that there will be a net turbulent transport of momentum in the z-direction.
@ u
w0 u0 ¼ ez 6¼ 0 (5.25)
@z
Now, half of the w 0 values will be positive, and the other half will be negative.
We will use this criterion to divide them into two parts:
þ
w0 u0 ¼ w0 u0 þ w0 u0 (5.26)
where w 0 u 0 + has a value when w 0 is positive and is equal to zero when w 0 is

negative. w 0 u 0 has a value when w 0 is negative and is equal to zero when w 0 is
positive. Consider the cases when w 0 is positive. Then Eq. 5.26 becomes
þ
w0 u0 ¼ w0 u0 þ 0 (5.27)
Let us assume that an eddy of length L is pulling a blob of fluid upward, as

illustrated in Fig. 5.6. On average, the blob will have an x-component of velocity
equal to uðz L=2Þ, where z is the location where u 0 is to be estimated. Thus, the
eddy pulls up, on average, the u value that is at z – L/2. This will become the deviation
from the temporal mean velocity at location z:
1
u0 ¼ u u uðz
L=2Þ uðzÞ LÞ uðzÞ
ffi ðuðz Þ (5.28)
2
Fig. 5.6 Illustration of the

relationship between velocity
profile, turbulent eddies, and
mixing length
u w′ = ⊕ z
z- /2
Equation 5.28 is a relation for a difference in velocity, which can be written as

a velocity gradient times a distance:

@ u L
u0 ¼ (5.29)
@x 2
Velocity ¼ velocity distance

difference gradient
Then,
þ w0 w0 @ u
w0 u0 LÞ uðzÞ L
½uðz (5.30)
2 2 @z

the development is similar for w0 u0 :
þ w0 @ u
w 0 u0 w 0 u0 L (5.31)
2 @z
Now combining Eqs. 5.26, 5.30, and 5.31 gives
@ u
w0 u0 ¼ w0 L (5.32)
@z
Because turbulent eddies tend to be close to spherical in shape:
jw0 j ju0 j (5.33)
and from Eq. 5.29:
@ u0
w0 L (5.34)
@z
100 J.S. Gulliver
Table 5.2 Dynamic Surface Type z0 (m)

roughness lengths, z0, for
typical atmospheric surfaces Urban 1.-3
(Turner [7]) Forest 1.3
Deciduous forest in winter 0.5
Desert shrubland 0.3
Wetland 0.3
Cropland (summer) 0.2
Cropland (winter) 0.01
Grassland (summer) 0.1
Grassland (winter) 0.001
Water 0.0001
If we substitute Eq. 5.34 into Eq. 5.32, and then substitute the result into
Eq. 5.24, we get
2
@ u
2 @u
w0 u0 ¼ ez ¼L (5.35)
@z @z
or

2 @ u
ez ¼ L (5.36)
@z
Equation 5.36 is Prandtl’s mixing length hypothesis, and it works well, consid-
ering that the basis for the equation is so empirical. However, Eq. 5.36 does present
a challenge for us that mixing length, L, still needs to be specified. Measurements
have shown us the following:
1. Near a wall, L = kz, where k is von Kármán’s constant [5] and is very close to
0.4, and z is the distance from the closest wall.
Prandtl also made another assumption in this region, that w 0 u 0 could be
approximated by a constant equal to the mean wall shear stress, or
w0 u0 ¼ t=r ¼ u2 (5.37)
Then, eliminating w0 u0 from Eqs. 5.35 and 5.37 results in the well-known
logarithmic velocity profile:

u 1 z
¼ ‘n (5.38)
u k z0
where u * is the shear velocity at the wall, t is the wall shear stress, and z0 is an
integration constant, often called the dynamic roughness. Table 5.2 provides some
typical dynamic roughness lengths for atmospheric boundary layers. Applying

Eq. 5.36 to 5.37 results in an equation for ez in this region:
ez ¼ ku z (5.39)
2. Very near a wall (approaching the laminar sublayer where the turbulence is so
small that it is eliminated by the viscosity of the fluid), that is, for zu*/n < 35,
L y2 [6].
Making the same assumption that u 0 w 0 is approximately equal to wall shear
stress, this relation for L results in the following relation for velocity profile very
near the wall:
u n
¼b (5.40)
u u z
Equation 5.40 is not used in mass transport calculations near a wall or interface
because the unsteady character of mass transport in this region is very important,
and Eq. 5.39 is for a temporal mean velocity profile.
3. Away from a wall, where the closest wall does not influence the velocity profile,
L is a function of another variable of the flow field (Prandtl [8]). For example,
consider the jet mixer given in Fig. 5.4. In this case, the mixing length, L, is
a function of the width of the jet or plume. As the jet/plume grows larger, the
value of L is larger.
Here, it is easier to simply give the experimental relation for eddy diffusivity:
ez ¼ b umax b (5.41)
where b is the width of the mixing zone, b is a constant, and umax is the maximum
velocity in the jet at the given location, x.
Figure 5.7 gives some relationships for eddy diffusion coefficient profiles
under different conditions that will be handy in applications of turbulent diffu-
sive transport.
Example Applications
Example 5.1: Profile of eddy diffusion coefficient Estimate the eddy diffusivity
profile for a wind velocity of 18 m/s measured at 10 m over a large lake (Fig. 5.8),
and calculate the elevation above the water surface where ez = D for water vapor.
There is only one assumption needed:
1. The wind fetch is sufficient so that U 10 is influenced by shear at the water surface
(10 m is inside the boundary layer of the lake surface at this point).
102 J.S. Gulliver
εz
w εz = κu* z
z
τo
a Unconfined boundary layer
h u εz
εz = κu* z(1 – z/h)
b Flows with a free surface
Q b(x)
(m3/s)
εy = βum b = constant
c Jetlike flows (not close to a boundary)
εy = (0.6 ± 0.3)u* h
εy
depth = h
d Depth averaged transverse turbulent diffusion in a river
Fig. 5.7 Profiles of eddy diffusion coefficient for various types of applications (From Gulliver [1])
Fig. 5.8 Velocity profile U10 = 18 m′sec

over a large lake
10 m
Then, mixing length theory may be used with momentum transport to derive:
@ u u
¼ (5.42)
@z kz
and,
e z ¼ k u z (5.43)
Now, Wu [9] has provided the following equation from a fit of field data:
1=2
u ¼ 0:01 U10 8 þ 0:65 U 10 (5.44)
which indicates that as the waves get larger at high wind speeds the boundary
1=2
roughness effect upon u∗ increases by the factor 8 þ 0:65 U10 , where U10 is
given in m/s.
Then,
ez ðm2 /sÞ ¼ 0:01 k u10 zð8 þ 0:65 u10 Þ1=2 ¼ 0:32 z
when z is given in meters. Now, the diffusion coefficient of water vapor in air is
calculated to be D = 2.6 105 m2/s. Then, the elevation at which the diffusivity of
water vapor would equal eddy diffusivity in this case would be,
0:32z ¼ 2:6 105
or,
z ¼ 8 105 m ¼ 0:08 mm ¼ 80 mm
104 J.S. Gulliver
Fig. 5.9 Lateral and

longitudinal cross sections
of a typical river
h u ε
z C
Co
Thus at z = 80 mm elevation above the water surface, eddy diffusivity will be equal
to the diffusivity of water. A similarly small elevation would result for almost any
environmentally relevant compound. We can thus see that both e and D need to be
considered simultaneously in Eq. 5.20 only very close to surfaces in turbulent flow,
where e approaches the diffusion coefficient. Otherwise, diffusivities can be ignored
in solving turbulent flow transport problems, since e +D is essentially equal to e.
Example 5.2: Concentration profile of suspended sediment in a river (assuming ez
is constant) We will apply Eq. 5.20 to solve for the concentration profile of
suspended sediment in a river, with some simplifying assumptions. Suspended
sediment is generally considered similar to a solute, in that it is a scalar quantity
in Eq. 5.20, except that it has a settling velocity. We will also change our notation,
in that the bars over the temporal mean values will be dropped. This is a common
protocol in turbulent transport, and will be followed here for conformity. Thus, if an
eddy diffusion coefficient,e, is in the transport equation,
u means u
v means v
w means and
w
C means C
throughout the remainder of this entry. Fig. 5.9 gives a longitudinal and lateral cross
section of our river. We will make the following assumptions:
1. The flow is steady over the long term, so that ∂C/∂t = 0.
2. The flow is fully developed, such that any gradient with respect to x is equal to
zero (∂C/∂x = 0).
3. The river can be divided into a series of longitudinal planes with no significant
interaction, such that v = 0 and ey = 0 (this is the assumption of the stream-tube
computational models).
4. The vertical eddy diffusivity, ez, is a constant value.
Assumptions 3 and 4 are more difficult to justify.

The solute will have a vertical velocity, w = vs, where vs is the settling velocity
of the suspended sediment.
Then, Eq. 5.20 becomes

@C @ @C
ns ¼ ðD þ ez Þ (5.45)
@z @z @z
where we have not yet applied assumption 4. We can move the settling velocity into
the partial term:

@ðns CÞ @ @C
¼ ðD þ ez Þ (5.46)
@z @z @z
and since both sides of Eq. 5.46 are a gradient with respect to z, the terms inside of
the gradients must also be equal:
dC
ns C ¼ ðD þ ez Þ (5.47)
dz
Equation 5.47 is converted to an ordinary differential equation because all

variables are only a function of z. Now, we will deal with assumption 4. Fig. 5.7
gives the equation developed by Rouse [10] for ez:
ez ¼ k u zð1 z=hÞ (5.48)
where u∗ is the shear velocity at the bottom of the channel, or

pffiffiffiffiffiffiffiffi
u ¼ t=r (5.49)
where t is the shear stress at the wall. For a fully developed open-channel flow in
a wide channel, the following relation is easily derived:
u ¼ ghS (5.50)
This derivation can be found in a text on fluid mechanics or open-channel flow.

Assumption 4 states that ez ¼ ez for all values of z, where ez is the depth average, or
Zh Zh
1 ku
ez ¼ ez dz ¼ zð1 z=hÞdz ¼ 0:067 u h (5.51)
h h
0 o
106 J.S. Gulliver
where h is the depth of the stream. The termez is almost always much greater than D
in a turbulent flow. Thus,
D þ ez ffi ez
Now, substituting these equations into Eq. 5.47 results in
dC
ez þ ns C ¼ 0
dz
We will solve this by separating variables,
dC ns
¼ dz
C ez
integrating, and taking both sides of the solution to the power of e:

vs
C ¼ b1 e ez z
Now, we need a boundary condition to determine b1. This is difficult with

suspended sediment profiles. We can develop a fairly good estimate of the distri-
bution of suspended sediment once we have a known concentration at some location
in the flow field. In the sediment transport field bed load and suspended load are often
discussed. The relation between the two, and some experience and measurements of
both simultaneously, can be used to predict an equivalent suspended sediment
concentration at the bed. Then, the relevant boundary condition is
1. At
z ¼ 0; C ¼ C0 :
where C0 is the concentration that has been determined from the bed
load–suspended load relationship. Applying this boundary condition gives b1 =
C0, and our solution is
C ¼ C0 ens z=ez
The result is illustrated in Fig. 5.10. This problem can also be solved without
assumption 4 [10].
Example 5.3: Concentration of organic compounds released into the air by an
industrial plant (application of the product rule to error function solutions) There
is some concern about the emissions from the adhesives produced in an industrial
plant. Specifically, the town of Scream Hollow is 1 km away from the plant, where
citizens have begun to complain about odors from the plant and of headaches.
Fig. 5.10 Suspended 5

sediment concentration
profile for Example 2
4
vs z / εz
2
0
0 0.2 0.4 0.6 0.8 1
C / Co
Town
wind
plant 10 m Acrolein
Fig. 5.11 Illustration of toxic chemical release into the atmosphere, with the wind blowing toward
a town
One culprit, aside from a haunting, may be the release of Acrolein, C3H4O,
a priority pollutant that is an intermediary of many organic reactions. The average
release from the 200 m 200 m 10 m plant sketched in Fig. 5.11 is assumed to be
20 g/h. If the wind is blowing directly toward Scream Hollow, at 3 m/s measured at
3 m height, with a dynamic roughness of 0.2 m for the farmland, what
concentrations will the Scream Hollow inhabitants experience? Is this above the
EPA threshold limit of 0.1 ppm(v)?
We will need to make some assumptions to formulate this problem. They are:
1. The Acrolein release is distributed over the most downwind plane of the
building. With the important concentration being 1 km away, this is not a bad
assumption. Then, the Acrolein will be released over the plane that is 200 m
10 m. If 20 g/h = 0.0056 g/s are released into a wind moving at 3 m/s, the initial
concentration is
0:0056 g/s
C0 ¼ ¼ 9:3 107 g/m3
3 m/sð200 mÞð10 mÞ
2. We will use a cross-sectional mean velocity of U ¼ u at 3 m height, or U = 3 m/s.

3. We will use ez ¼ ey ¼ ez at 3 m height.
4. We will not consider any of the source or sink terms for Acrolein.
108 J.S. Gulliver
Fig. 5.12 Illustration of the z

coordinate system for y
Example 5
w
x
We will also set up the coordinates so that (x,y,z) = (0,0,0) occurs on the ground
at mid-plant width, and will orient the wind in the x-direction.
With these assumptions, the governing equation becomes
@C @2C @2C
U ¼ ey 2 þ ez 2 (5.52)
@x @y @z
The boundary conditions are

1. At (x,y,z) = (0, 100 m ) 100 m, 0 ) 10 m), C = C0.
2. As x ) 1, y ) 1, or z ) 1, C ) 0.
3. Zero mass flux at z = 0.
These boundary conditions, illustrated in Fig. 5.12, will give us a concentration
front, but in two dimensions. In addition, we have a zero flux condition that will
require an image solution. We will use the solution of Example 5 in the entry
“▾Transport in the Environment” to develop a solution for this problem. The
solution, before applying boundary conditions, was

z
C ¼ bo þ b1 erf pffiffiffiffiffiffiffiffi
4Dt
Now, we need an image to the concentration front about the z = 0 plane. In the
y-direction we have a step-up at y = Dy and a step down at y = Dy. We will also
use the product rule (Example 3, Transport in the Environment) to indicate that the
solution to our governing equation for the y-direction should be multiplied times the
solution in the z-direction. Then the solution can be given as
" ! !#
C ðz þ DzÞ ðz DzÞ
¼ bo þ b1 erf pffiffiffiffiffiffiffiffiffiffiffiffiffiffi þ b2 erf pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Co 4ez x=U 4ez x=U
" ! !#
ðy þ DyÞ ðy DyÞ
b3 erf pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi þ b4 erf pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
4ey x=U 4ey x=U
where Dy = 100 m and Dz = 10 m.

Now, to see if our boundary conditions can be satisfied with the form of the
solution:
@ x ) 1; C ) 0: Thus b0 ¼ 0:
@ z ) 1; C ) 0: Thus b1 ¼ b2 :
@ y ) 1; C ) 0: Thus b3 ¼ b4 :
@ x ) 0; and ðy; zÞ ¼ ð0; 0Þ; C=C0 ¼ 1:
With the last boundary condition, Eq. E5.5.3 becomes
1 ¼ ðb1 b2 Þðb3 b4 Þ
or,
1 ¼ 2b1 2b3
or,
1 ¼ 2b2 2b4
Finally, @ x ) 0 and (y,z) = (0,Dz), C/C0 = ½. Thus b1 = ½.

Applying this last boundary condition results in b3 = ½, b2 = ½, and b4 = ½.
Thus, the solution to Eq. 5.52 is
( ! !)
C 1 ðz þ DzÞ ðz DzÞ
¼ erf pffiffiffiffiffiffiffiffiffiffiffiffiffiffi erf pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Co 2 4ez x=U 4ez x=U
( ! !) (5.53)
ðy þ DyÞ ðy DyÞ
erf pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi erf pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
4ey x=U 4ey x=U
Now, if we use Dz = 10 m, Dy = 100 m, U = 3 m/s, the only remaining parameter

to find is e. Using Eq. 5.43 given in Example 1:
ez ¼ ey ¼ ku z (5.43)
Note that the logarithmic boundary equation can be written as

u z
¼ ‘n (5.54)
u zo
where z0 is the dynamic roughness, assumed to be 0.2 m for the crop land between
the plant and Scream Hollow. Then,
110 J.S. Gulliver
u 3 m/s
u ¼ ¼ 3 m ¼ 1:1 m/s
‘nðz=zo Þ ‘n 0:2 m
and
ez ¼ 0:4ð1:1m=sÞð3mÞ ¼ 1:3 m2=s
If we now plug all of the parameters for the industrial plant into Eq. 5.53, we get C
= 0.25 mg/m3 = 2.5 107 g/m3. In terms of ppm(v), we will use rair = 1.2 g/m3, and
the molecular weights of air and Acrolein of 29 and 56 g/mole, respectively. Then,
2:5 107 g/m3 MWair

C¼
rair MWC3 H4 0
2:5 107 g/m3 20 g/mole
¼
1:2 g/m3 56 g/mole
moles C3 H4 0
¼ 1:08 107
mole air
This is right at the threshold for continuous exposure, and the pollution from the
plant should be investigated in more detail.
Conclusions
1. Although turbulent diffusion is a convection transport, and not a diffusive

transport, the result looks similar to diffusion, and can be described by
a turbulent diffusion coefficient.
2. Most environmental flows are turbulent. The exceptions are flow through porous
media and flows that are very close to an interface.
3. Reynolds averaging and the Boussinesq assumption result in a turbulent transport
equation that contains many features of the diffusive mass transport equation,
and can be solved by similar techniques.
4. Prandtl’s mixing length is a relatively accurate simplification for many turbulent
flows.
Future Directions
The future for turbulent transport in the environment is in the direction of computa-
tional mass transport. This requires a simultaneous fluid dynamics–mass transport
solution. On typical environmental scales, the computational power of our computers
still must be advanced to solve these large problems while resolving the scale of the
smallest turbulent eddies. Direct numerical simulation cannot deal with the scale of
these problems, and large eddy simulation cannot keep both the scale and grid
refinement required.
Bibliography
Primary Literature
1. Gulliver JS (2007) Introduction to chemical transport in the environment. Cambridge University

Press, Cambridge, UK, 288 pp
2. Reynolds O (1895) On the dynamical theory of incompressible viscous fluids and the determi-
nation of the criterion. Philos Trans R Soc 186:123–164
3. Boussinesq J (1877) Essai sur la théorie des eaux courantes. Mem Pres Acad Sci Paris 23:46
4. Prandtl L (1925) Bericht iiber Untersuchungen zur ausgebildeten Turbulenz. Z Angew Math
Mech 5:136–139
5. von Kármán T (1930) Mechanische Ahnlichkeit und Turbulenz. Nachr Ges Wiss Gottingen
5:58–76. Also Proceedings of third international congress on applied mechanics, vol I,
Stockholm, pp 85–93, 1930
6. Reichardt H (1951) Vollstandige Darstellung der turbulenten. Geschwindigkeitsverteilung.
Ann Angew Math Mech 31:7
7. Turner DB (1994) Workbook of atmospheric dispersion estimates, 2nd edn. Lewis, Boca
Raton
8. Prandtl L (1942) Bemerkungen zur Theorie der freien Turbulenz. Z Angew Math Mech
22:241–243
9. Wu J (1980) Wind-stress coefficients over sea surface. J Geophys Res 74:444
10. Rouse H (1937) Modern conceptions of the mechanics of fluid turbulence. Trans Am Soc Civil
Eng 102(1965):463–543
Books and Reviews
Nezu I, Nakagawa H (1993) Turbulence in open channel flow. Balkema, Rotterdam

White FM (1974) Viscous fluid flow. McGraw-Hill, New York
Chapter 6
Dispersive Transport
John S. Gulliver
Glossary
Convection The movement of a constituent with movement

of the fluid.
Diffusion The spreading of fluid constituents through the
motion inherent to atoms and molecules.
Diffusion coefficient A coefficient that describes the tendency of
molecules to spread a constituent mass.
Dirac delta An impulse of a given quantity (mass) that occurs
over an infinitely short time or space.
Dispersion The process of mixing caused by a variation in
velocity and transverse diffusion or turbulent
diffusion.
Dispersion coefficient A coefficient that can describe the mixing caused
by a transverse velocity profile and transverse
diffusion or turbulent diffusion. A dispersion
coefficient means that some sort of spatial mean
velocity is being used to describe the flow. Then,
the mixing lateral or longitudinal to the spatial
mean velocity due to a combination of a velocity
profile and diffusion or turbulent diffusion is
described by the dispersion coefficient. The
coefficient’s location in the mass transport
J.S. Gulliver (*)
Department of Civil Engineering, University of Minnesota, 500 Pillsbury Drive S.E.,
Minneapolis, MN 55454, USA
114 J.S. Gulliver
equation is similar to diffusion coefficients, and

the units are similar.
Laminar flow Flow that has no turbulent eddies, where the fluid
flows in laminas and diffusion creates the mixing
of the fluid.
Turbulent diffusion The mixing of fluids through turbulent eddies
created by convection.
Turbulent diffusion coefficient A coefficient that comes from the multiplication
of two turbulent velocities of the flow, divided by
density of the fluid. The coefficient’s location in
the mass transport equation is similar to diffusion
coefficients, and the units are similar, so it is
called a “turbulent diffusion coefficient.”
Definition of Dispersive Transport in the Environment
Dispersion is the enhanced mixing of material through spatial variations in velocity.

When it is of interest (when we are not keeping track of the three-dimensional
mixing), dispersion is typically one or two orders of magnitude greater than
turbulent diffusion. The process of dispersion is associated with a spatial mean
velocity, the assumption of plug flow, and a velocity profile. The means used in
association with diffusion, turbulent diffusion, and dispersion are identified in
Table 6.1.
The means by which diffusion and possibly turbulent diffusion are combined
with a spatial mean velocity to result in dispersion is illustrated in Fig. 6.1.
A velocity profile over space with mixing due to diffusion (and possibly turbulent
Table 6.1 Temporal or spatial means and scales used in association with various mixing processes
Process Variable representing process Mean Scale of mean
Diffusion Diffusion coefficient Temporal Molecular
Turbulent diffusion Turbulent diffusion coefficient Temporal Minutes
Dispersion Dispersion coefficient Spatial Scale of flow
Diffusion or Turbulent Dispersion

Diffusion
⇒
+ +
↑
Uniform Velocity Cross-Sectional
Velocity Profile
Assumption Mean Velocity
Fig. 6.1 Representation of the process by which diffusion or turbulent diffusion is related
to dispersion
6 Chemicals in the Environment, Dispersive Transport 115
diffusion) is combined into a cross-sectional mean velocity and a dispersion coeffi-

cient. Without the cross-sectional mean velocity, there is no dispersion coefficient.
Introduction
Dispersion was first developed as a means of dealing with reactors, where there was
little interest in the processes creating mixing inside of the reactor, but great need to
appropriately describe the output from the reactor. The physics of the mixing
process is lost in the conversion to a spatial mean velocity profile, but the end
result can still be modeled by dispersion.
A similar spatial mean velocity (bulk mean velocity) is used for the plug flow
reactor model. Thus, plug flow with dispersion is a natural match, where the mixing
that truly occurs in any reactor or environmental flow is modeled as dispersion. This
is the model that will be applied to utilize dispersion as a mixing model.
Dispersion in Laminar Flow
Any flow with a nonuniform velocity profile will, when spatial mean velocity and
concentration are taken, result in dispersion of the chemical. For laminar flow, the
well-described velocity profile means that we can describe dispersion analytically
for some flows. Beginning with the diffusion equation in cylindrical coordinates
(laminar flow typically occurs in small tubes):
2
@C @C @ C 1 @C @ 2 C
þU ¼D þ þ þS (6.1)
@t @x @r 2 r @r @x2
where the x-coordinate is aligned with the flow velocity, and v = w = 0. We will
outline the development of a dispersion coefficient for a fully developed laminar
pipe flow. This means that we are far enough from the entrance that the velocity
profile is essentially in equilibrium with the loss of pressure along the pipe. This
flow has a velocity profile (developed in most fluid mechanics texts):
U ¼ Umax ð1 r 2 =R2 Þ (6.2)
as illustrated in Fig. 6.2.

To convert Eq. 6.1 to cross-sectional mean values, we will assign:
ðR
1
C^ ¼ 2 C2pr dr (6.3)
pR
0
116 J.S. Gulliver
1
0.8
0.6
0.4
0.2
r/R
0
0 0.2 0.4 0.6 0.8 1
−0.2
U/ Umax
−0.4
−0.6
−0.8
−1
Fig. 6.2 The velocity profile in a fully developed tubular flow is a paraboloid
ðR
1 Umax
U¼ 2 U 2pr dr ¼ (6.4)
pR 2
0
Then, in Eq. 6.1:
C ¼ C^ þ C0 ðrÞ (6.5)
and
u ¼ U þ u0 ðrÞ (6.6)
Now, if we equate Eq. 6.2 and 6.6, we get

1 r2
u0 ðrÞ ¼ um (6.7)
2 R2
Substituting Eq. 6.5, 6.6, and 6.7 into Eq. 6.1:
@ C^ @C0 @ðC^ þ C0 Þ
þ þ ðU þ u0 Þ
@t " @t @x # (6.8)
@ 2 ðC^ þ C0 Þ 1 @ðC^ þ C0 Þ @ 2 ðC^ þ C0 Þ
¼D þ þ þS
@r 2 r @r @x2
An order of magnitude analysis will tell us that ∂2/∂r2> > ∂2/∂x2. In addition,
∂Ĉ/∂r = 0, by definition. Now, if we put everything that is known on the left-hand
side of Eq. 6.8, the result will be:
" #
@ C^ @ C^ @C0 @C0 ^
0 @C 0 @C
0
þU ¼ þU þU þU
@t @x @t @x @x @x
(6.9)
@ 2 C0 D @C0 @ 2 C^
þD 2 þ þ S ¼ DL 2 þ S
@r r @r @x
The second equality in Eq. 6.9 is a definition of longitudinal dispersion coeffi-

cient, DL. G. I. Taylor [1] assumed that some of the terms in Eq. 6.9 would cancel
and that longitudinal convective transport would achieve a balance with transverse
diffusive transport. He then solved the second equality in Eq. 6.9 for a fully
developed tubular flow, resulting in the relation,
R2 U 2 d 2 U 2
DL ¼ ¼ (6.10)
48 D 192 D
A similar relation can be developed for laminar flow down an inclined plate:
32 h2 U 2
DL ¼ (6.11)
945 D
The longitudinal dispersion coefficient is proportional to the square of the flow

scale (d or h), proportional to the square of the velocity scale (U), and inversely
proportional to the diffusion coefficient. The greater the diffusion, the less severe
the spread of the chemical by the velocity profile because of local mixing, and the
smaller the longitudinal dispersion coefficient. This result may seem illogical, but
can be explained by the following: Longitudinal dispersion describes mixing only
in terms of the cross-sectional mean concentration, and transverse mixing actually
slows down longitudinal dispersion. The governing Eq. 6.9 does not concern itself
with local mixing issues.
Knowing the relations given in Eq. 6.10 and 6.11, we no longer need the more
cumbersome middle portion of Eq. 6.8, and we can work to solve the equation,
@ C^ @ C^ @ 2 C^
þU ¼ DL 2 þ S (6.12)
@t @x @x
In expressing the equations for longitudinal dispersion, we will drop the “hat”
above the cross-sectional mean values. Thus, if DL is involved, we are discussing
cross-sectional mean concentrations.
Dispersion in Turbulent Flow
The dispersion that occurs in turbulent flow can also be calculated, as long as the
velocity profile is given. This was done by Taylor [2] for a tubular flow and by Elder
118 J.S. Gulliver
[3] for a two-dimensional, open-channel flow. Both investigators assumed that

a logarithmic velocity profile would apply in the entire flow field. The logarithmic
velocity profile is a fairly good description where shear stress can be assumed
constant. It is a good assumption for a fully developed turbulent flow field, because
the locations where the logarithmic profile applies are those with the greatest
change in velocity.
For a fully developed tubular flow, assuming a logarithmic velocity profile,
Taylor derived the equation,
DL ¼ 5:05 d u (6.13)
where u∗ is the shear velocity at the wall. Elder [3] derived the following equation
for a two-dimensional, open-channel flow:
DL ¼ 5:93 h u (6.14)
It is interesting to compare Eq. 6.13 and 6.14 with those for a fully developed
laminar flow, Eq. 6.10 and 6.11. In the chapter on Turbulent Transport, eddy diffusion
coefficient in a turbulent boundary layer was found to be linearly dependent upon
distance from the wall and on the wall shear velocity. If we replace the diffusion
coefficient in Eq. 6.11 with an eddy diffusivity that is proportional to h u∗, we get,
h2 U 2
DL (6.15)
hu
Noting that for a given boundary roughness we can generally say that U u∗,
Eq. 6.15 becomes,
DL hu (6.16)
which is what we have in Eq. 6.14.

The relations developed for longitudinal dispersion coefficient are given in
Table 6.2. The experimental results in rivers tend to have a large range because
of the variety of lateral velocity profiles that exist in natural rivers and streams.
Solutions to Transport with Convection
Diffusive transport with convection occurring simultaneously can be solved more

easily if we orient our coordinate system properly. First, we must orient one axis in
the direction of the flow. In this case, we will choose the x-coordinate so that u is
nonzero and v and w are zero. Second, we must assume a uniform velocity profile, u
= U = constant with y and z. Then Eq. 6.30 of the Transport in the Environment
chapter becomes
Table 6.2 Relationships for longitudinal dispersion coefficient in pipes and channels developed
from theory and experiments
Flow conditions DL Notation
Laminar flow in a pipe [1] R = radius of tube
R2 U 2 U = cross-sectional mean velocity
48 D D = diffusivity
Laminar flow down an inclined plate h = depth
32h2 U 2 Vp = velocity of upper plate
945 D Dz = spacing of plates
Laminar flow-linear velocity profile u∗ = shear velocity
(Couette Flow) Vp2 D2z
120 D
Turbulent flow in a pipe, assuming 10.1Ru∗
logarithmic velocity profile [2]
turbulent flow down an inclined plate, 5.93hu∗
assuming logarithmic velocity profile [3]
Open flume (experimental) 8–400 hu∗
Canals (experimental) 8–20 hu∗
Rivers (experimental) 8–7,500 hu∗
C
t = x R/U
x* = x − Ut/R = 0
0 t
Fig. 6.3 Pulse response in fixed and moving coordinate systems

@ C U @ C D @2 C @2 C @2C S
þ ¼ þ þ þ (6.17)
@ t R @ x R @ x2 @ y2 @ z2 R
where S is a source or sink term other than adsorption or desorption. In Eq. 6.17, we
are assuming that chemical reaction does not take place on the surface of a solid,
i.e., while the chemical is sorbed to solids. Now, we will convert our Eulerian
(fixed) coordinate system to one that moves (Lagrangian) with velocity U/R and
assign an independent variable
x ¼ x Ut=R (6.18)
such that x∗ = 0 at x = U t/R. The response of the system to a conservative pulse,

given in Fig. 6.3, indicates that in the Lagrangian coordinate system specified, there
is no convection term, only diffusion. Then Eq. 6.17 becomes
120 J.S. Gulliver

@C D @ 2 C @ 2 C @ 2 C S
¼ 2
þ 2þ 2 þ (6.19)
@t R @x @y @z R
Determination of Dispersion Coefficient from Tracer Clouds
The one-dimensional mass transport equation for plug flow with dispersion, and
a retardation coefficient of 1, is
@C @C @2C
þU ¼ DL 2 þ S (6.20)
@t @x @x
We will convert our fixed coordinate system to a coordinate system moving at

velocity U through the change of variables, x∗ = x Ut. Then Eq. 6.20 is given as
@C @2C
¼ DL þS (6.21)
@t @x2
Now, if we are determining the dispersion coefficient through the use of a pulse
tracer cloud, the boundary conditions are those of a Dirac delta:
1. At t = 0, a pulse of mass M is released at x∗ = 0.
2. As t ) 1, C ) 0.
The solution to Eq. 6.21 with these boundary conditions [4] is:
ðLUtÞ2
M=A
C ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi e
4D t
L (6.22)
4p DL t
or, in dimensionless variables,

" #
2
C C 1 ð1 CouÞ
¼ ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi exp (6.23)
C0 M=ðALÞ 4 pCou=Pe 4 Cou=Pe
Where Cou = U t/L = t/tr is a Courant number and Pe = U L/DL is a Peclet number.
Comparing Eq. 6.23 to a Gaussian probability distribution,

1 f2
PðfÞ ¼ pffiffiffiffiffiffi exp 2
s 2p 2s
we can see that the plug flow with dispersion, when Pe > 10, can be fit to a Gaussian
distribution in terms of C/C0 and (1y), with the relationships,
s2t
s2 ¼ ¼ 2 Cou=Pe (6.24)
t2r
where
Ð
1
t2 C=C0 dt
0
s2t ¼ 1
Ð tr 2 (6.25)
C=C0 dt
0
and
Ð
1
t2 C=C0 dt
0
tr ¼ 1
Ð (6.26)
C=C0 dt
0
In addition, if we could measure the tracer cloud over distance at one time, we
would use the relation,
s2x
s2 ¼ (6.27)
U 2 tr 2
where
Ð
1 Ð1
x2 C=C0 dx x C=C0 dx
0 0
s2x ¼ 1
Ð x and x ¼ 1
2
Ð (6.28)
C=C0 dx C=C0 dx
0 0
The response of a plug flow with dispersion model to a pulse input, Eq. 6.23, is
given in Fig. 6.4 for various values of the Peclet number, Pe = U L/DL.
Dispersion in Groundwater Flow
Dispersion in a flow through porous media occurs due to heterogeneity in the

media, i.e., the conductivity of the soil varies with space. This is shown on three
levels in Fig. 6.5. On the particle scale, a thread of tracer will be split a number of
times as it moves through the media. Each split of the tracer thread will move
through the media at a speed corresponding to the resistance that it encounters. If
122 J.S. Gulliver
3
Pe = 0.1
2.5 Pe = 1
Pe = 10
Pe = 100
2
C / C0
1.5
0.5
0
0 1 2 3 4
t / tr
Fig. 6.4 Response of the plug flow with dispersion model to a pulse input
you take a number of tracer threads coming out of the media at different times, and
collected them in an outlet pipe, what you would see at the end of the pipe would be
a dispersed pulse. This dispersion would be much greater than the diffusion that
would occur. A lateral dispersion would also occur because the media would move
some of the tracer threads laterally. Thus, in groundwater flow, there generally is
longitudinal and lateral dispersion, created by the character of the media.
On a larger scale, a similar process can occur. Fingering of the tracer is created
by layered beds with a low conductivity and lenses with a high conductivity.
The tracer that ends up in a lens travels at a relatively high speed. Those tracer
molecules will reach the measuring point sooner than the tracer molecules stuck in
the low conductivity beds, creating longitudinal dispersion. As the lenses are not all
parallel to each other, they will also create a lateral dispersion of the tracer.
Finally, on a still larger scale, about the scale of a small town, there will be
isopleths (lines of constant concentration) of our tracer that look something like
those given in the last illustration of Fig. 6.4. If enough particle and lens effects
have occurred with an apparent randomness to our tracer cloud, then the cloud
disperses in a manner similar to that illustrated.
Incorporating the retardation coefficient for the chemical, discussed in the
chapter on Transport in the Environment, the mass transport equation is then
written as:
@C @C ðD þ Dx Þ @ 2 C ðD þ Dy Þ @ 2 C
þ U=R ¼ þ
@t @x R @x2 R @y2
(6.29)
ðD þ Dz Þ @ 2 C
þ
R @z2
IN OUT
Moves
through
slowly
Moves
through
quickly
t t
Particle scale
Two flow paths on the particle scale move through the media at different rates.
On a larger level, fingering is caused by layered beds with a low k

(conductivity) and lenses with a high k
Lenses cause
both
longitudinal
and lateral
dispersion
Tracer
Lense scale
On a still-larger spatial and temporal scale, a “dispersion” cloud

y New
Brighten
city limits
V x
x Town scale
Fig. 6.5 Illustration of dispersion in groundwater flow at various scales (From [5])
where R is a retardation coefficient, and Dx, Dy, and Dz are the dispersion
coefficients in the x-, y-, and z-directions, respectively.
The dispersion coefficients are dependent upon the character of the media and
the flow velocity. It is difficult to predict these coefficients within an order of
124 J.S. Gulliver
magnitude, so they are normally measured or fit to measured data in the field. For
example, if a concentration variance in the x, y, and z directions can be measured in
response to a pulse release, then an approximate solution would be:

MR1:5 Rx2 Ry2 Rz2
C¼ exp (6.30)
8ðp3 t3 Dx Dy Dz Þ0:5 4 Dx t 4 Dy t 4 Dz t
where x∗ is x Ut/R. Comparing Eq. 6.30 to that for a three variable Gaussian
distribution, we can see that:
Dx ¼ 9 Rsx 2 = ð2tÞ (6.31a)
Dy ¼ 9R sy 2 =ð2tÞ (6.31b)
Dz ¼ 9 Rsz 2 =ð2tÞ (6.31c)
where s2x ; s2y and s2z are the variance of concentration in the x, y, and z directions,
respectively. They are given by the equations,
Ð Ð 2
x2 C dx x C dx
s2x ¼ Ð Ð (6.32a)
C dx C dx
Ð
y2 C dx
s2y ¼ Ð (6.32b)
C dx
Ð 2
z C dx
s2x ¼ Ð (6.32c)
C dx
The last term in Eq. 6.32a is the distance to the center of mass. In the diffusion
equation, it is equal to Ut/R.
Dispersion coefficients in groundwater flow. In a uniform media of particles, the
longitudinal dispersion coefficient, DL, and the transverse dispersion coefficient, Dt,
are both functions of the grain diameter and velocity. (In our previous example, DL
was Dx, and Dt would indicate Dy and Dz). The relevance of longitudinal and
transverse dispersion relative to diffusion may therefore be very roughly
characterized by a Peclet number, Pe:
Pe ¼ Ud=D (6.33)
where d is the grain diameter and U is a bulk velocity, Q/A, where the cross section
includes porous media. If the consideration is not of uniform media but is for
a heterogeneous region of high and low groundwater flow permeability, then the
appropriate length scale would be the size of these permeability regions, normal to
the flow. The characterization is [6];
DL =D Pe (6.34)
Koch and Brady [7] have characterized transverse dispersion coefficient as

a fraction of longitudinal dispersion coefficient:
Dt =DL 0:1 (6.35)
Thus, longitudinal dispersion coefficient is roughly 10 times the value of trans-

verse dispersion coefficient in a uniform media.
In the field, however, all media are heterogeneous, resulting in far greater
dispersion than in uniform porous media. Because of the heterogeneities, the
velocity profile can be highly variable over long distances, creating a much greater
dispersion. Gelhar et al. [8] provided a plot of field data that can be manipulated to
result in an equation that applies between a scale of 1 and 100,000 m:
UL
PeD ¼ ¼ 2:5 X 1011:4 (6.36)
L DL
where L is the horizontal scale of the measurement. The field data are given in
Fig. 6.6, with the curves of Eq. 6.36.
Dispersion in Rivers
Rivers are an excellent environmental flow for describing the flow as a mean
velocity with dispersion. The flow is confined in the transverse and vertical
directions, such that a cross-sectional mean velocity and concentration can be easily
defined. In addition, there is less variation in rivers than there is, for example, in
estuaries or reactors, both of which are also described by the plug flow with
dispersion model. For that reason, the numerous tracer tests that have been made
in rivers are useful to characterize longitudinal dispersion coefficient for use in
untested river reaches. A sampling of the dispersion coefficients at various river
reaches which were determined from tracer tests is given in Table 6.2. Also given
are the relevant mean parameters for each reach.
The question that we need to ask ourselves is whether the longitudinal dispersion
can be predicted accurately for these rivers. Equation 6.16, which predicts that
DL/(u∗h) = constant, is shown in Table 6.2 to have a large range of constants,
probably because of the variations in cross section and morphology seen in natural
streams. Fischer [9] observed that this constant seemed to depend upon mean
surface width, W, and substituted W for h in the numerator of Eq. 6.16 to develop
126 J.S. Gulliver
104
103
102
Longitudinal Dispersivity (m)
101
fractured
100
porous
tracer
10−1 tests
contam.
events
10−2 envir.
tracers
10−3 −1
10 100 101 102 103 104 105 106
Scale (m)
Fig. 6.6 Field data on dispersion coefficients, taken from Gelhar et al. [8] with Eq. 6.36 added.
Longitudinal dispersivity is DL/U
the following empirical equation to characterize longitudinal dispersion coefficient

in rivers:
U2 W 2
DLP ¼ 0:011 (6.37)
u h
where DLp is the predicted value of DL, h the mean depth, U the cross-sectional
mean velocity, and u∗ the mean shear velocity of the river reach. When compared
to the data given in Table 6.3 the root mean square error relative to the measure-
ment, given by the equation,
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
1X DLP 2
Relative ¼ 1 (6.38)
rms Errors n n DL
where n, the number of measurements, is 1.71. This relative rms error means that
roughly 67% of the predictions are within a factor of 1.71 (within 71%) of the
observed.
Table 6.3 Measurements of longitudinal dispersion coefficient in laboratory flumes and rivers (after Fisher et al., 1979)
DLR
Observed predicted
Mean Shear despersion by
Depth. Width. velocity, velocity, coefficient, Eq. (6.37)
2
References Channel h(m) W(m) U (m/sec) u* (m/sec) DL (m /sec) DL/(hu*) (m2/sec)
Thomas (1958) Chicago Ship Canal 8.07 48.8 0.27 0.0191 3.0 20 12.4
State of California (1962) Sacramento River 4.00 0.53 0.051 15 74
Owens et al. (1964) River Derwent 0.25 0.38 0.14 4.6 131
Glover (1964) South Platte River 0.46 0.66 0.069 16.2 510
Schuster (1965) Yuma Mesa A Canal 3.45 0.68 0.345 0.76 8.6
Fischer (1967) Trapezoidal laboratory 0.035 0.40 0.25 0.0202 0.123 174 0.156
channel with 0.047 0.43 0.45 0.0359 0.253 150 0.244
roughened sides 0.035 0.40 0.45 0.0351 0.415 338 0.298
0.035 0.34 0.44 0.0348 0.250 205 0.202
0.021 0.33 0.45 0.0328 0.400 392 0.248
0.021 0.19 0.46 0.0388 0.220 270 0.103
6 Chemicals in the Environment, Dispersive Transport
Fischer (1968b) Green-Duwamish River, 1.10 20 0.049 6.5–8.5 120–160

Washington
Yotsukura et al. (1970) Missouri River 2.70 200 1.55 0.074 1500 7500 3440
Godfrey and Frederick Copper Creek, Virginia 0.49 16 0.27 0.080 20 500 5.24
(1970) (below gauge) 0.85 0.60 0.100 21 250 15.1
0.49 16 0.26 0.080 9.5 245 4.86
Clinch River, Tennessee 0.85 47 0.32 0.067 14 235 22
2.10 60 0.94 0.104 54 245 73
2.10 53 0.83 0.107 47 210 28
Copper Creek, Virginia 0.40 19 0.16 0.116 9.9 220 2.19
(above gauge)
Powell River, Tennessee 0.85 34 0.15 0.055 9.5 200 6.12
Cinch River, Virginia 0.58 36 0.21 0.049 8.1 280 22.1
127
(continued)
Table 6.3 (continued)
128
DLR
Observed predicted
Mean Shear despersion by
Depth. Width. velocity, velocity, coefficient, Eq. (6.37)
2
References Channel h(m) W(m) U (m/sec) u* (m/sec) DL (m /sec) DL/(hu*) (m2/sec)
Coachella Canal, California 1.56 24 0.71 0.043 9.6 140 47.6
McQuivey and Keefer Bayou Anacoco 0.94 26 0.34 0.067 33 520 13
(1974) 0.91 37 0.40 0.067 39 690 38
Nooksack River 0.76 64 0.67 0.27 35 170 98
Wind Bighorn Rivers 1.10 59 0.88 0.12 42 330 232
2.16 69 1.55 0.17 160 440 340
John Day River 0.58 25 1.01 0.14 14 170 88
2.47 34 0.82 0.18 65 150 20
Comite River 0.43 16 0.37 0.05 14 650 16
Sabine River 2.04 104 0.58 0.05 315 3090 330
4.75 127 0.64 0.08 670 1760 190
Yadkin River 2.35 70 0.43 0.10 110 470 44
3.84 72 0.76 0.13 260 520 68
Gulliver (1977) MERS Experimental Streams 0.32 32 0.95 0.52 0.11 17 0.61
J.S. Gulliver
Tracer Determination of Longitudinal Dispersion Coefficient in Rivers
Tracers are generally used to determine longitudinal dispersion coefficient in rivers.

Some distance is required, however, before the lateral turbulent diffusion is bal-
anced by longitudinal convection, similar to Taylor’s [1] analysis of dispersion in
a laminar flow. This transport balancing distance, X∗ is given by the equation:
0:2 UW 2
X ¼ (6.39)
ey
The region x < x∗ can be visualized as a mixing region, which can skew the
results of a tracer study. Downstream of this region, where turbulent diffusion is
balanced by longitudinal convection, the variance of a tracer pulse grows linearly
with distance. It is best to begin the measurements a distance X∗ below the tracer
release. Once dispersion coefficient at one or more river discharges have been
measured, Eq. 6.37 can be used to adjust the dispersion coefficient to all discharges:
2 2
D LQ 1 U W u h
¼ (6.40)
D LQ 2 u h Q1 U 2 W 2 Q2
The technique used to perform the analysis of tracer studies will be the subject of
Example 4.
Example Applications
Example 1: Determination of retardation coefficient As part of a forensic investi-
gation of a continuous Malathion spill, you need to determine the retardation
coefficient of the soil at the site for Malathion. You have decided to do so in
a column experiment with the soil, illustrated in Fig. 6.7. Also given in the figure
are the results of a pulse test with the non-sorptive tracer, chloride, and the results of
a pulse test with Malathion. What is the retardation coefficient, R?
Chloride:
tr = tr1 = 20 min.
st1 = 3 min.
Malathion:
tr = tr2 = 400 min.
st2 = 66 min.
This test can result in a retardation coefficient from a comparison of both residence
times and the variance.
Residence times: The residence time of the chloride can be given as tr1 = L/U, and
the residence time of the Malathion can be given as tr2 = R L/U. Thus,
130 J.S. Gulliver
Fig. 6.7 Illustration of the

column test for retardation 1.0
coefficient and results of the
tracer tests
0.8
2DLW2
s x2 ε y
0.6
0.4
0.2
0
0.2 0.4 0.6 0.8 1.0 1.2
x* εy/UW2
tr2 400 min

R¼ ¼ ¼ 20
tr1 20 min
Variance of tracer curves: Eq. 6.27 provides the relationships:
s2 ¼ st 2 =tr 2 ¼2DL Cou=ðULÞ:
Therefore, assuming that Cou = 1,
2DL R 2 2DL 2 tr2

s2t2 ¼ tr2 ¼ tr2 ¼ s2t1 22 ¼ R2 s2t1
LR U UL tr1
or,
R ¼ st2 =st1 ¼ 66 min=3 min ¼ 22
Our two means of determining the retardation coefficient in the column gave R =
20 and R = 22. We can also check whether the organic carbon content of the soil fits
what is generally known from the literature. First, there is the relation for R:
rB
R¼1þ Kd (6.41)
e
Second, we have the equation from Karikhoff et al. [10]:

cm3
kd ¼ 0:41 f Kow (6.42)
g
Soil
Sandstone
V
Atrazine contamination
Fig. 6.8 Illustration of well and Atrazine spill (From [5])
Combining Eq. 6.29 and 6.30 results in an equation for the organic fraction:
ðR 1Þe
f ¼ ¼ 0:030 0:033 (6.43)
0:41 rB ðg=cm3 Þ Kow
For Malathion, Kow = 230. From soil tests, e = 0.3, and rb = 2.0 g/cm3. Then
Eq. 6.43 gives f = 0.030 0.033. This number is about right for the organic-rich soil
that was tested. Thus, we know that our column tests are of the right order.
Example 2: Atrazine spill into an irrigation well (three-dimensional dispersion
with convection)Three kilograms of 1,000 g/m3 Atrazine pesticide is accidentally
dumped down an old farm irrigation well, placed to pump water out of porous
sandstone, as illustrated in Fig. 6.8. Estimate the movement of the Atrazine plume
over time and the concentrations in the plume.
Given:
U = 10–4 m/s d (sandstone) ffi 1 mm
f = 10–4 (not soil) D = 10–10 m2/s
e = 0.3 rb/e = 6 g/cm3
Our solution is [4]:
M
C¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
8eðpt=RÞ3=2 Dx Dy Dz
" 2 # (6.44)
R x UR t Ry2 Rz2
exp
4 Dx t 4 Dy t 4 Dy t
Now, estimate the maximum concentration location of this maximum and spread
of the Atrazine cloud as a function of time.
Dispersion coefficients: As a first guess, let us use the empirical relations
provided in Eq. 6.34 and 6.35. We will assume no heterogeneity in rock porosity
and no lenses.
132 J.S. Gulliver
Dx DL Ud
¼ ffi Pe ¼ or
D D D
Dx ¼ ð104 m/sÞð103 mÞ ¼ 107 m2 /s
Dy Dz
¼ ffi 0:1 Pe or
D D
Dy ; Dz ¼ 0:1 ð107 m2 /sÞ ¼ 108 m2 =s
Retardation Coefficient:
Lehman et al. [11] gives:
log (K0w) = 2.75 for Atrazine
∴K0w = 102.75 = 562
Using Karikhoff et al. [10] relationship:
Kd = b f K0w where b = 0.41 cm3/g and f = 104
Then
Kd = 0.41 cm3/g (104)(562) = 0.023 cm3/g
and
rb g
R ¼ 1þ Kd ¼ 1 þ 6 ð0:023 cm3 =gÞ ¼ 1:14
e cm3
Even with a fairly sorptive organic compound, the retardation coefficient in rock
is not much different from 1.0. Eq. 6.44 gives a maximum at y = 0, z = 0, x = U t/R.
Thus,
M
Cmax ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
8eðpt=RÞ3=2 Dx Dy Dz
and
Ut
Xmax ¼ ¼ 0:88 104 ðm=sÞ t ðsÞ
R
We will indicate spread by 4s, which corresponds to 95% of the total mass of the
cloud, for a Gaussian distribution like this equation provides.
For 1-D diffusion: s2 = 2 Dt/R
pffiffiffiffiffiffiffiffiffiffi
2 Dt=R
For 3-D diffusion: s2 ¼ 29 Dt=R s ¼ 3
and
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
4 2Dx t=R
4sx ¼
3
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2Dy t=R
4
4sz ¼ 4sy ¼
3
Table 6.4 Estimated concentration over time and space with transport through a uniformly
porous sandstone
Time Cmax (g/m3) Xmax (m) 4sx (m) 4sy (m)
1 hr 4.0 108 0.32 0.04 0.014
1 day 3.4 106 7.6 0.17 0.054
1 month 2.1 104 229 0.90 0.29
1 year 500 2,800 3.1 0.98
2 years 180 5,500 4.4 1.39
10 years 16 27,000 9.9 3.13
Table 6.5 Estimated concentration over time and space with transport through a sandstone media
with 1 m heterogeneities
Time Cmax(g/m3) Xmax(m) 4sx(m) 4sy(m)
1 hr 1.25 0.32 1.3 0.4
1 day .011 7.6 5.5 1.7
1 month 6.6 105 229 29 9.2
1 year 1.57 106 2,800 99 31
2 years 5.6 107 5,500 140 44
10 years 5.0 108 27,000 320 99
Note that in Table 6.4, the concentrations at 1 h, 1 day, and 1 month are above
the initial Atrazine concentration (1,000 g/m3). This is one problem with Dirac delta
boundary conditions because they initially have no volume, only mass. At greater
elapsed time, however, the inaccuracies of the Dirac delta solution have a minimal
impact on the resulting concentration.
However, virtually all media have heterogeneities of high and low porosity
regions, as well as lenses that form around cracks. If we assume that the spacing
of these regions is a mean of 1 m, instead of the 1 mm grain size, then our estimated
dispersion coefficients are increased by a factor of 103. Applying the above
equations to these parameters results in a reduction in Cmax by a factor of 3.2
104 and an increase in both 4s values by a factor of 32. The resulting estimates of
the pertinent parameters are given in Table 6.5. These values are likely to be more
realistic for transport through a groundwater aquifer.
The unknown dispersion coefficient is not uncommon in groundwater transport
problems. It is typically one of the parameters fitted to measurements in groundwa-
ter transport.
Example 3: Drinking water pollution by trichloroethylene (steady state ground-
water transport with lateral dispersion) A military ammunition plant in Arden
Hills, Minnesota, used trichloroethylene (TCE) as a metal cleaning solvent for many
years. Trichloroethylene is currently believed to be a carcinogen. Unaware of the
hazardous nature of TCE, plant personnel placed the waste grease and TCE in
134 J.S. Gulliver
a trench to burn (the grease) or soak into the ground (the TCE) and disappear from
sight for many years. What was not known, however, is that they were placing the
TCE into an aquifer that surfaces near the armory. Four kilometers downstream, the
City of New Brighton used this aquifer as a source of municipal water supply. What is
the expected TCE concentration in the New Brighton water supply and what should
be done in the adjacent cities? Is the TCE plume sufficiently captured by the New
Brighton wells? The following conditions were approximated from available data:
Supply of TCE = 100 kg/day
U = 1.6 105 m/s
Aquifer thickness, H = 30 m
R ffi 1:0 in the aquifer for TCE.
Aquifer porosity, e = 0.3
New Brighton extraction, Q = 0.25 m3/s
Drinking water recommended limit for TCE = 5 mg/L (5 103 g/m3)
As a first assumption, we will assume that the New Brighton well was located in
the center of the plume, compute the capture zone, and then the concentrations
within this capture zone. The capture zone is given by:
Q
Y¼ ¼ 1740 m (6.45)
eHU
At a velocity of 1.6 105 m/s, the 4 km distance would be covered in 8 years,

which is short compared to the 40 years of dumping TCE. We will therefore
assume that the system is at steady state. We have these boundary conditions:
At x = 0, M_ ¼ 100 kg=day ¼ 1:2 g=s
At y = 1, C = 0
At x = 1, C = 0.
The solution with these boundary conditions is:

M_ Uy2
C¼ exp
He ð4pxUDt Þ1=2 4 Dt x
with a capture zone mean of
ð
Y=2
2
C ¼ Cdy
Y
0
where Y is the width of the capture zone, and a leakage from the capture zone, ML of
ð
1
ML ¼ 2eH Cdy
Y=2
1.E+01
y=0
y = 100 m
y = 300 m
y = 500 m
y = 700 m
y = 870 m
1.E+00
Concentration (g/m3)
Capture Zone Mean
1.E−01
Concentration at the edge of

the Capture Zone
1.E−02
0 1000 2000 3000 4000
Distance from Armory (m)
Fig. 6.9 Predicted concentration of trichloroethylene in groundwater plume versus distance for
the ammunitions plant release – City of New Brighton case study
Using the rough approximations of Eq. 6.35 and 6.36, we get
DL ¼ 0:0026m2 /s
and
Dt ¼ 2:6 X 104 m2 =s:
Note that Eq. 6.36 gives DL = 0.0026 m2/s with a 67% confidence interval of
between 0.064 and 1 104 m2/s, or 1.4 orders of magnitude. This variation would
need to be considered in any preliminary analysis of this problem.
The solution for this application is given in Fig. 6.9 at various lateral distances
from the peak concentration. The capture zone mean is 0.42 g/m3, or almost
100 times the recommended limit, which would raise concern in New Brighton.
The leakage from the capture zone at x = 4,000 m for this scenario is computed to be
47 g, or sufficient mass to result in a concentration of 9 104 g/m3 for a similar
capture zone. It is possible, then, that almost all of the plume was captured by the
City of New Brighton, so the problem may not cover a wider area than the
136 J.S. Gulliver
immediate downstream cities. This, at least, is one positive result of the low
transverse dispersion of groundwater plumes.
Example 4: Determination of DL in a river A tanker car carrying a solvent
derailed on a bridge and fell into the Nemadji river, Wisconsin. The forensic
investigation team has a computational model that will simulate the spill, if some
coefficients are determined, including DL and tr . The most cost-effective means
of determining these parameters would be to perform a conservative tracer pulse
test and adjust the parameters from discharge on the day of the tracer test (30 m3/s)
to discharge on the day of the spill (70 m3/s) with some predictive equations that
have been developed.
The location of the pulse input is identified as x = 0. The measured tracer
concentrations and other relevant data are given in Fig. 6.10. From this data,
determine the DL and tr parameters on the day of the test for the reach from x =
8,000 m to x = 35,000 m.
Because the variance of the tracer curve grows linearly with distance after x =
x∗, we can make the following statements:
Dtr ¼ tr2 tr1
and
Ds2t ¼ s2t j2 s2t j1
and finally,
Ds2t 2DL 2DL Dtr

¼ ¼
Dt2r UDx Dx2
At x = 8,000 m,
Ð P
Ct dt Ct Dt
tr1 ¼ Ð ffi P
C dt C Dt
Using all of the Dt values set equal to 0.1 h, Dt can be cancelled out of the
equation, and
1:30 h g=m3
tr1 ¼ ¼ 1:94 h
0:67 g=m3
Also,
Ð 2 P 2
Ct dt Ct Dt
s2t j1 ¼ Ð t2r1 ffi P t2r1
C dt C Dt
HWY 35 BRIDGE
W = 30 m X = 8000 m M = 500 g A = 10.78 m2 S = 10−3 h = 0.6 m

0.17
0.16
0.15
0.14
0.13
0.12
Rh concentration (g/m3)
0.11
0.1
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
0.6 1 1.4 1.8 2.2 2.6 3
Time since release (hrs)
CNTY RD C BRIDGE
W = 60 m X = 35,000 m M = 500 g A = 10.90 m2 S = 3 × 10−4 h = 0.6 m
0.035
0.03
0.025
Rh concentration (g/m3)
0.02
0.015
0.01
0.005
0
6 8 10 12 14
Time since release (hrs)
Fig. 6.10 Tracer measurements taken at the Highway 35 and County Road C bridges
138 J.S. Gulliver
or, again cancelling the equal Dt’s:
2:59 h2 g=m3
s2t j1 ¼ ð1:94 hÞ2 ¼ 0:125 h2
0:666 g=m3
We have assumed that the location 8,000 m downstream of the pulse injection
will be out of the “mixing” region, as specified by Eq. 6.39. We are now ready to
check this assumption with the parameters of the stream flow at x = 8,000 m. First,
U = Q/A = 8 m3/s/10.8 m2 = 0.74 m/s, and from the chapter on Turbulent Transport:
ey ffi ð0:6 0:3Þ u h (6.46)
where
pffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u ¼ ghS ¼ ð9:8 m=s2 Þð0:6 mÞ ð103 Þ ¼ 0:077 m=s
Then,
ey ¼ ð0:6 0:3Þ ð0:077 m=sÞð0:6 mÞ

¼ 0:028 0:014 m2 =s
and Eq. 6.41 gives,
0:2 ð0:74 m=sÞ ð30 mÞ2

X ¼ ¼ 4760 m
0:028 m2 =s
With the approximations in Eq. 6.39, x = 8,000 m will be assumed sufficient to

begin our determination of DL, especially since that is one of the few access points
(a bridge) into this reach of the river.
We will now perform a similar calculation on the tracer cloud at x = 35,000 m.
tr2 ¼ 11:54 hr
1 4 hr2 g=m3
s2t j2 ¼ ð11:54 hrÞ2 ¼ 0:162 hr2
0:105 g=m3
Dx ¼ 35; 000 m 8; 000 m ¼ 27; 000 m
Now,
Ds2t Dx2 ð0:162 0:125Þ hr2 ð27; 000 mÞ2

DL ¼ ¼
2Dt3r 2 ð11:54 1:94Þ3 hr3
or,
DL ¼ 15; 200m2 /hr ¼ 4:2m2 =s
We can use Eq. 6.37 to adjust our dispersion coefficient from the 8 m3/s with
0.6 m mean depth on the day of the tracer test to the 3 m3/s with 0.4 m mean depth
that existed as the river discharge on the day of the spill:
DL ¼ 0:011 ðQ=WhÞ2 W 2 =ð ghS hÞ (6.47)
and then, assuming that the slope does not change and that the banks are fairly
steep, such that dA = W dh, and assigning the subscripts t and s to indicate tracer and
spill:
2 7=2
Qs ht
DLS ¼ DLt
Qt hs
3
m 3 m =s 0:6 m
2
¼ 4:2 ¼ 2:4 m2 =s
s 8 m3 =s 0:4 m
In addition, tr can be adjusted as well:
Qt hs 8m3 =s 0:4 m
Dtrs ¼ Dtrt ¼ ð11:54 1:94Þ h
Q s ht 3 m3 =s 0:6 m
so
Dtrs ¼ 17 h
The use of an empirical relation, such as Eq. 6.37, to adjust parameters for
discharge is more accurate than simply using the equation itself, because the
coefficient and other variables that can have substantial uncertainty are eliminated
from the equation.
Conclusions
Dispersive transport is handled in a manner similar to turbulent diffusion and

diffusion, with almost the same equations and solutions. The primary difference
is that a dispersion coefficient is one or two orders of magnitude greater than
turbulent diffusion coefficients and about 10 orders of magnitude greater than
diffusion coefficients. Assuming that the dispersion coefficient will result in
140 J.S. Gulliver
a Gaussian distribution of a chemical is only a rough approximation, however,

which can be inaccurate at the trailing edge of the chemical cloud.
Future Directions
As computational power of our computers continues to improve, less attention will

be paid to dispersive transport in the environment. The merging of velocity profiles
and mixing to make a dispersion coefficient will be of less value when a more exact
solution is available on desktop computers. The simplicity of the analysis, however,
still will make the dispersion equations valuable in developing an understanding of
mixing and transport problems in environmental settings.
Bibliography
Primary Literature
1. Taylor GI (1953) Dispersion of soluble matter in solvent flowing slowly through a tube. Proc
R Soc Lond Ser A 219:186
2. Taylor GI (1954) The dispersion of matter in turbulent flow through a pipe. Proc R Soc Lond
Ser A 223:446
3. Elder JW (1959) The dispersion of marked fluid in turbulent shear flow. J Fluid Mech 5:544
4. Crank J (1975) The mathematics of diffusion, 2nd edn. Oxford University Press, Oxford
5. Gulliver JS (2007) An introduction to chemical transport in the environment. Cambridge
6. Freeze RA, Cherry JA (1979) Groundwater. Prentice-Hall, Englewood Cliffs
7. Koch DL, Brady JF (1985) Dispersion in fixed beds. J Fluid Mech 154:399
8. Gelhar L, Welty C, Rehfeldt KR (1992) A critical-review of data on field-scale dispersion in
aquifers. Water Resour Res 28(7):1955
9. Fisher HB (1973) Longitudinal dispersion and turbulent mixing in open-channel flow. Ann
Rev Fluid Mech 5:59
10. Karikhoff SW, Brown DS, Scott TA (1979) Sorption of hydrophobic pollutants on natural
sediments. Water Res 13:241
11. Lehman WJ, Reehl WF, Rosenblatt DH (1990) Handbook of chemical property estimation.
American Chemical Society, Washington, DC
Books and Reviews
Fisher HB, List JE, Koh RCY, Imberger J, Brooks NH (1979) Mixing in inland and coastal waters.
Academic, San Diego
Kreyszig E (1982) Advanced engineering mathematics, 4th edn. Wiley, New York
Levenspiel O (1962) Chemical reaction engineering. Wiley, New York
Chapter 7
Transport with Jets and Plumes of Chemicals
in the Environment
Wenming Zhang, Nallamuthu Rajaratnam, and David Z. Zhu
Glossary
Bubbly jet The jet produced by injecting gas-liquid mixture into a liquid.
Buoyant jet The plume with momentum or jet with buoyancy.
Circular jet The jet produced through a nozzle with a circular cross section.
Diffuser The device which has multiple nozzles to quickly mix the
discharged substances (e.g., effluent or air) with the surrounding
ambient fluid.
Jet The flow generated by the release of momentum usually through
a nozzle or slot.
Jet in coflow The jet discharged in the direction of a flowing ambient fluid.
Jet in crossflow The jet discharged at an oblique angle to a flowing ambient fluid.
Plane jet Also called “slot jet” or “two-dimensional jet”, the jet produced
through a slender slot.
Plume The flow generated by the release of buoyancy.
Surface jet The jet discharged at (or near) the surface of an ambient fluid.
Slurry jet The jet produced by injecting solid-liquid mixture into a liquid.
Wall jet The jet discharged tangentially or at a certain angle to a solid
boundary (wall).
W. Zhang • N. Rajaratnam • D.Z. Zhu (*)
Department of Civil and Environmental Engineering, University of Alberta,
T6G 2W2 Edmonton, AB, Canada
e-mail: [email protected]; [email protected]; [email protected]
142 W. Zhang et al.
Definition of the Subject and Its Importance
Jets and plumes are common in our environment. Some examples of jets are:
wastewater discharged from an outfall, emission from an aircraft or vehicle, and
the eruption of volcano. Some examples of plumes are: the smoke from a chimney
stack or cigarette, the thermal plumes from a fire, municipal wastewater or hot
water discharged in deep water, and oil spill from sea bed. One of the most
important features of jets or plumes is its ability of entraining ambient fluid to
achieve self-dilution. This greatly triggers our interests to study jets and plumes.
This book chapter is a review of the studies on turbulent jets and plumes, with
a focus on the transport of conservative pollutants.
Introduction
The earliest experimental study of turbulent jets appears to be the work of Trupel on
circular jets in 1915 [1]. Förthmann performed an experimental study of plane
turbulent jets in 1934 and his work also considered plane turbulent wall jets [23].
The results of these investigations showed the similarity of the velocity profiles at
different distances from the sources of the jets. These studies were followed by the
extensive investigations of Hinze and Zijnen [28] on circular jets and Albertson
et al. [4] for plane and circular jets. Turbulence characteristics of plane jets were
studied by Heskestad [27] for plane jets and Wygnanski and Fielder [105] for
circular jets. Theoretical solutions for plane jets were developed by Tollmien in
1926 for plane jets and by Goertler in 1942 for circular jets [73]. Numerical studies
of turbulent jets followed, starting with the work of Rodi and Spalding [82].
Abramovich [1], Rajaratnam [73], and Fischer et al. [21] provided comprehensive
treatment of jets.
Turner [97] provides an introduction to study turbulent plumes in his book on
Buoyancy Effects in Fluids. Rouse et al. [84] performed an experimental study of
plane and circular plumes wherein they found that the velocity and density defect
profiles were similar if proper scales were chosen for velocity, width, and density
defect. Morton et al. [58] published an integral study of plumes, wherein the
concept of entrainment coefficient was introduced. Since then numerous
studies have been conducted on turbulent plumes and forced turbulent plumes
(buoyant jets). These studies have been summarized in Chen and Rodi [15] and
Lee and Chu [45]. Turbulent jets and plumes have been studied extensively not only
because these flows are very interesting but also that they are of considerable
practical importance in the fields of hydraulic, mechanical, aeronautical, environ-
mental, and chemical engineering and many other fields.
This book chapter will first review the most classic and well-established theories
on simple jets or plumes in stagnant water, and then consider effects of boundaries
including: the bed (wall) and the surface of ambient fluid; coflowing and
7 Transport with Jets and Plumes of Chemicals in the Environment 143
cross-flowing ambient fluid; and the interaction of neighboring jets in the case of
multiple jets. Next, two kind of multiphase jets and plumes – bubbly jets and
plumes, and slurry jets and plumes – will be briefly introduced. Multiphase jets
and plumes are much more complicated compared to the single-phase ones, but
have gained more interest in recent years because of their wide applications.
Finally, some directions for future researches will be highlighted.
Turbulent Jets and Plumes in Stagnant Environment
In this section, the focus is on the transport of conservative pollutants in a steady-

state turbulent jet or plume issuing from a simple (plane or circular) nozzle into
stagnant ambient fluid of large extent. Such jets or plumes are called simple jets or
simple plumes. Theories in this area have been well established. Close to the nozzle
exit, there is a wedge-like or cone-like region termed “potential core” where the
width of initial velocity distribution decays to a point. The length of the potential
core is very short, about 10.4b0 for a plane jet where b0 is the half slot width, or 6.2d0
for a circular jet where d0 is the nozzle diameter [45]. Therefore, for practical
purposes, our attention will be limited on the flow beyond the potential core, i.e., in
the “fully developed flow” region (see Fig. 7.1). The reader who is interested in flow
development region may refer to Rajaratnam [73].
Simple Jets
The integral method is the most common method for analyzing simple jets or
plumes. The following is a brief introduction on this method. For a plane jet as
C/Cm Entrainment
u/um
1 1
u/um C/Cm
2 2
b bc
Fully
developed
flow region
Flow Potential
development core
region
U0
Fig. 7.1 Schematic of y

a simple plane jet (for circular
jet, replace 2b0 by d0 and y by r) 2b0
144 W. Zhang et al.
shown in Fig. 7.1, the Reynolds-averaged Navier-Stokes equation in x direction,

continuity equation, and pollutant conservation equation, respectively, can be
simplified as [73]:
@u @u 1 @t
u þv ¼ (7.1)
@x @y r @y
@u @v
þ ¼0 (7.2)
@x @y
@uC @vC @2C

þ ¼e 2 (7.3)
@x @y @y
where u and v are the time-averaged velocities in x and y directions, respectively;

C is the pollutant concentration; r is the density of the fluid; t is the turbulent shear
stress and e is the mean value of turbulent diffusion coefficient. After multiplying
Eq. 7.1 by r and then integrating from y = 0 to y = 1, Eq. 7.1 becomes:
Z 1
d
ru2 dy ¼ 0 (7.4)
dx o
Equation 7.4 states that the jet momentum flux at different x-sections is
conserved. Using Eq. 7.2 and integrating the first term of Eq. 7.2 from y = 0 to y
= 1, we have:
Z 1
dqx d
¼ u dy ¼ vy¼1 (7.5)
dx dx 0
where qx is the jet volume flux per unit slot length. The entrainment hypothesis
assumes that the entrainment velocity ve ¼ vy¼1 ¼ ae um , where ae is called the
jet entrainment coefficient, um is jet centerline (maximum) velocity, and the negative
sign indicates the ambient fluid is entrained into the jet. Equation 7.5 says that the jet
volume flux increases with traveling distance x due to the entrainment of ambient
fluid, which explains the ability of jet in diluting pollutants. Similarly, integrating
Eq. 7.3 from y = 0 to y = 1, we have:
Z 1
d
uC dy ¼ 0 (7.6)
dx 0
Equation 7.6 states that the mass fluxes of pollutants at different x-sections are
conserved if the chemical or biological reactions of pollutants are not considered.
Numerous laboratory experiments and numerical simulations have confirmed
that beyond the potential core, the jet velocity or concentration exhibits
self-similarity. The most widely used expression for such similarity is the Gaussian
distribution, which represents laboratory data satisfactorily:
y 2
u
¼ exp k1 (7.7)
um b
" 2 #
C y
¼ exp k1 (7.8)
Cm k2 b
where Cm is the jet time-averaged centerline (maximum) concentration, b is the

jet velocity half-width where the velocity is 50% (if k1 = 0.693, refer to Fig. 7.1) or
37% (if k1 = 1) of um, k2b defines the jet concentration half-width where the concen-
tration is 50% or 37% of Cm, and k2 is the ratio of concentration half-width to velocity
half-width. Using Eqs. 7.4–7.8, the analytical solutions for plane jets can be derived, as
shown in Table 7.1. The coefficients of the solutions are mainly determined from
experimental results and integrations.
Using the same procedures as above, the equations of momentum flux, volume
flux, and pollutant mass flux for a circular jet can be derived. These equations
suggest that the jet momentum flux at any x-section is conserved and the jet volume
flux across any x-section increases with traveling distance due to the entrainment
of ambient fluid. Although the entrainment causes the pollutant to get diluted
within the jet core, the mass flux of any conservative pollutant at any x-section
is conserved. The solutions of these equations for circular jets are summarized
in Table 7.1.
Simple Plumes
A plume is produced from a steady discharge of a fluid whose motion is controlled by

its buoyancy, with negligible effect of initial momentum. First, a plane plume of
density r0 issued into a stagnant unstratified ambient fluid of density ra is considered.
Dr
It is assumed that ra
0
1 (true for most practical cases), where the initial density
defect Dr0 ¼ ra r0 . After some manipulation, it can be shown that the Reynolds-
averaged Navier-Stokes equation in x direction becomes:
@u @u 1 @t Dr
u þv ¼ þg (7.9)
@x @y ra @y ra
where Dr ¼ ra r and r is the plume density. For a plane plume, the continuity
equation and pollutant conservation equation can be simplified in the same form as
for a plane jet (Eqs. 7.2 and 7.3).
146 W. Zhang et al.
Table 7.1 Summary of mean properties of simple plane jet and circular jet
Parameter Plane jet Circular jet
Maximum C0 ¼
Cm pffiffiffixffi , where a1 = 3.37 in [21], CCm ¼ ax1 , where a1 = 4.96 in [21],
a1
0
(centerline) b0 d0
concentration Cm 3.45 in [74], 3.21 in [45] 5.34 in [74], 5.26 in [45]
Cavg ¼ a2 , where a2 = 1.2 in [21], Cavg ¼ a2 , where a2 = 1.4 in [21],

Cross-sectional Cm Cm
average 1.25 in [45] 1.76 in [74], 1.68 in [45]

concentration
Cavg
Maximum U0 ¼
um pa3ffiffiffixffi , where a3 = 3.50 in [73], Uum ¼ ax3 , where a3 = 6.3 in [73], 6.2
0
b0 d0
(centerline)
3.41 in [21], 3.65 in [45] in [21] and [45], 6.13 in [74]
velocity um
Velocity half-width b = a4x, where a4 = 0.10 in [73], b = a4x, where a4 = 0.10 in [73],
ba 0.116 in [21], 0.097 in [74], 0.12 0.107 in [21], 0.096 in [74], 0.114
in [45] in [45]
b ¼ a5, where a5 = 1.35 in [21] and b ¼ a5 , where a5 = 1.19 in [21],
Concentration bC bC
half-width bCa [45], 1.17 in [74] 1.17 in [74], 1.2 in [45]

Entrainment ae = 0.053 in [73] and [45] ae = 0.026 in [73], 0.028 in [74],
coefficient ae 0.057 in [45]
Source: [21, 45, 73, 74].
a
In [21] and [45], b (or bC) is defined as where the velocity (or concentration) is 37% of um (or Cm);
while in others, defined as 50% of um (or Cm).
Using the integral method, Eq. 7.9 can be reduced to:

Z 1 Z 1
d
ru2 dy ¼ gDrdy (7.10)
dx 0 0
Equation 7.10 says that axial momentum flux increases in x direction, and the
increase rate is equal to the buoyancy per unit length (in x direction) of the plume.
For a plane plume, the continuity equation and pollutant conservation equation can
be reduced the same as Eqs. 7.5 and 7.6. Equation 7.5 states that the volume flux of
a plume increases due to the entrainment of ambient fluid, and Eq. 7.6 states that
although the concentration of pollutant decreases, its total mass flux is conserved.
Note that C in Eq. 7.6 can be also interpreted as gDr, and then Eq. 7.6 becomes the
integral form of buoyancy conservation equation.
Experimental results show that: similarly as for jets, the plume velocity or
concentration also exhibit self-similarity and the Gaussian distribution can well
describe it. Using the Gaussian profiles (Eqs. 7.7 and 7.8), the analytical solutions
for a plane plume can be derived as shown in Table 7.2. The coefficients differ
slightly in different references as they are determined using the results of different
experiments. Here, to constitute the solutions, an useful dimensionless parameter is
indroduced – the densimetric Froude number at the slot exit F0 ¼ qffiffiffiffiffiffiffiffiffiffiffiffi
ffi
U0
Dr
g ra b0
(for circular plume, F0 ¼ qffiffiffiffiffiffiffiffiffiffiffiffi

U0
Dr
).ffi
g ra d0
Table 7.2 Summary of mean properties of plane plume and circular plume
Parameter Plane plume Circular plume
Maximum (centerline) Cm Drm a1 F0 23 Cm Drm a1 F0 3
2
concentration Cm or C ¼ Dr ¼ x=b , where F0 ¼ C ¼ Dr ¼ 53 , where

0 0 0 0 0
density defect Drm x
d0
U0 U
qffiffiffiffiffiffiffiffiffiffiffiffiffi , a1 = 3.78 in [21], 0
F0 ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffi , a1 = 7.75 in [21],
g Dr
ra b0 g Dr
ra d0
3.84 in [74], 4.25 in [45] 7.83 or 9.37 in (from different
methods in Ref. [74]), 8.90 in
[45]
Cross-sectional average Cm Cm
¼ a2 , where a2 = 1.32 in ¼ a2 , where a2 = 1.40 in [21],
concentration Cavg Cavg Cavg
[74], 1.25 in [45] 1.70 in [45]
Maximum (centerline) um ¼ a3 , where a = 2.09 in um
¼
a3
u0 F0 23
3
U0 13 , where a3 = 4.34 in
velocity um 2
F0 3 dx0
[21], 2.52 in [74], 2.85 in [45]
[21], 4.00 or 4.33 in (from
different methods in Ref. [74]),
4.35 in [45]
Velocity half-width ba b = a4x, where a4 = 0.116 in [21] b = a4x, where a4 = 0.100 in [21],
and [45], 0.128 in [74] 0.085 in [74], 0.105 in [45]
Concentration bC =b: ¼ a5 , where a5 =1.35 in bC =b ¼ a5 , where a5 =1.20 in [21],
half-width bCa [21] and [45], 1.17 in [74] 1.16 in [74], 1.19 in [45]
Entrainment coefficient ae = 0.136 in [74], 0.103 in [45] ae = 0.047 in [74], 0.088 in [45]
ae
Source: [21, 45, 74].
a
In [21] and [45], b (or bC) is defined as where the velocity (or concentration) is 37% of um (or Cm);
while in [74], defined as 50% of um (or Cm).
For a circular plume, the Reynolds-averaged Navier-Stokes equations, continu-

ity equation and pollutant conservation equation, can be simplified, and the integral
method can be used to obtain their integral forms of the equations. These equations
indicate that the momentum flux in the plume increases with axial (x) direction, and
the increase rate is equal to the buoyant force per unit axial length; the volume flux
in the plume increases due to the entrainment of ambient fluid; and the pollutant
mass flux (or the flux of density defect) remains invariant in the axial direction.
The analytical solutions for a circular plume are also presented in Table 7.2.
The constants in these equations change slightly in different references where
different experimental results were used.
Buoyant Jets
For a buoyant jet, the initial momentum flux cannot be neglected. Near the nozzle
(slot) exit, it is expected that the buoyant flow will be like a jet; after some distance
from the exit where the increase of momentum flux is much larger than the initial
momentum flux, the buoyant flow will behave like a plume. For a plane buoyant jet,
148 W. Zhang et al.
based on dimensional analysis, a characteristic length scale to judge whether the

buoyant flow behaves like a jet or plume may be defined as:
M0
LM ¼ 2 (7.11)
B0 3
where the initial specific momentum flux M0 = qU0; the initial specific buoyancy
flux B0 ¼ qg Drra . If the jet centerline trajectory l lM, the buoyant jet can be treated
0
as a pure jet and the simple jet equations can be used; if l lM, it can be treated as
a pure plume and the simple plume equations are valid. Buoyant jets will be plume-
like beyond lMl 4 5 [41, 65].
Characteristic length scale of jet/plume can only roughly help us calculate the
evolution of a buoyant jet. The Reynolds equations, continuity equation and
pollutant mass conservation equations can be simplified and the integral method
can be used to obtain the analytical solutions. For a plane buoyant jet, the integral
forms of the momentum, continuity, pollutant mass (or buoyancy) conservation
equations are the same as Eqs. 7.10, 7.5 and 7.6, respectively. Here, for a buoyant
jet, the integral form of energy equation needs to be introduced (as a result of
multiplying Eq. 7.9 by u and integrating from y = 0 to y = 1):
Z 1 Z 1 Z 1
d ru2 @u
udy ¼ t dy þ gDru dy (7.12)
dx 0 2 0 @y 0
Equation 7.12 states that the flux of kinetic energy in the jet plume is decreased by
turbulence production (first term in the right hand side of Eq. 7.12) and increased by
the work done by buoyancy (second term in the right hand side).
The Gaussian type self-similarity equations are still valid for plane buoyant jets.
Using Eqs. 7.7 and 7.8, as well as the experimental results on the jet spreading rate
db bC
dx and on the ratio of the concentration half-width to the velocity half-width b ,
and after some mathematical manipulations, the solutions for a plane buoyant
jet can be obtained as shown in Table 7.3. It is interesting to note that in Table 7.3,
the jet centerline concentration or velocity equation are composed of two
parts, corresponding to two limits (pure jet-like or pure plume-like conditions).
The characteristic length scale for a buoyant circular jet or plume is:
3
M0 4
LM ¼ 1 (7.13)
B0 2
Similarly, for a buoyant circular jet, the integral equations of the momentum,
continuity, pollutant mass (or buoyancy) conservation, and kinetic energy can
be obtained. Using the Gaussian distribution for jet velocity or concentration and
some experimental data, the analytical solutions for buoyant circular jets are shown
in Table 7.3.
Table 7.3 Summary of mean properties of turbulent buoyant jet
Parameter Plane buoyant jet Circular buoyant jet
3
Characteristic length for M0 M0 4
LM ¼ 2
jet/plume LM LM ¼ 1
B0 3 B0 2
Maximum (centerline) Cm Drm a11 1 U0 Cm Drm a11

¼ ¼ ffi,
313 x , where F0 ¼ qffiffiffiffiffiffiffiffiffiffiffiffi ¼ ¼ , where a11 = 100,
concentration Cm or density C0 Dr0 2 C0 Dr0 5 3 13
b0 g Drra b0 a12 x x
defect Drm 6 a12 7 F0 2 d0 þ a13 d0
4F0 2 þ a133 5
2
x a12 = 1,920, a13 = 6,720 in [74]
b0
a11 = 12.75, a12 = 21.19, a13 = 50.0 in [74]
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Cross-sectional average Cm 1 Cm
¼ 1:80 in [74]
concentration Cavg ¼ 1 þ 2 ¼ 1:24, where k2 ¼ bbC ¼ 1:35 in [45] Cavg
Cavg k2
Cm 1 bC
¼ 1 þ 2 ¼ 1:69, where k2 ¼ ¼ 1:2 in [45]
Cavg k2 b
Maximum (centerline) velocity 2 313 2 313
um um 6 a31 a32 7 um 6 a31 a32 7
¼ 4 2 þ 3 5 , where a31 = 21.2, a32 = 50.7 in [74] ¼ 4 þ 3 5 , where a31 = 64.75, a32 =
U0 F0 x 2 U0 2 x x
F0 d0
b0 d0
223.25 in [74]
Velocity half-width b b = a4x, where a4 = 0.097 in [74] b = a4x, where a4 = 0.097 in [74]
7 Transport with Jets and Plumes of Chemicals in the Environment
Concentration half-width bC bC bC
¼ a5 , where a5 =1.18 in [74] ¼ a5 , where a5 =1.16 in [74]
b b
Entrainment Coefficient ae a61 ðaep aej Þ a61
ae ¼ aej þ 3 , where a61 = 0.44, a62 = 20.74, a63
ae ¼ aej þ 2 , where aej and aep are the
a62 F0 a62 þ a63 F0 3
a61 þ 3 2 x
x d0
b0
entrainment coefficients for the plane jet and plume, = 2.38 in [74]
respectively, a61 = 17.1, a62 = 41.4 in [74]
Source: [45, 74].
149
150 W. Zhang et al.
Effect of Boundaries on Jets and Plumes
In this section, the effect of different types of boundaries on jets and plumes will be
considered, including the solid bed (wall), the free surface of ambient fluid, the coflowing
or crossflowing ambient fluid, and the neighboring jets in the case of multiple jets.
Wall Jets
Wall jets are the jets discharged tangentially or at certain angles to a solid boundary
(wall) (see Fig. 7.2). A simple case is first considered: a plane jet discharged
tangentially to a smooth flat plate in deep still ambient fluid of the same kind. For
turbulent plane wall jets with high Reynolds numbers ( R0 ¼ U0vb0 ¼ 104 105 ,
where v is the kinematic viscosity of the fluid) at the slot exit, the length of
the potential core will be (6.1 6.7)b0 [73], which is in the same range as for
1
b0 u
2 m
um
u0 um b um
X
a Potential core
u
m /2
u
m
e
zzl
No
bz
z
by
u
m0 /
2
u
m0
b
x
Fig. 7.2 Schematic of (a) plane and (b) bluff wall jets
Table 7.4 Summary of mean properties of plane wall jet and bluff wall jet
Parameter Plane wall jet Bluff wall jet
Maximum (centerline)
concentration Cm
Cross-sectional average Cavg a1
¼ x , where a1
concentration Cavg C0 b0
= 4.032 in [73]
Maximum (centerline) um a3
¼ qffiffiffiffi , where a3
velocity um U0 x
b0
= 3.50 in [78] & [73]
Velocity half-width b b = a4x, where a4 by x
db ¼ 0:90 þ a41
= 0.068 in [73]; d0 h
dx bz x
= 0.073 in [44] ¼ a42 1:25
B B
where a41 =0.045, a42 =0.20,
B is the nozzle (horizontal)
width in [77] & [73];
dby
¼ 0:048 and db dx ¼ 0:26
z
dx
in [44]
Concentration
half-width bC
Entrainment ae = 0.035 in [73]
Coefficient ae
Source: [44, 73, 77, 78].
simple jets. As expected, experiments show that, near the wall, there exists a thin
layer (boundary layer) where the jet velocity increases from zero at the wall to
a maximum velocity um; above the boundary layer (named free mixing region), the
et velocity decreases from um to zero at some large distance y from the wall.
Similarly as for simple jets, the jet width may be defined as where the jet velocity
is 50% (or 37%) of um and @u @y < 0 (i.e., in the free mixing region). In the boundary
layer region, the boundary layer theories may be used to further divide this
region into two or three sub-layers: in the sub-layer very close to the wall, the
velocity distribution is linear with y; some distance away from the wall, the velocity
distribution can be described by the logarithmic law [87]. For the velocity
distributions of the entire wall jet, after some distance (about 20b0) from the slot
exit, they exhibit self-similarity [23, 99]. Verhoff [99] proposed an empirical
equation which agreed well with the experimental data:
u y17 h y i
¼ 1:48 1 erf 0:68 (7.14)
um b b
Using the equations of motion and the integral method, the following results
could be obtained for plane wall jets: um / x2 ; b / x. The detailed results are
1
listed in Table 7.4. To study the effect of wall roughness on wall jets, readers may
152 W. Zhang et al.
refer to Rajaratnam [72], Tachie et al. [95], Dey et al. [19], and Rostamy et al. [83].
To study the jets impinging on walls, readers may refer to Beltaos and Rajaratnam
[10], Rajaratnam [73], and Chan et al. [14].
For non-buoyant circular wall jets (herein termed “bluff wall jet” to include
semicircular and rectangular wall jets with aspect ratio not very different from
unity; the properties of bluff jets are not very different from those of circular jets),
a number of studies have been conducted: square wall jets by Sforza and Herbst
[88], circular wall jets by Newman et al. [61], bluff (including square, rectangular,
circular, elliptic, and equilateral triangular) wall jets by Rajaratnam and Pani [77],
square wall jets by Lübcke et al. [53], and circular wall jets by Agelin-Chaab [3].
These experiments show that after some distance from the potential core, the
velocity distributions both in the vertical central plane and in the horizontal plane
(see Fig. 7.2) are self-similar. From similarity analysis on the equations of motion
or from dimensional analysis, the following results can be obtained for bluff wall
jets: um / x1 ; by / x; bz / x [73]. Experimental results support these predictions,
and the results are listed in Table 7.4.
Surface Jets
A surface jet can be produced by discharging a fluid at the surface of an ambient fluid
(see Fig. 7.3). One typical example is the surface discharge of heated water from
a power plant through either an open-channel or a pipe into an ocean, a lake, or river.
Rivers flowing into lakes, reservoirs, and oceans and storm water discharges into
rivers may be also viewed as surface jets. In this section, our attention will be limited
in the region from the end of the jet potential core to the end of the near field (where
the mixing is still dominated by the jet momentum and buoyancy). The length of the
pffiffiffiffiffi
near field is in the order of 100 A0, where A0 is the cross-sectional area of the flow at
discharge [74]. For the mixing in the far field (where the turbulence in rivers, lakes,
or oceans dominates further mixing), readers can refer to Fischer et al. [21] and
Rutherford [85].
Non-buoyant plane surface jets in stagnant water are now considered.
Equations 7.1–7.8 for plane submerged jets also work for plane surface jets.
Using the integral method and some mathematical manipulations, some useful
results can be obtained: um / x2 ; b / x; cm / x2 , which are in the same form
1 1
as those for plane submerged jets (Table 7.1). Experiments on plane surface jets
were conducted by Chu and Vanvari [16], Rajaratnam and Humphries [76], and
others. The experimental results are listed in Table 7.5. The results confirm that
essentially a plane surface jet is quite similar to half of the corresponding plane
submerged jets, but with slightly different coefficients. For example, the jet spreading
rate of plane surface jets db dx ¼ 0:07 , smaller than the value of 0.10 for plane
submerged jets.
Fig. 7.3 Schematic b0 um

of (a) plane and (b) bluff U0 x
b u
surface jets 1
u
2 m Surface jet
y
Bed
a
um
d0 U0 by u x
1
u
2 m
Side view
Bed
Side wall
Plan view
z
u
um
U0 bz 1 x
u
2 m
Next, non-buoyant bluff surface jets in stagnant water are considered. From the
experiments of Rajaratnam and Humphries [76], the Gaussian function describes
the velocity distributions well both in vertical (half-Gaussian) and transverse
directions, unless there is excess wave generation at the water surface (in this
case um occurs some distance below the water surface). Using the integral method,
it can be shown that: um / x1 ;bz / x; by / x, which are in the same forms as for
circular submerged jets. Rajaratnam and Humphries’ experimental results show
dx ¼ 0:09, twice of that in the
that the jet spreading rate in the transverse direction db z
db
vertical direction dxy ¼ 0:044. The phenomenon of the several times faster trans-
verse spreading has also been observed in the studies of Anthony and Willmarth [6],
Gholamreza-kashi et al. [24], and Cuthbertson and Davies [18]. These studies
further found that there exists a thin layer (called “surface current”) at the free
surface, which exhibits even faster transverse spreading compared to that below the
154 W. Zhang et al.
Table 7.5 Summary of mean properties of non-buoyant plane surface jet and bluff surface jet
Non-buoyant plane Non-buoyant bluff
Parameter surface jet surface jet
Maximum (centerline)
concentration Cm
Cross-sectional average
concentration Cavg
Maximum (centerline) um a3 um a3
¼ qffiffiffiffi , where ¼ , where
velocity um U0 x U0 x=d0
b0
a3 = 13 in [24]
a3 = 3.1 in [76]
Velocity half-width b db In transverse direction:
¼ a4 , where a4 = 0.07
dx dbz =dx ¼ a41
in [74] & [76] In vertical direction:
dby =dx ¼ a42
where a41 = 0.09 in [74]
& [76], 0.12 (below the
free surface) and 0.22
(at the free surface) in
[24]; and a42 = 0.044
in [74] & [76],
0.025 in [24]
Concentration half-width bC bC =b ¼ a5 , where
a5 = 1.15 in [74]
Entrainment coefficient ae ae = 0.037 in [74]
Source: [24, 73, 74, 76].
layer. Comparing Table 7.5 with Table 7.4, one may find that surface jets are
somehow similar to wall jets, e.g., they both spread faster in the horizontal direction
than in the vertical direction due to the boundary constraint in the vertical direction,
and they have similar forms of jet equations.
For the surface discharges such as heated water into rivers or wastewater into the
oceans, the effect of buoyancy needs to be considered. Experimental results
have indicated that the behavior of buoyant surface jets is mainly controlled

by
three parameters: the Richardson number at the outfall, Ri0 ¼ gd0 Dr0 ra U0 ; the2
depth (thickness) of the surface jet, d0 (or b0 for a plane jet); and the depth of
the surface stratified layer formed at the end of the near field of the surface jet, b1 .
For a buoyant surface jet with a fixed Ri0, depending on the value of bd0 , there could
1
be four possible hydraulic phenomena: a surface jet, a surface (density) jump at the
outfall, a surface jet followed by a surface jump, or a drowned jump. Rajaratnam and
Subramanyan [79] presented a graph to distinguish which of the four possibilities
may happen for a plane buoyant surface jet. For the case of a pure plane buoyant
surface jet (without any jump), the experiments of Rajaratnam and Subramanyan
[79] show that: initially the jet spreading rate db
dx follows the equation of the plane
non-buoyant surface jet, but after some longitudinal distance, the spreading rate
slows down, and eventually the jet thickness approaches a constant. In other words,
generally the buoyancy effect constrains the spreading of a plane surface jet. Their
results also indicate that uum is self-similar at different x-sections; however, these
self-similarities can no longer be described by the Gaussian distribution and seem to
be related with Ri0. The results of Chu and Vanvari [16] reveal that the entrainment
coefficient ae of a plane buoyant surface jet decreases continuously with the increase
of bulk Ri (defined as U0rd0ugDr
3
0
) with x; and ae equals to zero when Ri increases to 0.2.
a m
For bluff buoyant surface jets, a number of experiments have shown that the jet
behavior is strongly affected by Ri0 at the outfall. For convenience, bluff surface
jets may be classified into two classes, the small Ri0 class (Ri0 0.1) and the large
Ri0 class (Ri0 > 0.1). From a number of experiments, the common findings for
the two classes are that the vertical velocity profile u(y) in the center-plane and the
transverse (across the jet) velocity profile u(z) just below the water surface are
self-similar, and the self-similarities can be well described by half-Gaussian or
Gaussian distribution. Using the similarity analysis of the simplified equations of
motion, the following relations can be obtained: for the small Ri0 class, um / x1 ,
Cm / x1 , by / x, bz / x; and for the large Ri0 class, um / x3 , Cm / x3 , by =
1 2
constant, bz / x. The detailed results are listed in Table 7.6.
Jets and Plumes in Coflow
Similarly as solid bed or free surface of ambient fluid, coflowing or crossflowing

ambient fluid itself can be viewed as some sort of boundary affecting jet behaviors.
When jets are discharged in the direction of flowing ambient fluids, this is the
problem of jets in coflow (Fig. 7.4). Extensive experimental and numerical studies
show that beyond the potential core, the jet concentration and the jet excess velocity
relative to the ambient velocity exhibit self-similarity. The self-similarity may be
described by the Gaussian distribution, exponent function, or cosine expression [45,
73]. In the following, plane jets in uniform coflow will be briefly introduced,
followed by circular jets.
For coflowing plane jets, the integral form of equation of motion is:
Z 1
d
ruDudy ¼ 0 (7.15)
dx 0
where Du ¼ u Ua , the jet excess velocity. Equation 7.15 states that the excess
momentum flux is conserved in x direction. Using Eq. 7.15 and similarity analysis,
Rajaratnam [73] obtained the following asymptotic relations: for the strong jet
region (i.e., Du
Ua 1), Dum / x
m 12
and b / x; and for the weak jet region (i.e., Du
Ua
m
1), Dum / x2 and b / x2 . Based on the experimental results in the literature,
1 1
Rajaratnam [73] derived that:

Table 7.6 Summary of Mean properties of buoyant plane surface jet and bluff surface jet
156
Buoyant bluff surface jet Buoyant bluff surface

Parameter Buoyant plane surface jet with Ri0 0.1 jet with Ri0 > 0.1
Maximum (centerline) Cm a1
¼ , where a1 = 2.83 in [74]
concentration Cm C0 x 23
pffiffiffiffi
A0
Cross-sectional average
concentration Cavg
u 1 um a3
Maximum (centerline) um exhibits self-similarity, but cannot um a3 Ri0 10 ¼
¼ qffiffiffiffi , where a3 = 15.3 1=3 , where a3 = 1.25
velocity um be well described by Gaussian profile, x
U0
U0 pxffiffiffiffi 5:0
A
A0 0
as uum 0 at by 1:4 [79]
in [74] & [75] in [74]
Velocity half-width b dby by a41
First bb0 increases linearly as non-buoyant ¼ a41
dx
¼
surface jet to some point bx0 (the dbz
d0 Ri0 1=8
location depends on Ri0); then the ¼ a42 bz x
dx pffiffiffiffiffi ¼ a42 pffiffiffiffiffi þ 2:0
increase rate db where a41 decreases with Ri0 A0 A0
dx decreases nonlinearly,
finally approaches asymptotically (a41 = 0.044 for Ri0 =0, 0.02 where by∗ is the average of by which
a horizontal line [79]. for Ri0 =0.038, 0 for Ri0 =0.09); changes slightly with y; a41 =0.29
a42 increases with Ri0 in [74] in [63], 0.26 in [74]; a42 =0.54
& [75] in [74]
Concentration byC byC
¼ a51 ¼ a51
half-width bC by by
bzC bzC
¼ a52 ¼ a52
bz bz
where a51 = 1.0 in [74] & [75] and where a51 = 1.12 in [63] & [74];
a52 = 1.15 in [74] and a52 = 1.9 in [63], 1.6 in [74]
Entrainment ae decreases from about 0.04 to 0 when
coefficient ae Ri increases from 0 (non-buoyant
plane surface jet) to 0.2 [16]
Source: [16, 63, 74, 75, 79].
W. Zhang et al.
Fig. 7.4 Schematic of plane Ua

jets in coflow (for circular
jets, replace 2b0 by d0 and y Ua Ua
by r)
Δu Δum/2
U0 y b
u
2b0 x
Δum Δum
um
Dum 3:41
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ qffiffiffiffi (7.16)
U0 ðU0 Ua Þ x
b 0
b x 1
¼ 0:118 qffiffiffiffi (7.17)
b0 b0 1 þ pffiffiffiffiffiffiffiffiffiffi
0:41 ffi x
b0aða1Þ
where a ¼ U Ua , the ratio of jet exit velocity to ambient velocity. Note that Rajaratnam
o
[73] also summarized other more complex forms of equations for Dum and b, which
were derived by Patel [68] and Pande and Rajaratnam [62].
For circular jets in coflow, the integral momentum equation can be derived:
Z 1
d
r2pruDudr ¼ 0 (7.18)
dx 0
which says that the jet excess momentum is conserved in axial direction. Using
Eq. 7.18 and similarity analysis on the equations of motion, Rajaratnam [73]
presented the following asymptotic relations: for the strong jet region (i.e., Du
Ua 1
m
Dum
), Dum / x1 and b / x; for the weak jet region (i.e., 1), Dum / x3 and
2
Ua
1
b / x3 . Pande and Rajaratnam [62] proposed a complex expression for Dum.
Lee and Chu [45] also derived asymptotic solutions for circular jets in coflow: for
the strong jet region, the jet solution is assumed to be the same as in stagnant water
(see Table 7.1); for the weak jet region,
23
Dum x
¼ 2:14 (7.19)
Ua lm
13
b x
¼ 0:385 (7.20)
lm lm
158 W. Zhang et al.
1=2
Me0
where lm* is the excess momentum length scale defined as Ua ; Me0 ¼ ðU0 Ua Þ
U0 pd0 2
4 , the jet specific excess momentum at discharge.
To completely model circular jets in coflow, Lee and Chu [45] formulated
an integral model based on a Lagrangian jet spreading hypothesis:
1
U 2 þ U ¼0 (7.21)
pB 2
dB U
¼ bs (7.22)
dx 1 þ U
where U ¼ DU
Ua ; B ¼ lm ; DU and B are, respectively, the excess velocity and half
B
of the width of the top-hat profile (instead of the Gaussian profile) of an equivalent
jet, which carries the same mass flow and excess momentum flux as the actual jet;
Dum
pffiffiffiffiffi
and bs ¼ dBdx in stagnant water. It can be proved that DU ¼ 2 and B ¼ 2b, where
Dum and b are, respectively, the maximum excess velocity and 37% half-width for
the Gaussian profile. From Eqs. 7.21 and 7.22, DU and B can be solved. The actual
jet centerline dilution can be obtained:

C0 l2 pB2 Ua þ 2
2 DU
Sc ¼ ¼ 2
1þl
(7.23)
Cm 2Q0
where Q0 and C0 are the initial jet discharge and concentration, respectively;
l is the ratio of concentration half-width to velocity half-width using the Gaussian
profile (l 1.2). The modeling results of Lee and Chu [45] reveal that the centerline
dilution of a circular jet in coflow is only slightly smaller than that in stagnant water;
and the centerline excess velocity decays in a similar way as in stagnant water.
Jets and Plumes in Crossflow
Now consider a non-buoyant circular jet discharged at an oblique angle (not 0 or

180 degree) to a flowing ambient fluid. In fact, most outfalls or diffusers in oceans
or rivers discharge effluents as jets in crossflow, as the jet (effluent) dilution can be
considerably enhanced even in a weak crossflow [45, 73]. Jets in crossflow have
been studied extensively by Abramovich [1], Rajaratnam [73], Fischer et al. [21],
Wright [102, 103], Andreopoulos [5], Hodgson and Rajaratnam [29], Margason
[54], Smith and Mungal [91], Lee and Chu [45], Huang et al. [31], Kikkert et al.
[38], and others. According to these studies, the evolution of jets in crossflow can be
divided into three regions: the potential core region, the maximum deflection
region, and the vortex region (see Fig. 7.5). The length of the potential core has
Fig. 7.5 Schematic of jets

in crossflow x
y
Ua
Vertex zone
A
Maximum Wz
deflection Wy
A
zone
Section A-A
Potential core region
x
U0
d0
been found to be mostly controlled by the relative strength of the jet compared to
U
the crossflow (a ¼ 0 ), and typically in the range of 2–6d0 which is smaller than
Ua
that of a free jet [70].
Beyond the potential core region, the jet would be largely deflected due to the
stagnation pressure exerted by the free stream and the entrainment of ambient fluid
(and thus horizontal momentum). Jet deflection probably is the most distinct feature
in crossflow. After the maximum deflection region, the jet would be gradually
parallel to the direction of ambient flow. Laboratory experiments [1, 29, 45]
and numerical simulations [45] have found that after some distance beyond the
potential core, the jet cross section would be like a kidney shape with a pair of two
counter-rotating vortices (see Fig. 7.5). The vortex pair significantly entrains
ambient fluid in the form of tornado vortices into the jet [45], which explains the
considerable enhancement of jet dilution in crossflow. The concentrations at the
centers of the two vortices have been found to be about 1.1–1.6 times of the jet
centerline concentration [29, 45].
For non-buoyant jets in crossflow, it is common to analyze them in three regions:
the momentum dominated near field (MDNF), the momentum dominated far field
(MDFF), and the transition between MDNF and MDFF. If the jet trajectory x lm,
where lm is defined as:
1
Mv0 2
lm ¼ (7.24)
Ua
and Mv0 is the vertical momentum at the exit, then the jet is in MDNF, where the
effect of jet momentum is much stronger than that of the ambient crossflow. In
MDNF, the classic equations for jets in stagnant ambient fluid are approximately
valid. If x lm, the jet is in MDFF, where the effect of ambient crossflow is
dominant over the jet momentum. In MDFF, the jet properties can be studied with
physical and numerical models. Fischer et al. [21] used dimensional analysis to find
the asymptotic formulas for MDNF and MDFF. Lee and Chu [45] studied the jets in
160 W. Zhang et al.
MDFF based on the analogy to advected line puffs. Using length-scale analysis and
numerical models, they proposed the formulas which can represent satisfactorily
the experimental results:
13
yc x
¼ 1:56 (7.25)
lm lm
bvc ¼ 0:28yc (7.26)
U a yc 2
Sc ¼ 0:46 (7.27)
Q0
where yc is the vertical location of the centerline concentration, bvc is the vertical
centerline half-width defined by 37% of the centerline concentration and Sc is the
centerline dilution.
The most common jet discharge angle in crossflow is 90 degree, i.e., jets are
discharged at right angle to the ambient crossflow. Rajaratnam [73] summarized the
early studies in 1950s to 1970s that mostly focused on jet trajectories. Hodgson and
Rajaratnam [29] conducted detailed laboratory experiments on circular jets at right
angle to crossflow and proposed the following equations:
0:56
ax
Sc ¼ 1:09 (7.28)
d0
0:26
yc x
¼ 1:46 (7.29)
ad0 ad0
0:29
Wz x
¼ 1:20 (7.30)
ad0 ad0
0:37
Wy x
¼ 0:78 (7.31)
ad0 ad0
where Wz and Wy are the jet width and thickness (see Fig. 7.5). Equations 7.28–7.31
have also been validated by a field experiment in the Lesser Slave River, Canada.
Hodgson and Rajaratnam’s equations are mainly derived based on the experiments
conducted in the range of adx 0 ¼ 1 1; 000. It is interesting to note that Hodgson
and Rajaratnam’s expressions fit the experimental data satisfactorily both in
MDNF, MDFF, and the transition between the two.
Now consider a circular plume in crossflow. Similarly as lm, a length scale lb
needs to be defined to compare the relative strength of plume buoyancy with
crossflow:
B0
lb ¼ (7.32)
Ua 3
If the jet trajectory x lb , then the plume is in the buoyancy dominated near
field (BDNF) where the effect of buoyancy is dominant over crossflow; if x lb,
then plume is in the buoyancy dominated far field (BDFF) where the effect of
crossflow is more pronounced than the buoyancy. In BDNF, the plume is essentially
vertical and only slightly advected, thus the equations for plumes in stagnant fluid
are approximately valid. Similarly as jets in crossflow, after some distance from the
nozzle, the plume cross section will become a kidney shape that is made up of
a vortex pair, and the concentration at the centers of the vortices have been found to
be 1.4–1.7 times of the plume centerline concentration. In BDFF, the plume bends
over and finally approaches the ambient flow direction. The analysis on the plume
properties in BDFF relies on experiments or numerical models. Based on the
equations of motion and the use of similarity solutions, Fischer et al. [21] derived
asymptotic formulas for the BDNF and BDFF. As the plume in the BDFF behaves
similarly as the advected line thermal, Lee and Chu [45] used numerical models to
obtain the plume characteristics. The predictions are comparable to experimental
results. The formulas Lee and Chu derived are:
23
yc x
¼ 1:3 (7.33)
lb lb
bvc ¼ 0:4yc (7.34)
U a yc 2
Sc ¼ 0:46 (7.35)
Q0
Note Eq. 7.35 for BDFF is exactly in the same form as the equation for MDFF.
Now consider the case of a circular buoyant jet in crossflow. In this case, the
relative strengths of buoyancy, momentum, and crossflow need to be considered.
If lb lm, then the jet would in sequence experience in MDNF, MDFF, and
BDFF; if lb lm, then the sequence would be MDNF, BDNF, and BDFF [21, 43,
45]. Equations should be selected carefully according to the studied location of
the jet (e.g., in MDNF or MDFF or BDNF or BDFF). The transitions between
MDNF and MDFF and between BDNF and BDFF are better treated using numer-
ical models [45].
For jets directed at an oblique angle to crossflow, the reader may refer to Platten
and Keffer [69] and Kikkert et al. [38]. For plane jets and plume in crossflow, the
reader may refer to Girshovich [25], Jones and Wille [36], Kalita et al. [37], and
Huang et al. [31].
162 W. Zhang et al.
Fig. 7.6 An example Bed

of a unidirectional Ports
diffuser
Risers
Diffuser base
Effluent
Multiple Jets
Effluents may be discharged via single port outfalls or multiport diffusers.

Multiport diffusers are commonly used given their fast mixing and diluting ability
and thus less adverse impacts on the environment. The jets issuing from the ports of
a multiport diffuser are usually viewed as multiple jets. The characteristics of
multiple jets are primarily determined by the arrangements of multiport diffusers.
Generally, multiport diffusers can be classified into three categories: unidirectional
diffuser (where net horizontal momentum flux is imparted perpendicular to diffuser
line), staged diffuser (where net horizontal momentum flux is imparted parallel to
the diffuser line), and alternating diffuser (where no net horizontal momentum flux
is imparted) [20]. Figure 7.6 illustrates one typical example of a unidirectional
diffuser. It is expected that different types of diffusers have significantly different
jet mixing and spreading properties.
Studies on multiple jets (or diffusers) have been reported extensively in the past
decades ([2, 21, 32, 34, 35, 39, 40, 45, 46, 94, 96, 100, 107], etc.). Most of these
studies focused on the deep water ambient condition (e.g., in oceans and lakes), and
limited studies dealt with the shallow water condition (e.g., in rivers). In this section,
only the deep water condition will be considered. Theoretically, multiple jets in
sequence experience: the individual free jet zone (where jets has no effect on each
other), the jet merging zone (where the interaction between jets are strong), and the
two-dimenstional zone thereafter [66, 107] (see Fig. 7.7). In practice, multiple jets
are usually simplified as one line momentum source, neglecting the interactions
between individual jets that are complex and not well understood [35, 46].
In the previous sections, some basic characteristics of three-dimensional free
jets and two-dimentional (plane) jets have been introduced; hence, in the following,
the studies on the merging process of unidirectional non-buoyant circular jets will
be briefly summarized. As yet, only limited studies on jet merging have been
reported [30, 40, 64, 71, 100]. To calculate the concentration or velocity field in the
jet merging process, the most widely used method is superposition. However, as the
momentum equation is nonlinear, simple superposition of individual jets would
overestimate the jet velocity. Knystautas [40] studied the velocity field of merging
jets in still ambient fluid and showed that the jet velocity can be modeled by
superposing the momentum (u2) of individual jets (based on Reichardt’s hypothesis).
Fig. 7.7 (a) Schematic

of multiple merging jets, dx
with the indication of (b) jet y n=1
dp z
cross-section deformation
(Modified from [100]) x U
n=0 0
n = –1
L/2
L/2
One jet
a
bvertical
y
z bhorizontal
b
Hodgson et al. [30] reported the following equation for the velocity field of the
merging jets in still water:
" #
X
i¼þn X
i¼þn
z iL 2
u ¼
2
ui ¼ um
2 2
exp 150 (7.36)
i¼n i¼n
x
where (2n + 1) is the number of jets; the centerline velocity of each individual jet
um ¼ 6:13U
x
0 d0
; L is the distance between the centers of neighboring jets; z is the
transverse distance from the central jet axis. Hodgson et al. [30] extended the
Reichardt’s hypothesis on lateral transport of momentum to the lateral transport
of pollutant, and showed that uC is additive. After solving the velocity field from
Eq. 7.36, the pollutant concentration field can be obtained from Eq. 7.37:
" #
iX
¼þn iX
¼þn
z iL 2
uC ¼ ui Ci ¼ um Cm exp 130 (7.37)
i¼n i¼n
x
where the centerline concentration of each individual jet Cm ¼ 5:34C

x
0 d0
. Hodgson’s
experimental results validated the use of Eqs. 7.36 and 7.37.
164 W. Zhang et al.
Hodgson et al. [30] revealed some basic physics in the jet merging, where the jet
spreading rate db
dx and the ratio of the jet concentration half-width to the velocity
b
half-width bC are both assumed to be constant during merging. Wang and Davidson
[100] developed a similar model for jet merging in stagnant ambient fluid, but
bC
allows the change of db dx and b during merging, as well as the change of these
parameters in horizontal (the jet merging) plane and vertical (the free entrainment)
plane. Theoretical analysis and experimental data indicate that the jet merging in
stagnant ambient fluid occurs at 4:5 < Lx < 12. Note that in Wang and Davidson
[100], the start of merging refers to the location where the jet interaction begins to
influence the bulk properties of the central jet, which is beyond the location where
the physical jet boundaries start intersecting. By studying their experimental data
and that of Knystautas [40], Wang and Davidson [100] found that, during the jet
dbC
merging, the jet spreading rate db dx (or dx ) increases by 30% in the vertical plane,
while it decreases by a similar amount in the horizontal plane; the ratio of velocity
b
half-width in the vertical plane to that in the horizontal plane b vertical increases from
horizontal
b
1 to 1.5 during merging; and the ratio of concentration half-widths b C;vertical
C;horizontal
increases from 1 to 1.8. Obviously, the jet merging process constrains the jet
spreading and thus dilution in the jet merging plane, while accelerates them in
the free entrainment plane. This phenomenon is similar to the boundary effects
found in wall jets or surface jets.
In recent years, researchers started to study the jet merging in coflow [64, 71].
Pun et al. [71] developed a multiple-point hybrid model for merging jets in coflow,
which combines a length-scale model and an Eulerian-integral model. The model
allows multiple transition points for each parameter (jet velocity, spread and dilution),
instead of a single transition point for all these parameters. The multiple-point hybrid
model is shown to be able to significantly reduce transition errors during merging
compared to the single-point model, and predicts favorable results compared to the
integral solution. Pani et al. [64] developed a model based on Reichardt’s hypothesis
for multiple coflowing jets. Instead of momentum (u2) is additive in stagnant water,
Pani et al. showed that the excess momentum (uDu) is additive in coflow and follows
Gaussian distribution. Using the method of superposition and a generalized spreading
hypothesis, Pani et al. presented the equations for predicting the velocity field and
centerline dilution downstream of multiple circular jets in coflow, which appeared to
agree with the experimental data.
Multiphase Jets and Plumes
In this section, two kind of multiphase jets and plumes will be introduced: bubbly
jets and plumes and slurry jets and plumes, which have wide engineering
applications.
Bubbly Jets and Plumes
Bubbly jets are produced by injecting gas-liquid mixtures into liquids, while
bubble plumes are produced by injecting gases into liquids. Bubble plumes
and bubbly jets are widely used to achieve artificial aeration, circulation and
mixing in confined reactors, aeration tanks, polluted water bodies, ice-covered
rivers, and deep stratified lakes and reservoirs [48, 86, 101, 104]. Such kind of
gas-liquid two-phase flow is also common in some hydraulic structures, e.g.,
the super-gas saturation downstream of hydro-power dams. So far, most of the
early studies were conducted in confined setups, where the sizes and geometry
of the setup further complicates the characteristics of bubbly jets and plumes
[51]. In this section, the studies in stagnant water of relatively large setup will
be introduced and then the case with flowing ambient fluid will be considered.
For two-phase flows, the dissolving of the gas phase into the liquid phase can be
derived from Fick’s law of diffusion as [60]:
dC
¼ KL aðCs CÞ (7.38)
dt
where C is the dissolved gas (e.g., oxygen) concentration in the liquid, CS is the
saturation dissolved gas concentration, t is the time, KL is the mass transfer
coefficient, and a is the gas-liquid interfacial area per unit liquid volume
(also named the specific interfacial area). From Eq. 7.38, the gas transfer rate
is mainly controlled by KL and a, which differ significantly in different setups.
Previous studies [9, 59] have shown that these two parameters are greatly
influenced by bubble size. Bubble size depends on a number of factors: nozzle
sizes and types, initial gas volume fractions, the solubility and mass transfer
ability of the gas, turbulence intensity and flow structure of the ambient liquid,
impurities and surfactants in the ambient liquid, etc. [17, 49–52]. Lima Neto
et al. [50] proposed a criterion to judge the sizes and shapes of the bubbles
produced by injecting a mixture of air and water into water: if the nozzle
Reynolds number Re ¼ Uw0 vw < 8000 (where U w0 is the superficial water
d0
velocity based on the water discharge at the nozzle and nozzle diameter d0),
then large and irregular bubbles will be produced; if Re 8,000, smaller and
uniform bubbles will be produced. A decrease in gas discharge or an
increase in liquid discharge will decrease the bubble size [50, 52, 98].
Now consider the vertical injection of a pure gas into a pure stagnant liquid. The
bubbles produced at the orifice will coalescence/breakup and rise, inducing ambient
liquid entrained into the bubble core and the dissolving of the gas into the ambient.
Lima Neto et al. [49] studied air injection into still water with six different nozzles
(single orifice, multiple orifices, and airstone), and found that the water entrained into
the bubble core under different initial air discharges Qa and nozzle types can be
described as a function of Qa and vertical distance from the nozzles.
166 W. Zhang et al.
For a vertical bubbly jet in stagnant liquid, Milgram [56], Brevik and Kristainsen
[12], and Lima Neto et al. [50] reported that the bubble area typically only occupies
50–90% of the bubbly jet in the radial direction. Lima Neto et al. [50] studied bubbly
jets produced by injecting a mixture of air-water into stagnant water and found that
the more uniform and smaller the bubble sizes, the wider the bubble core can spread
in the radial direction; within the bubbly jets, the radial distributions of the time-
averaged bubble concentration (void fraction) and water velocity of the mean flow
can be well described by Gaussian distributions, similarly as for single-phase jets or
plumes; and db/dx of the bubbly jets is close to that of the pure water jet.
Although the existence of bubbles seems not to change the Gaussian profiles, the
entrainment of the ambient into the bubbly jets is significantly enhanced. Milgram
[56], Socolofsky and Adams [92], Brevik and Kristainsen [12], and Lima Neto et al.
[50] reported the entrainment coefficient of bubbly jets is in the range of 0.03–0.15,
much larger than the values of pure jets or plumes. The additional entrainment
probably associates with the bubble wakes [47] and additional liquid turbulence
caused by interactions of the bubbles and their wakes [50]. At a specific height
of the centerline of a bubbly jet, Lima Neto et al. [50] compared the liquid volume
flux Qw of the bubbly jet with that of a pure water jet with the same nozzle diameter
and water flow rate:
32
Qw C0
¼ 1 þ 6:426
10 6
(7.39)
QwðC0 ¼0Þ Re
where the initial gas volume fraction C0 ¼ ðQa0QþQ a0

w0 Þ
; Qa0 and Qw0 are the initial
volumetric flow rates of air and water, respectively. It is expected that the average
dilution of a bubbly jet is larger than that of a pure jet due to the additional entrainment
of ambient water as shown in the second term of the the right hand side of Eq. 7.39.
For a bubbly jet injected horizontally into a stagnant liquid, there are only
limited experimental studies [22, 52, 57, 98]. Lima Neto et al. [52] studied the
injection of air-water mixtures into a water tank of relatively large size and reported
that: first, the bubbly jet comes out of the nozzle as a whole quasi-horizontal bubbly
jet where bubble breakup/coalescence occurs and only a few bubbles escape
from the bubbly jet and rise vertically due to buoyancy; then, there follows
a separation zone where the quasi-vertical bubble plume partially separates from
the water jet (when the initial gas volume fraction C0 < 0.15, the bubble plume will
completely separate from the water jet); finally, the bubble plume continues
rising and the water jet impinges the water surface and becomes surface jet. In
Lima Neto et al. [52], the length and width of the bubble plume as well as the
centerline trajectories of the bubble plume and water jet were also proposed in
dimensionless forms.
For bubbly jets, bubble properties and mean liquid flow could be non-
dimensionalized as functions of the initial gas volume fraction and nozzle Reynolds
number for the vertical injection or as functions of the initial gas volume fraction
and nozzle densimetric Froude number for the horizontal injection [50, 52]. The
variation of bubble properties and mean liquid flow along the jet centerline and
across the jet needs further studies. For bubble plume modeling, the reader may
refer to Bravo et al. [11] and others.
Crossflow will exert significant effects on bubbly jets or bubble plumes, of
which the most distinguishing one is the possibility of separation of bubble plumes
(named generally as dispersed phases) from the entrained ambient fluids (continu-
ous phases) [92]. So far, very limited studies have been reported on bubbly jets or
bubble plumes in crossflow. Socolofsky and Adams [92] conducted laboratory
experiments on bubble plumes produced by injecting air, air and oil, as well as
air and alcohol in uniform crossflow. In weak crossflow, the separation between the
bubble plumes and the entrained fluid does not occur before the plumes reach the
surface. While in strong crossflow, the separation is significant and the separation
height hs can be given by an empirical relation:
5:1B0
hs ¼ (7.40)
ðUa us 2:4 Þ0:88
where Ua is the horizontal crossflow velocity; us is the bubble slip (terminal rise)
ðr r Þg
velocity; B0 = Q0g0 ; Q0 is the discharge of the plume fluid; g0 ¼ a r p ; ra is the
a
ambient density of water; and rp is the density of the plume fluid. Socolofsky and
Adams also reported that before hs, the plumes can be treated as single-phase
plumes; after hs, the bubble plumes follow the trajectory of the vector sum of us
and Ua (i.e., the bubble plumes rise in a linear line), and the separated entrained
fluid behaves like a momentum jet (the momentum is gained by the acceleration of
bubbles before the separation).
Slurry Jets and Plumes
Slurry jets are produced by injecting a mixture of liquid phase and solid phase (such
as sand or clay particles). Slurry jets have wide applications in pumping industrial
(e.g., mining or petroleum) tailings into settling tanks, dredging and land reclama-
tion, discharging storm water and industrial waters that have solid particles,
etc. A number of experimental and numerical studies in this area have been reported
[7, 13, 26, 33, 55, 67, 90]. Usually, two injection ways were used, vertically upward
and vertically downward. In the following, the main focus will be on the vertically
downward injection of slurry jets into stagnant ambient fluids. Compared to single-
phase jets, the adding of the solid phase will change the properties of the flow [89].
Previous experiments have indicated that the velocity and concentration of the
solid phase across slurry jets follow self-similar Gaussian distributions [26, 33, 90].
In the slurry jets with dilute solid particles, Jiang et al. [33] reported that the
velocity and concentration of the liquid phase also exhibit self-similar Gaussian
168 W. Zhang et al.
profiles. However, this may not be valid for the slurry jets with high concentration
solid particles.
The spreading of the solid phase has been found to increase linearly along the
axial direction [13, 26, 55]. Brush [13] reported that the spreading rate of the
velocity of the solid phase db dx depended on the particle size. Mazurek et al. [55]
s
confirmed this by photographic measurements on sand jets, and further generalized

the spreading as a function of the initial densimetric Froude number of the solid
particle. Recently, Hall et al. [26] conducted detailed experiments on pure sand jets
and sand-water slurry jets in stagnant water. With the densimetric Froude number at
the nozzle exit (F0 ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffi
U0
where U0 is the initial velocity of the sand particle
gd0 ðrs rw Þ
rw
from a nozzle of diameter d0 and rs is sand density) in the range of 2 6, db dx

s
measured 0.087 0.109, not very different from that of single-phase jets in
stagnant fluids. The difference lies in the ratio of concentration to velocity spread-
b
ing rates of the solid phase ( bs;Cs ). Hall et al. [26] reported that, for sand jets
b
and slurry jets with high F0 (F0 6), bs;Cs ¼ 0:86 0.92, which means the
sand concentration spreads slower than the velocity; and for slurry jets with low
b
F0 (F0 2), bs;Cs 1:0 which means they have almost equal spreading rates. This
finding is contrary to the classic single-phase jet theory, which states that the
concentration scale spreads faster than the velocity scale ( bbC 1:2).
Similar as single-phase jets, along the axial direction of slurry jets, the velocity
and concentration of both solid phase and liquid phase decay rapidly. According to
2
Hall et al. [26], beyond the potential core (about 2:9 d0 F0 5), the axial concentration
of the solid phase can be well described by:
Cm 17:12
¼
C0 53 (7.41)
x
2 þ 11:39
d0 F0 5
In Eq. 7.41, the 5/3 power relation is very similar to that of single-phase plume
(Table 7.2), as Eq. 7.41 was established in the region at x > 5LM where the buoyant
slurry jet behaved like a slurry plume [65]. Similar 5/3 power relation was also
built for the sand concentration in sand jets. For both sand jets and slurry jets, the
axial velocity of the solid phase was found to decrease rapidly and then reach a final
plateau region. Generally, before the plateau region, the axial velocity of the solid
phase in slurry jets can be well represented by:
um F 0 1:63
¼ 13 (7.42)
U0 x
d0 F0 þ 0:56
Similar as Eq. 7.42, for sand jets, before reaching the velocity plateau region, the
axial sand velocity was also found to follow the 1/3 power relation, which is very
similar to that of single-phase plumes (Table 7.2). The terminal (settling) centerline
velocity of the solid phase was found in the range of 0.32 0.43 m/s, which is
larger than the settling velocity of 0.033 m/s for individual solid particles. The
larger terminal velocity probably can be attributed to the interactions between solid
particles, i.e., the wake of previous solid particles tends to decrease the drag forces
for the following particles.
For the studies dealing with particle interactions, the reader may refer to Lain and
Garcia [42], Tamburello and Amitay [93], and Yan et al. [106]. For the effect of solid
particle size on velocity distribution, concentration profile and turbulent properties,
the reader may refer to Azimi et al. [8]. For vertically upward slurry jets, the reader
may refer to Jiang et al. [33].
Future Directions
Some of the basic characteristics of a variety of jets and plumes have been
reviewed, including simple jets and plumes, buoyant jets, surface jets, wall jets,
jets and plumes in coflow and crossflow, multiple jets, bubbly jets and plumes, and
slurry jets and plumes. The turbulence and turbulence structures in these flows have
not been discussed. Interested readers may refer to the works of Heskestad [27],
Wygnanski and Fielder [105], and Launder and Rodi [44]. Also, the behaviors of
jets and plumes in stratified environment have not been considered. The readers
may refer to the works of Morton et al. [58], Turner [97], and Roberts et al. [80, 81].
To facilitate the applications of jets and plumes theories, software packages have
been developed. The USEPA-supported CORMIX is perhaps the most commonly
used expert system for dealing with environmental problems involving jets and
plumes. Other models are VISJET and Visual Plume. In addition to the simple jets
and plumes, all the other varieties of jets and plumes are still currently under active
studies. This constitutes the general tone for the future research. Herein, some of
key areas are listed as follows:
• Physics and models of bubbly jets, and their application in aeration and mixing
of ponds, lakes, and wastewater treatment plants
• Physics and models of slurry jets, and their industrial applications
• Physics and models of three phase jets and plumes, especially oil-water-gas
plumes produced by oil spills in oceans
• Physics and models of jets and plumes in stratified environment
• Development, improvement, and validation of computational fluid dynamics
(CFD) modeling (such as direct numerical modeling (DNS); large eddy simula-
tion (LES); k-epsilon modeling; and others).
170 W. Zhang et al.
Bibliography
Primary Literature
1. Abramovich GN (1963) The theory of turbulent jets. MIT Press, Cambridge, MA, 671 p
2. Adams EE (1982) Dilution analysis for unidirectional diffusers. ASCE J Hydraul Div
108(HY3):327–342
3. Agelin-Chaab M (2010) Experimental study of three-dimensional offset jets and wall jets.
Ph.D. thesis, Department of Mechanical and Manufacturing Engineering, University of
Manitoba, Winnipeg
4. Albertson ML, Dai YB, Jensen RA, Rouse H (1950) Diffusion of submerged jets. Trans
ASCE 115:639–697
5. Andreopoulos J (1985) On the structure of jets in a crossflow. J Fluid Mech 157:163–197
6. Anthony DG, Willmarth WW (1992) Turbulence measurements in a round jet beneath a free
surface. J Fluid Mech 243:699–720
7. Awaya Y, Fujisaki K, Matsunaga K (1985) Transition in the behavior of sediment laden
vertical buoyant jet. J Hydrosci Hydr Eng 3(1):63–74
8. Azimi AH, Zhu DZ, Rajaratnam N (2010) Effect of particle size on the characteristics of sand
jets in water. J Eng Mech (in press)
9. Barnhart EL (1969) Transfer of oxygen in aqueous solutions. J Sanit Eng Div 95(3):645–661
10. Beltaos S, Rajaratnam N (1972) Plane turbulent impinging jets. J Hydr Res 11:29–59
11. Bravo HR, Gulliver JS, Hondzo M (2007) Development of a commercial code-based
two-fluid model for bubble plumes. Environ Modell Softw 22(4):536–547
12. Brevik I, Kristiansen Ø (2002) The flow in and around air bubble plumes. Int J Multiph Flow
28:617–634
13. Brush LMJ (1962) Exploratory study of sediment diffusion. J Geophys Res 67(4):1427–1433
14. Chan T, Zhou Y, Liu M, Leung C (2003) Mean flow and turbulent measurements of the
impingement wall jet on a semi-circular convex surface. Exp Fluids 34(1):140–149
15. Chen CJ, Rodi W (1980) Vertical turbulent buoyant jets – a review of experimental data.
Pergamon, Oxford. 83 p
16. Chu VH, Vanvari MR (1976) Experimental study of turbulent stratified shearing flow.
J Hydraul Div ASCE 102(6):691–706
17. Clift R, Grace JR, Weber ME (1978) Bubbles, drops, and particles. Academic, New York, 380 p
18. Cuthbertson AJS, Davies PA (2008) Deposition from particle-laden, round, turbulent,
horizontal, buoyant jets in stationary and coflowing receiving fluids. J Hydraul Eng
134(4):390–402
19. Dey S, Nath TK, Bose SK (2010) Fully rough submerged plane wall-jets. J Hydro Environ
Res 4(4):301–316
20. Doneker RL, Jirka GH (2007) CORMIX user manual – a hydrodynamic mixing zone model
and decision support system for pollutant discharges into surface waters. U.S. EPA-823-K-
07-001, pp 1–236
21. Fischer HB, List EJ, Imberger J, Brooks NH (1979) Mixing in inland and coastal waters.
Academic, New York, 483 p
22. Fonade C, Doubrovine N, Maranges C, Morchain J (2001) Influence of a transverse flowrate
on the oxygen transfer performance in heterogeneous aeration: case of hydro-ejectors. Water
Res 35(14):3429–3435
23. Förthmann E (1936) Turbulent jet expansion. English translation N.A.C.A. TM-789.
(Original paper in German, 1934. Ing. Archiv., 5)
24. Gholamreza-kashi S, Martinuzzi RJ, Baddour RE (2007) Mean flow field of a nonbuoyant
rectangular surface jet. J Hydraul Eng 133(2):234–239
25. Girshovich TA (1966) Theoretical and experimental study of a plane turbulent jet in
a crossflow. Fluid Dyn 9:84–86
26. Hall N, Elenany M, Zhu DZ, Rajaratnam N (2010) Experimental study of sand and slurry jets
in water. J Hydraul Eng 136(10):727–738
27. Heskestad G (1965) Hot-wire measurements in a plane turbulent jet. Trans ASME J Appl
Mech 32:1–14
28. Hinze JO, Van Der Hegge Zijnen BG (1949) Transfer of heat and matter in the turbulent
mixing zone of an axially symmetrical jet. Appl Sci Res A1:435–461
29. Hodgson JE, Rajaratnam N (1992) An experimental study of jet dilution in crossflows.
Can J Civ Eng 19:733–743
30. Hodgson JE, Moawad AK, Rajaratnam N (1999) Concentration field of multiple circular
turbulent jets. J Hydr Res 37(2):249–256
31. Huang JF, Davidson MJ, Nokes RI (2005) Two-dimensional and line jets in a weak cross-
flow. J Hydr Res 43(4):390–398
32. Huang H, Fergen RE, Rroni JR, Tsai JJ (1996) Probabilistic analysis of ocean outfall mixing
zones. J Environ Eng 122(5):359–367
33. Jiang JS, Law AWK, Cheng NS (2005) Two-phase analysis of vertical sediment-laden jets.
J Eng Mech 131(3):308–318
34. Jirka GH, Akar PJ (1991) Hydrodynamic classification of submerged multiport-diffuser
discharges. J Hydr Eng 117(9):1113–1128
35. Jirka GH, Harleman DRF (1973) The mechanics of submerged multiport diffuser for buoyant
discharges in shallow water. R.M. Parsons Laboratory Report, Massachusetts Institute of
Technology, Boston
36. Jones WP, Wille M (1996) Large-eddy simulation of a plane jet in a cross-flow. Int J Heat
Fluid Flow 17:296–306
37. Kalita K, Dewan A, Dass AK (2002) Predication of turbulent plane jet in crossflow.
Numer Heat Transfer A 41:101–111
38. Kikkert GA, Davidson MJ, Nokes RI (2009) A jet at an oblique angle to a cross-flow.
J Hydro Env Res 3(2):69–76
39. Kim DG, Seo IW (2000) Modeling the mixing of heated water discharged from a submerged
multiport diffuser. J Hydraul Res 38(4):259–269
40. Knystautas R (1964) The turbulent jet from a series of holes in Line. Aeronaut Q 15:1–28
41. Kotsovinos NE, List EJ (1977) Plane turbulent buoyant jet. Part I. Integral properties. J Fluid
Mech 81:25–44
42. Lain S, Garcia J (2006) Study of four-way coupling on turbulent particle-laden jet flows.
Chem Eng Sci 61:6775–6785
43. Lam KM, Lee WY, Chan CHC, Lee JHW (2006) Global behaviors of a round buoyant jet in
a counterflow. J Hydr Eng 132(6):589–604
44. Launder BE, Rodi W (1983) The turbulent wall jet - measurement and modeling. Ann Rev
Fluid Mech 15:429–459
45. Lee JHW, Chu VH (2003) Turbulent jets and plumes – a lagrangian approach. Kluwer,
Norwell, 390 p
46. Lee JHW, Jirka GH (1980) Multiport diffuser as line source of momentum in shallow water.
Water Resour Res 16(4):695–708
47. Leitch AM, Baines WD (1989) Liquid volume flux in a weak bubble plume. J Fluid Mech
205:77–98
48. Lima Neto IE, Zhu DZ, Rajaratnam N, Yu T, Spafford M, McEachern P (2007) Dissolved
oxygen downstream of an effluent outfall in an ice-cover river: natural and artificial aeration.
J Env Eng 133(11):1051–1060
49. Lima Neto IE, Zhu DZ, Rajaratnam N (2008) Air injection in water with different nozzles.
J Environ Eng 134(4):283–294
50. Lima Neto IE, Zhu DZ, Rajaratnam N (2008) Bubbly jets in stagnant water. Int J Multiph
Flow 34(12):1130–1141
51. Lima Neto IE, Zhu DZ, Rajaratnam N (2008) Effect of tank size and geometry on the flow
induced by circular bubble plumes and water jets. J Hydr Eng 134(6):833–842
172 W. Zhang et al.
52. Lima Neto IE, Zhu DZ, Rajaratnam N (2008) Horizontal injection of gas-liquid mixtures in
a water tank. J Hydr Eng 134(12):1722–1731
53. Lübcke HM, Rung Th, Thiele F (2003) Predication of the spreading mechanism of 3D
turbulent wall jets with explicit Reynolds-stress closures. Int J Heat Fluid Flow 24:
434–443
54. Margason RJ (1993) Fifty years of jet in cross flow research. Computational and experimental
assessment of jets in cross flow. AGARD, Report CP-534
55. Mazurek KA, Christison K, Rajaratnam N (2002) Turbulent sand jet in water. J Hydr Res
40(4):527–530
56. Milgram H (1983) Mean flow in round bubble plumes. J Fluid Mech 133:345–376
57. Morchain J, Moranges C, Fonade C (2000) CFD modeling of a two-phase jet aerator under
influence of a crossflow. Water Res 34(13):3460–3472
58. Morton BR, Taylor GI, Turner JS (1956) Turbulent gravitational convection from maintained
and instantaneous sources. Proc R Soc Lond A 234:1–23
59. Motarjemi M, Jameson GJ (1978) Mass transfer from very small bubbles – the optimum
bubble size for aeration. Chem Eng Sci 33(11):1415–1423
60. Mueller JA, Boyle WC, Pöpel HJ (2002) Aeration: principles and practice. CRC Press,
New York
61. Newman BG, Patel RP, Savage SB, Tjio HK (1972) Three-dimensional wall jet originating
from a circular orifice. Aeronaut Quart 23:188–200
62. Pande BBL, Rajaratnam N (1979) Turbulent jets in co-flowing streams. ASCE J Eng Mech
105:1025–1038
63. Pande BBL, Rajaratnam N (1977) An experimental study of bluff buoyant turbulent surface
jets. J Hydr Res 15(3):261–275
64. Pani BS, Lee JHW, Lai ACH (2009) Application of Reichardt’s hypothesis for multiple
coflowing jets. J Hydro-environ Res 3(3):121–128
65. Papanicolaou PN, List EJ (1988) Investigations of round vertical turbulent buoyant jets.
J Fluid Mech 195:341–391
66. Parr AD, Sayre WW (1979) Multiple jets in shallow flowing receiving waters. ASCE
J Hydraul Div 105(HY11):1357–1374
67. Parthasarathy RN, Faeth GM (1987) Structure of particle laden turbulent water jets in still
water. Int J Multiph Flow 13(5):699–716
68. Patel RP (1971) Turbulent jets and wall jets in uniform streaming flow. Aeronaut Q
22:311–326
69. Platten JL, Keffer JF (1971) Deflected Turbulent Jet Flows. Trans ASME J Appl Mech 38E
(4):756–758
70. Pratte BD, Baines WD (1967) Profiles of the round turbulent jets in a cross flow. J Hydr Div
ASCE 92:53–64
71. Pun KL, Davidson MJ, Wang HJ (2000) Merging jets in a co-flowing ambient fluid – a hybrid
approach. J Hydraul Res 38(2):105–114
72. Rajaratnam N (1967) Plane turbulent wall jets on rough boundaries. Water Power
49:149–242
73. Rajaratnam N (1976) Turbulent jets. Elsevier Scientific, Amsterdam, 304 p
74. Rajaratnam N (1984) Turbulent jets and plumes, Lecture Notes, University of Alberta,
Edmonton, Canada
75. Rajaratnam N (1985) An experimental study of bluff surface discharges with small
Richardson number. J Hydraul Res 23(1):47–55
76. Rajaratnam N, Humphries JA (1984) Turbulent non-buoyant surface jets. J Hydraul Res
22(2):103–105
77. Rajaratnam N, Pani BS (1974) Three-dimensional turbulent wall jets. Proc ASCE J Hydraul
Div 100:69–83
78. Rajaratnam N, Subramanya K (1967) Diffusion of rectangular wall jets in wide channels.
J Hydraul Res 5:281–294
79. Rajaratnam N, Subramanyan S (1985) Plane turbulent buoyant surface jets and jumps.
J Hydraul Res 23(2):131–146
80. Roberts PJW, Snyder WH, Baumgartner DJ (1989) Ocean outfalls I: Submerged waste field
formation. J Hydraul Eng 115(HY1):1–25
81. Roberts PJW, Snyder WH, Baumgartner DJ (1989) Ocean outfalls II: Spatial evolution of
submerged waste field. J Hydraul Eng 115(HY1):26–48
82. Rodi W, Spalding DB (1970) A two-parameter model of turbulence and its application to free
jets. Wärme Stoffübertragung 3:85–95
83. Rostamy N, Bergstrom D, Sumner D (2010) An experimental study of a turbulent wall jet
on smooth and rough surfaces. In: Nickels TB (ed) IUTAM symposium on the physics of
wall-bounded turbulent flows on rough walls, Cambridge, UK
84. Rouse H, Yih CS, Humphreys HW (1952) Gravitational convection from a boundary source.
Tellus 4:201–210
85. Rutherford JC (1994) River mixing. Wiley, West Sussex, 347 p
86. Schierholz EL, Gulliver JS, Wilhelms SC, Henneman HE (2006) Gas transfer from air
diffusers. Water Res 40(5):1018–1026
87. Schlichting H, Gersten K (2000) Boundary layer theory. Springer, Berlin Heidelberg, 801 p
88. Sforza PM, Herbst G (1970) A study of three-dimensional incompressible turbulent wall jets.
Am Inst Aeronaut Astron J 8:276–283
89. Sheen HJ, Jou BH, Lee YT (1994) Effect of particle size on a two-phase turbulent
jet. Exp Therm Fluid Sci 8:315–327
90. Singansetti SR (1966) Diffusion of sediment in submerged jet. J Hydraul Div 92(2):153–168
91. Smith SH, Mungal MG (1998) Mixing, structure and scaling of the jet in crossflow. J Fluid
Mech 357:83–122
92. Socolofsky SA, Adams EE (2002) Multi-phase plumes in uniform and stratified crossflow.
J Hydraul Res 40(6):661–672
93. Tamburello DA, Amitay M (2008) Active manipulation of a particle laden jet. Int J Multiph
Flow 34:829–851
94. Tang HS, Paik J, Sotiropoulos F, Khangaonkar T (2008) Three-dimensional numerical
modeling of initial mixing of thermal discharges at real-life configurations. J Hydraul Eng
134(9):1210–1224
95. Tachie MF, Balachandar R, Bergstrom DJ (2004) Roughness effects on turbulent plane wall
jets in an open channel. Exp Fluid 37(2):281–292
96. Tian X, Roberts PJW, Daviero GJ (2004) Marine wastewater discharges from multiport
diffuser. II: unstratified flowing water. J Hydraul Eng 130(12):1147–1155
97. Turner JS (1973) Buoyancy effects in fluids. Cambridge University Press, Cambridge, 367 p
98. Varley J (1995) Submerged gas-liquid jets: bubble size prediction. Chem Eng Sci
50(5):901–905
99. Verhoff A (1963) The two-dimensional turbulent wall jet with and without an external
stream. Report 626, Princeton University
100. Wang HJ, Davidson MJ (2003) Jet interaction in a still ambient fluid. J Hydraul Eng
129(5):349–357
101. Whipple WJ, Yu SL (1970) Instream aerators for polluted rivers. J Sanit Eng Div
96(5):1153–1165
102. Wright SJ (1977) Mean behavior of buoyant jet in a crossflow. J Hydraul Div ASCE
103(HY5):499–513
103. Wright SJ (1984) Buoyant jets in density-stratified crossflow. J Hydraul Eng ASCE
110(5):643–656
104. Wüest A, Brooks NH, Imboden DM (1992) Bubble plume modeling for lake restoration.
Water Resour Res 28(12):3235–3250
105. Wygnanski I, Fielder H (1969) Some measurements in the self-preserving jet. J Fluid Mech
38:577–612
174 W. Zhang et al.
106. Yan J, Luo K, Fan J, Tsuji Y, Cen K (2008) Direct numerical simulation of particle
dispersion in a turbulent jet considering inter-particle collisions. Int J Multiph Flow 34:
723–733
107. Zhang W, Zhu DZ (2011) Near-field mixing downstream of a multiport diffuser in a shallow
river. J Environ Eng 137(4): 230–240
Books and Reviews
Cheremisinoff NP (ed) (1986) Encyclopedia of fluid mechanics, vol 2, Dynamics of single-fluid

flow and mixing. Gulf, Houston, 1505 p
Cheremisinoff NP (ed) (1986) Encyclopedia of fluid mechanics, vol 3, Gas-liquid flow. Gulf,
Houston, 1535 p
Cheremisinoff NP (ed) (1993) Encyclopedia of fluid mechanics, vol 2, Advances in
multiphase flow. Gulf, Houston, 674 p
Davies PA, Valente Neves MJ (eds) (1994) Recent research advances in the fluid mechanics of
turbulent jets and plumes. Kluwer, Dordrecht, 514 p
Fay JA (1973) Buoyant plumes and wakes. Ann Rev Fluid Mech 5:151–160
Jirka GH (2004) Integral model for turbulent buoyant jets in unbounded stratified flow. Part 1:
single round jet. Environ Fluid Mech 4:1–56
Jirka GH (2006) Integral model for turbulent buoyant jets in unbounded stratified flow. Part 2:
plane jet dynamics resulting from multiport diffuser jets. Environ Fluid Mech 6:43–100
List EJ (1982) Turbulent jets and plumes. Ann Rev Fluid Mech 14:189–212
Philip AA (1997) Jets, plumes and mixing at the coast. In: Allen PA (ed) Earth surface processes.
Blackwell, Oxford, pp 241–266, Chapter 7
Rodi W (1982) Turbulent buoyant jets and plumes. Pergamon, Oxford, 184 p
Rubin H, Atkinson J (2001) Environmental fluid mechanics. Marcel Dekker, New York, 728 p
Shen HH, Cheng AHD, Wang KH, Teng MH, Liu CCK (eds) (2002) Environmental fluid
mechanics theories and applications. ASCE, Reston, 467 p
Wood AW (2010) Turbulent plumes in nature. Ann Rev Fluid Mech 42:391–412
Wood IR, Bell RG, Wilkinson DL (eds) (1993) Ocean disposal of wastewater. World Scientific,
Singapore, 425 p
Chapter 8
Atmosphere-Water Exchange
Bernd Jähne
Glossary
Bulk coefficients ci relate the transfer velocity k for a species i to

the wind velocity Ur in a reference height, typi-
cally at 10 m above the mean water level: ci = ki/
Ur . From the bulk coefficient, the flux density ji
of a species can be computed as ji = ci(Cr C0)
Ur , where Cr and C0 are the corresponding
concentrations at the reference height and right
at the water surface, respectively. For momen-
tum density (rU) the bulk coefficients is also
known as the drag coefficient cD. It can also be
expressed as cD = (u∗/Ur)2 with the momentum
flux given by jm ¼ ru2 ; u∗ is the friction
velocity.
Friction velocity u∗ is a measure for the tangential force per area
applied by the wind at the water surface, the shear
stress t ¼ ru2 , which is also equal to the vertical
momentum flux density jm.
Mass boundary layer Thickness of the layers at both sides of the water
surface in which transport of mass by turbulence
is smaller than by molecular diffusion.
B. Jähne (*)
Institute for Environmental Physics and Heidelberg Collaboratory
for Image Processing (HCI), University of Heidelberg, Speyerer Straße 6,
69115 Heidelberg, Germany
176 B. Jähne
Schmidt and Prandtl numbers, The Schmidt and Prandtl numbers are the ratio of
Sc and Pr kinematic viscosity n (molecular diffusion coeffi-
cient for momentum) and the molecular diffusion
coefficients for the corresponding chemical spe-
cies, D, and heat, Dh, respectively. Thus, these
numbers express how much slower chemical spe-
cies and heat, respectively, are transported by
molecular processes than momentum. In air,
these numbers are in the order of one; in water,
the Prandtl number is about 10 and the Schmidt
number about 1,000.
Transfer velocity k is the velocity by which a momentum, heat, and
chemical are transported across the surface; because
of the concentration discontinuity at the water sur-
face, the transfer velocity on the air side is different
from the transfer velocity on the water side.
Viscous boundary layer Thickness of the layers at both sides of the water
surface in which transport of momentum by tur-
bulent is smaller than by molecular friction,
resulting in a linear velocity profile in this layer.
Gaseous and volatile chemical species reside not only in the atmosphere. Because
they dissolve in water, they are also distributed in the hydrosphere. The by far
largest part of the hydrosphere is the ocean. Therefore, the exchange between
atmosphere and oceans is the most important process for the fate of gaseous and
volatile chemical species (Table 8.1).
For long-lived species, that is, life time by reaction larger than residence times
in either the atmosphere or the ocean, two basic factors determine the fate of
these species. For time scales longer than the turnover time of ocean mixing,
which is much slower than the same process in the atmosphere, the solubility of
the species determines where it ends up. On much shorter time scales, the speed by
which the species are exchanged between atmosphere and ocean and the concen-
tration difference determines the flux between these two compartments.
While turbulent mixing is quite fast within the ocean and the atmosphere, it
becomes more and more inefficient, the closer a species is transported to the
surface. The final step toward the interface must be overcome by molecular
diffusion, which is a slow process, especially in water. Therefore, very thin mass
boundary layers at both sides of the water surface, in which turbulent transport is
8 Atmosphere-Water Exchange 177
Table 8.1 Comparison of size (as depth in m over entire earth surface)
and mass of atmosphere and global water storage
System Depth (m) Mass (1018 kg)
Oceans 2,624 1,382
Ice caps and glaciers 47 25
Total ground water 46 24
Lakes 0.35 0.18
Rivers 0.004 0.002
Atmosphere 8,000 5.3
slower than molecular transport, are the bottleneck of the exchange between
atmosphere and ocean.
In this way, tiny layers only about 20–200 mm thick in the water and 200–2,000 mm
thick in the air control the exchange process. In consequence, the crucial question is
which processes determine the thickness of these layers.
Introduction
Research in air–water gas exchange dates back more than a century. Bohr [1], for
example, was one of the first to study the exchange of carbon dioxide in a stirred
bottle. With the early observation of increasing carbon dioxide concentration in the
atmosphere interest in air-sea gas exchange rose considerably, triggered by the
question how much of the excess carbon dioxide would go into the ocean [2, 3].
Since then significant and continuous research work was devoted for a better
understanding of air–water mass exchange, but the mechanisms of the exchange
processes are still not fully understood even today.
This entry is divided into three major parts and partly based on previous reviews of
the author on the same subject [4–6]. Section “Mechanisms of Atmosphere–Water
Exchange” describes the basic mechanisms of air–water exchange and the many
parameters influencing it. Focusing on field data, section “Experimental Techniques
and Results” summarizes the various experimental techniques and gives a synthesis of
the available experimental data. Using the information collected in the previous two
sections, section “Empiric Parametrization” tries to give the best possible empirical
parametrization of the exchange rate and section “Future Directions” points to future
directions of research.
Mechanisms of Atmosphere–Water Exchange
Turbulent and Molecular Transport
The transfer of gases and volatile chemical species between the atmosphere
and oceans is driven by a concentration difference and the transport by molecular
178 B. Jähne
and turbulent motion. Both types of transport processes can be characterized by

diffusion coefficients, denoted by D and Kc, respectively. The resulting flux area
density jc is proportional to the diffusion coefficient and the concentration gradient:
jc ¼ ðDþK c ðzÞÞrc : (8.1)
In a stationary homogeneous case and without sinks and sources by chemical

reactions, the flux density j is in vertical direction and constant. Then integration of
Eq. 8.1 yields vertical concentration profiles:
Zzr
1
Cðzr Þ Cð0Þ ¼ jc dz: (8.2)
D þ Kc ðzÞ
0
The molecular diffusion coefficient is proportional to the velocity of the

molecules and the free length between collisions. The same concept can be applied
to turbulent diffusion coefficients. Far away from the interface, the free length
(called mixing length) is set proportional to the distance from the interface and the
turbulent diffusion coefficient Kc for mass transfer is:
k
Kc ¼ u z; (8.3)
Sct
where k = 0.41 is the von Kármán constant, u*, the friction velocity, a measure for
the velocity fluctuations in a turbulent flow, and Sct = Km/Kc the turbulent Schmidt
number. Closer to the interface, the turbulent diffusion coefficients are decreasing
even faster. Once a critical length scale l is reached, the Reynolds number Re = u*l/
n (n is the kinematic viscosity, the molecular diffusion coefficient for momentum)
becomes small enough so that turbulent motion is attenuated by viscosity. The
degree of attenuation depends on the properties of the interface. At a smooth solid
wall, Kc / z3, at a free water interface it could be in the range between Kc / z3 and
Kc / z2 depending on surface conditions.
Viscous and Mass Boundary Layers
Boundary layers are formed on both sides of the interface (Fig. 8.1). When the
turbulent diffusivity becomes equal to the kinematic viscosity, the edge of
the viscous boundary layer is reached. As the name implies, this layer is dominated
by viscous dissipation and the velocity profile becomes linear because of a constant
diffusivity. The edge of the mass boundary layer is reached when the turbulent
diffusivity becomes equal to the molecular diffusivity. The relative thickness of both
boundary layers depends on the dimensionless ratio Sc = n/D (Schmidt number).
The viscous and mass boundary layers are of about the same thickness in the air,
because values of D for various gaseous species and momentum are about the same
(Scair is 0.56 for H2O, 0.63 for heat, and 0.83 for CO2). In the liquid phase,
the situation is completely different. With Schmidt numbers in the range from
100 to 3,000 (Fig. 8.2, Table 8.3), molecular diffusion for a dissolved volatile
chemical species is two to three orders of magnitude slower than diffusion of
momentum (Table 8.2). Thus, the mass boundary layer is significantly thinner
than the viscous boundary layer in the liquid phase. This means that the transfer
of gases is much slower and almost entirely controlled by the tiny residual turbu-
lence in the small top fraction of the viscous boundary layer. This basic fact makes
it difficult to investigate the mechanism of air–water gas transfer both theoretically
and experimentally. In addition, the transfer process depends strongly on the water
temperature because the Schmidt number decreases by about a factor of 6 from 0 C
to 35 C (Fig. 8.2, Table 8.3).
Description of Transport
The amount of species exchanged between the air and water across the interface
can be described by a quantity with the units of a velocity. It represents the velocity
with which a tracer is pushed by an imaginary piston across the surface. This
quantity is known as the transfer velocity k (also known as the piston velocity,
gas exchange rate or transfer coefficient). It is defined as the flux density divided by
the concentration difference between the surface and the bulk at some reference
level zr:
jc
k¼ : (8.4)
Cs Cb
The inverse of the transfer velocity is known as the transfer resistance R:
Cs Cb
R¼ : (8.5)
jc
The indices s and b denote the surface and bulk. Both quantities can directly
be related to vertical concentration profiles by introducing Eqs. 8.4 and 8.5 into
Eq. 8.2:
Zzr
1 1
R¼ ¼ dz: (8.6)
k D þ Ke ðzÞ
0
180 B. Jähne
Fig. 8.1 Schematic graph of

the mass boundary layers at
a gas–liquid interface for
a tracer with a solubility a = 3
Cas Cws = α Cas

Kc
Kc
zr
Cwb
Fig. 8.2 Schmidt number/

solubility diagram including
various volatile tracers, 105 H2O SO2
momentum, and heat for
a temperature range (C) as
indicated. Filled circles refer 104
to only a temperature of 20 C.
The regions for air-side,
mixed, and water-side control 103
of the transfer process between
gas and liquid phase are
marked. At the solid lines, the 102
solubility
transfer resistance is equal in

both phases. The following
dimensional transfer 10
resistances were used: ra = 31,
rw = 12Sc2/3 (smooth), rw =
6.5Sc1/2 (wavy surface) with ra 1 CO2
= Rau*a and rw = Rwu*w N2O
(Adapted from Jähne and
Haußecker [6]) 10–1 C6F6
O2 CH4
H2
10–2
SF6
10–3
1 10 102 103 104
Schmidt number Sc=n/D
Table 8.3 Schmidt numbers of various gases and volatile species in the temperature range from
0 C to 40 C
Schmidt number
Species 0 C 5 C 10 C 15 C 20 C 25 C 30 C 35 C 40 C
Heat 13.45 11.19 9.46 8.10 7.02 6.14 5.42 4.82 4.32
3
He 329 254 200 160 130 107 88 74 63
4
He 379 293 230 184 149 122 102 85 72
4
Hea 411 319 252 201 164 135 112 94 79
Ne 768 579 445 347 276 221 180 148 122
Kr 2,045 1,478 1,090 819 625 483 379 301 241
Xe 2,701 1,930 1,409 1,047 791 606 471 370 294
222
Rn 3,168 2,235 1,611 1,182 883 669 514 400 314
H2 633 473 360 278 219 174 140 114 94
H2a 663 502 387 304 242 195 159 131 109
CH4 1,908 1,400 1,047 797 616 483 383 308 250
CO2 1,922 1,397 1,036 782 600 466 367 293 236
DMS 2,593 1,905 1,428 1,089 844 662 527 423 344
CH3Br 2,120 1,545 1,150 870 669 522 412 329 266
F12 (CCl2F2) 3,639 2,624 1,931 1,447 1,102 851 666 527 422
F11(CCl3F) 3,521 2,549 1,883 1,416 1,082 839 658 523 420
SF6 3,033 2,208 1,640 1,239 952 741 585 467 376
a
In sea water
Thus, the transfer resistances over several height intervals can be added in the
same way as electrical resistances that are connected in series. Typical values of the
transfer velocity across the water-side mass boundary layer are 106–105 m/
s (1–10 m/day). With respect to typical mixed layer depths in the ocean of about
100 m, gas transfer is a very slow process. It takes a time constant t = h/k =
10–100 days for the concentration of dissolved gases in the mixed layer to come
into equilibrium with the atmosphere.
Boundary Layer Thickness
The boundary layer thickness ~ z is defined as the thickness of a fictional layer

in which the flux is maintained only by molecular transport: j ¼ DðCs Cb Þ=~z .
Then with Eq. 8.4 the boundary layer thickness is given by:
D
z¼
~ : (8.7)
k
Geometrically, ~ z is given as the intercept of the tangent to the concentration

profile at the surface and the bulk (Fig. 8.1). With thicknesses between 20 and
200 mm, the mass boundary layer is extremely thin.
182 B. Jähne
Table 8.2 Diffusion coefficients for various gases and volatile chemical species in deionized
water and in some cases in sea water collected from Jähne et al. [18] and King et al. [20]. Die data
for momentum (kinematic viscosity) are from Sündermann [22]. Column 3 and 4 contain the
parameters for the fit of the diffusion coefficient: D = A exp[Ea/(RT)], the last four columns the
diffusion coefficients for 5 C, 15 C, 25 C, and 35 C
A Diffusion coefficient (10–5 cm2/s)
Molecular (10–5 cm2/ Ea (kJ/ s(Fit)
Species mass s) mole) % 5 C 15 C 25 C 35 C
Momentum – 1,517 1,139 893.4 724.3
Momentuma – 1,560 1,182 934.5 763.9
Heat – 379.2 2.375 135.80 140.72 145.48 150.08
3
Heb,c 3.02 941 11.70 2.1 5.97 7.12 8.39 9.77
4
He 4.00 818 11.70 2.1 5.10 6.30 7.22 8.48
4
Hea 886 12.02 1.8 4.86 5.88 7.02 8.03
Ne 20.18 1,608 14.84 3.5 2.61 3.28 4.16 4.82
Kr 83.80 6,393 20.20 1.6 1.02 1.41 1.84 2.40
Xe 131.30 9,007 21.61 3.5 0.77 1.12 1.47 1.94
222
Rnb 222.00 15,877 23.26 11 0.68 0.96 1.34 1.81
H2 2.02 3,338 16.06 1.6 3.17 4.10 5.13 6.23
H2a 1,981 14.93 4.3 3.05 3.97 4.91 5.70
CH4 16.04 3,047 18.36 2.7 1.12 1.48 1.84 2.43
CO2 44.01 5,019 19.51 1.3 1.07 1.45 1.91 2.43
DMSb 62.13 2,000 18.10 0.80 1.05 1.35 1.71
CH3Brb 94.94 3,800 19.10 0.98 1.31 1.71 2.20
F12b 120.91 4,100 20.50 0.58 0.79 1.05 1.37
(CCl2F2)
F11b(CCl3F) 137.37 3,400 20.00 0.60 0.81 1.07 1.38
SF6b 146.05 2,900 19.30 0.69 0.92 1.20 1.55
a
In sea water
b
Values of diffusion coefficients from fit, not measured values
c
Set 15% higher than 4He
Boundary Layer Time Constant
The time constant ~t for the transport across the mass boundary layer is given by:
z~ D
~t ¼ ¼ : (8.8)
k k2
Typical values for ~t are 0.04–4 s. Any chemical reaction with a time constant
larger than ~t does not significantly affect the transfer process. Therefore, CO2 can be
regarded as an inert gas, but not fast hydrating acid gases such as SO2.
The definitions of the three parameters k, ~ z , ~t are generally valid and do not
depend on any models of the boundary layer turbulence. According to Eqs. 8.7 and
8.8, they are coupled via the molecular diffusion coefficient. Therefore, only one of
them needs to be measured to get knowledge of all three parameters provided the
molecular diffusion coefficient of the species is known.
Partitioning of Transfer Between Air and Water
Because a mass boundary layer exists on both sides, it is important to determine

which one controls the transfer, that is, exhibits the largest transfer resistance (or
lowest transfer velocity). At the surface itself, the thermodynamic solubility equi-
librium is assumed to be established between the tracer concentrations ca in the gas
phase and cw in the liquid phase:
Cws ¼ aCas ; (8.9)
where a is the dimensionless solubility (partition coefficient). A solubility a 6¼ 1

causes a concentration jump at the surface (Fig. 8.1). Thus, the resulting total
transfer velocity kt or transfer resistance Rt can either be viewed from the gas
phase or the liquid phase. Adding them up, the factor a must be considered to
conserve the continuity of the concentration profile:
1 1 1
air side: ¼ þ ; Rat ¼ Ra þ Rw =a; (8.10)
kat ka akw
1 a 1
water side: ¼ þ ; Rwt ¼ aRa þ Rw :
kwt ka kw
The total transfer velocities in air and water differ by the factor a: kat = akwt. The
ratio akw/ka determines which boundary layer controls the transfer process. A high
solubility shifts control of the transfer process to the gas-phase boundary layer, and
a low solubility to the aqueous layer. The solubility value for a transition from air-
sided to water-sided control depends on the ratio of the transfer velocities. Typi-
cally kw is about 100–1,000 times smaller than ka. Thus, the transfer of even
moderately soluble volatile chemical species with solubilities up to 30 is controlled
by the water side. Some environmentally important species lie in a transition zone
where it is required to consider both transport processes (Fig. 8.2). The transfer of
highly soluble volatile and/or chemically reactive gas is controlled by the air-side
transfer process and thus analogous to the transfer of water vapor. The following
considerations concentrate on the water-side transfer process.
Gas Exchange at Smooth Water Surfaces
At smooth water surfaces, the theory of mass transfer is well established because it
is equivalent to mass transfer to a smooth solid wall. The turbulent diffusivity can
be described by the classic approach of Reichardt [7] with an initial z3 increase
that smoothly changes to a linear increase in the turbulent layer as in Eq. 8.3.
184 B. Jähne
Then integration of Eq. 8.6 yields the following approximation for Schmidt num-
bers higher than 100 [12, 16]:
1
kw ¼ uw Sc2=3 Sc > 60: (8.11)
12:2
This equation establishes the basic analogy between momentum transfer and gas
exchange. The transfer coefficients are proportional to the friction velocity in water,
which describes the shear stress (tangential force per unit area) t ¼ rw u2w applied
by the wind field at the water surface. Assuming stress continuity at the water
surface, the friction velocity in water is related to the friction velocity in air by:
1=2
ra
uw ¼ ua : (8.12)
rw
The friction velocity in air, u*a, can further be linked via the drag coefficient to
the wind speed UR at a reference height: cD = (u*a/UR)2. Depending on the
roughness of the sea surface, the drag coefficient has values between 0.8 and 2.4 ·
103. In this way, the gas exchange rate is directly linked to the wind speed. The gas
exchange further depends on the chemical species and the water temperature via the
Schmidt number.
Influence of Waves
A free water surface is neither solid nor is it smooth as soon as short wind waves are
generated. On a free water surface, velocity fluctuations are possible. Thus, there can
be convergence or divergence zone at the surface; surface elements may be dilated or
contracted. At a clean water surface, dilation or contraction of a surface element does
not cause restoring forces, because surface tension only tries to minimize the total
free surface area, which is not changed by this process. As a consequence of this
hydrodynamic boundary condition, the turbulent diffusivity normal to the interface
can now increase with the distance squared from the interface, Kc / z2 . Then:
1
kw ¼ uw Sc1=2 ; (8.13)
b
where b is a dimensionless constant.

In comparison to the smooth case in Eq. 8.11, the exponent n of the Schmidt
number drops from 2/3 to 1/2. This increases the transfer velocity for a Schmidt
number of 600 by about a factor of 3. The total enhancement depends on the value
of the constant b.
Wind waves cannot be regarded as static roughness elements for the liquid flow
because their characteristic particle velocity is of the same order of magnitude as
Fig. 8.3 Transfer velocity of

carbon dioxide plotted against
the transfer velocity of water with waves (unlimited fetch)
vapor. Measurements from a no waves
small annular wind/wave
facility Jähne [15] 10−2
kCO2 (Sc=600) [cm/2)

10−3
10−4
kH2O [cm/s]
0.1 1 10
the velocity in the shear layer at the surface. This fact causes a basic asymmetry
between the turbulent processes on the air and on the water sides of the interface.
Therefore, the wave effect on the turbulent transfer in the water is much stronger
and of quite different character than in the air. This basic asymmetry can be seen if
the transfer velocity for CO2 is plotted against the transfer velocity for water vapor
(Fig. 8.3, [15]). At a smooth water surface, the points fall well on the theoretical
curve predicted by the theory for a smooth rigid wall. However, as soon as waves
occur at the water surface, the transfer velocity of CO2 increases significantly
beyond the predictions.
Even at high wind speeds, the observed surface increase is well below 20%
[24]. When waves are generated by wind, energy is not only transferred via shear
stress into the water but a second energy cycle is established. The energy put by
the turbulent wind into the wave field is transferred to other wave numbers by
nonlinear wave–wave interaction and finally dissipated by wave breaking, viscous
dissipation, and turbulence. The turbulent wave dissipation term is the least
known term and of most importance for enhanced near-surface turbulence. Evi-
dence for enhanced turbulence levels below wind waves has been reported from
field and laboratory measurements. Experimental results also suggest that the
gas transfer rate is better correlated with the mean square slope of the waves as an
integral measure for the nonlinearity of the wind wave field than with the
wind speed.
186 B. Jähne
It is not yet clear, however, to what extent microscale wave breaking can account
for the observed enhanced gas transfer rates. A gravity wave becomes instable and
generates a steep train of capillary waves at its leeward face and has a turbulent
wake. This phenomenon can be observed even at low wind speeds, as soon as wind
waves are generated. At higher wind speeds, the frequency of microscale wave
breaking increases.
At high wind speeds, wave breaking with the entrainment of bubbles may
enhance gas transfer further. This phenomenon complicates the gas exchange
between atmosphere and the oceans considerably [19, 26]. First, bubbles constitute
an additional exchange surface. This surface is, however, only effective for gases
with low solubility. For gases with high solubility, the gas bubbles quickly come
into equilibrium so that a bubble takes place in the exchange only for a fraction of
its life time. Thus bubble-mediated gas exchange depends – in contrast to the
exchange at the free surface – on the solubility of the gas tracer. Second, bubble-
mediated gas transfer shifts the equilibrium value to slight supersaturation due to
the enhanced pressure in the bubbles by surface tension and hydrostatic pressure.
Third, breaking waves also enhance near-surface turbulence during the breaking
event and the resurfacing of submerged bubbles [17].
Experimental data are still too sparse for the size and depth distribution of
bubbles and the flux of the bubbles through the interface under various sea states
for a sufficiently accurate modeling of bubble-mediated air-sea gas transfer and
thus a reliable estimate of the contribution of bubbles to the total gas transfer rate.
Some experiments from wind/wave tunnels and the field suggest that significant
enhancements can occur, other experiments could not observe a significant influ-
ence of bubbles.
Influence of Surface Films
A film on the water surface creates pressure that works against the contraction
of surface elements. This is the point at which the physicochemical structure of
the surface influences the structure of the near-surface turbulence as well as the
generation of waves [13, 21]. As at a rigid wall, a strong film pressure at the surface
maintains a two-dimensional continuity at the interface just as at a rigid wall.
Therefore, Eq. 8.11 should be valid for a smooth film-covered water surface and
has indeed been verified in wind/wave tunnel studies as the lower limit for
the transfer velocity. As a consequence, both Eqs. 8.11 and 8.13 can only be
regarded as limiting cases. A more general approach is required that has not yet
been established. One possibility is a generalization of Eqs. 8.11 and 8.13 to:
1
kw ¼ uw ScnðsÞ ; (8.14)
bðsÞ
where both b and n depend on dimensionless parameters describing the surface

conditions s. Even films with low film pressure may easily decrease the gas transfer
rate to half of its value at clean water surface conditions. But still too few
measurements at sea are available to establish the influence of surfactants on gas
transfer for oceanic conditions more quantitatively.
Experimental Techniques and Results
Laboratory Facilities
Laboratory facilities play an important role in the investigation of air-sea gas

transfer. Only laboratory studies allow a systematic study of the mechanisms and
are thus an indispensable complement to field experiments. Almost all basic knowl-
edge about gas transfer has been gained by laboratory experiments in the past.
Geochemical Tracer Techniques
The first oceanic gas exchange measurements were performed using geochemical
tracer methods such as the 14C [2], 3He/T [23], or 222Rn/226Ra [8, 9] methods. The
volume and time-average flux density is given by mass balance of the tracer
concentration in a volume of water Vw:
Vw c_w ¼ Fw j or j ¼ hw c_w ; (8.15)
where Fw, and hw are the surface area and the effective height Vw /Fw of a well-
mixed water body, respectively. The time constant tw = hw /k is in the order of days
to weeks. It is evident that the transfer velocities obtained in this way provide only
values integrated over a large horizontal length scales and time scales in the order of
tw . Thus a parametrization of the transfer velocity is only possible under steady
state conditions over extended periods. Moreover, the mass balance contains many
other sources and sinks besides air-sea gas exchange and thus may cause severe
systematic errors in the estimation of the transfer velocity. Consequently, mass
balance methods are only poorly suited for the study of the mechanisms of
air–water gas transfer.
188 B. Jähne
Tracer Injection
The pioneering lake studies for tracer injection used sulfur hexafluoride (SF6).
However, the tracer concentration decreases not only by gas exchange across the
interface but also by horizontal dispersion of the tracer. This problem can be
overcome by the dual tracer technique [10] simultaneously releasing two tracers
with different diffusivities (e.g., SF6 and 3He). When the ratio of the gas transfer
velocities of the two tracers is known, the dilution effect by tracer dispersion can be
corrected, making it possible to derive gas transfer velocities. But the basic problem
of mass balance techniques, that is, their low temporal resolution, remains also with
artificial tracer approaches.
Eddy Correlation Flux Measurements
Eddy correlation techniques are used on a routine basis in micrometeorology, that

is, for tracers controlled by the boundary layer in air (momentum, heat, and water
vapor fluxes). Direct measurements of the air-sea fluxes of gas tracers are very
attractive because the flux densities are measured directly and have a much better
temporal resolution than the mass balance based techniques. Unfortunately, large
experimental difficulties arise when this technique is applied to gas tracers con-
trolled by the aqueous boundary layer [Broecker 1986, 14]. The concentration
difference in the air is only a small fraction of the concentration difference across
the aqueous mass boundary layer. But after more than 20 years of research has this
technique delivered useful results. Some successful measurements under favorable
conditions have been reported and it appears that remaining problems can be
overcome in the near future.
Passive and Active Thermography
The basic idea of this technique is to determine the concentration difference across the
mass boundary layer when the flux density j of the tracer across the interface is known.
The local transfer velocity can be determined by simply measuring the concentration
difference Dc across the aqueous boundary layer (cold surface skin temperature)
according to Eq. 8.4 with a time constant ~t for the transport across the boundary
layer Eq. 8.8. This technique is known as the controlled flux technique (CFT).
Heat proves to be an ideal tracer for the CFT. The temperature at the water
surface can then be measured with high spatial and temporal resolution using IR
thermography. A known and controllable flux density can be applied by using
infrared radiation. Infrared radiation is absorbed in the first few ten micrometers at
the water surface. Thus a heat source is put right at top of the aqueous viscous
boundary layer. Then the CFT directly measures the water-side heat transfer velocity.
Fig. 8.4 Summary of gas

exchange field data
normalized to a Schmidt
number of 600 and plotted 14
C
versus wind speed together
SF6 - 3He
with the empirical 222
Rn
relationships of Liss and
CO2 eddy correlation
k600 [cm/h]
Merlivat [11] and ? (After
Jähne and Haußecker [6]) heat (CFT)
wind speed u10 [m/s]
A disadvantage of the CFT is that the transfer velocity of gases must be

extrapolated from the transfer velocity of heat. The large difference in the Schmidt
number (7 for heat, 600 for CO2) casts some doubt whether the extrapolation to so
much higher Schmidt numbers is valid.
Two variants of the technique proved to be successful. Active thermography
uses a CO2 laser to heat a spot of several centimeter in diameter on the water
surface. The heat transfer rates are estimated from the temporal decay of the heated
spot. Passive thermography uses the naturally occurring heat fluxes caused by latent
heat flux jl, sensible heat flux js and longwave emission of radiation jr . The net heat
flux jn = jl + js + jr results according to Eq. 8.4 in a temperature difference across
the interface of DT = jh/(rcpkh). Because of the turbulent nature of the exchange
process any mean temperature difference is associated with surface temperature
fluctuations which can be observed in thermal images. With this technique,
the horizontal structure of the boundary layer turbulence can be observed. Surface
renewal is directly observable in the IR image sequences, which show patches of
fluid being drawn away from the surface.
With some knowledge about the statistics of the temperature fluctuations, the
temperature difference DT across the interface as well as the time constant ~t of heat
transfer can be computed from the temperature distribution at the surface. Results
obtained with this technique are shown in Fig. 8.4.
Summary of Field Data
A collection of field data is shown in Fig. 8.3. Although the data show a clear increase
of the transfer velocity with wind speed, there is substantial scatter in the data that can
only partly be attributed to uncertainties and systematic errors in the measurements.
190 B. Jähne
Thus, in addition, the field measurements reflect the fact that the gas transfer velocity
is not simply a function of the wind speed but depends significantly on other
parameters influencing near-surface turbulence, such as the wind-wave field and
the viscoelastic properties of the surface film.
Empiric Parametrization
Given the lack of knowledge, all theories about the enhancement of gas transfer by
waves are rather speculative and are not yet useful for practical application. Thus, it
is still state of the art to use semiempiric or empiric parameterizations of the gas
exchange rate with the wind speed. Most widely used is the parametrization of Liss
and Merlivat [11]. It identifies three physically well-defined regimes (smooth,
wave-influenced, bubble-influenced) and proposes a piece-wise linear relation
between the wind speed U and the transfer velocity k:
8
> 2=3
>
> 0:472UðSc=600Þ U 3:6 m/s
>
<
7:971ðU 3:39ÞðSc=600Þ1:2 U > 3:6 m/s and
k ¼ 106 (8.16)
>
> U 13 m/s
>
>
:
16:39ðU 8:36ÞðSc=600Þ1=2 U > 13 m/s
At the transition between the smooth and wavy regime, a sudden artificial jump
in the Schmidt number exponent n from 2/3 to 1/2 occurs. This actually causes
a discontinuity in the transfer rate for Schmidt number unequal to 600.
The empiric parametrization of Wanninkhof [25] simply assumes a quadratic
increase of the gas transfer rate with the wind speed:
k ¼ 0:861 106 ðs=mÞU 2 ðSc=600Þ1=2 : (8.17)
Thus, this model has a constant Schmidt number exponent n = 1/2. The two
parameterizations differ significantly (see Fig. 8.3). The Wanninkhof parametriza-
tion predicts significantly higher values. The discrepancy between the two
parameterizations and many others proposed (up to a factor of two) mirrors the
current uncertainty in estimating the air-sea gas transfer rate.
Future Directions
In the past, progress toward a better understanding of the mechanisms of air–water

gas exchange was hindered by inadequate measuring technology. However, new
techniques have become available and will continue to become available that will
give a direct insight into the mechanisms under both laboratory and field conditions.
This progress will be achieved by interdisciplinary research integrating such

different research areas as oceanography, micrometeorology, hydrodynamics,
physical chemistry, applied optics, and image processing.
Optical and image processing techniques will play a key role because only
imaging techniques give direct insight to the processes in the viscous, heat, and
mass boundary layers on both sides of the air–water interface. Eventually, all key
parameters including flow fields, concentration fields, and waves will be captured
by imaging techniques with sufficient spatial and temporal resolution. The experi-
mental data gained with such techniques will stimulate new theoretical and
modeling approaches.
Bibliography
Primary Literature
1. Bohr C (1899) Definition und methode zur bestimmung der invasions- und evasionscoef-
ficienten bei der auflösung von gasen in flüssigkeiten. Werthe der genannten constanten sowie
der absorptionscoefficienten der kohlensäure bei auflösung in wasser und in chlornatrium-
lösungen. Ann Phys Chem 68:500–525
2. Bolin B (1960) On the exchange of carbon dioxide between the atmosphere and the sea. Tellus
12(3):274–281. ISSN 2153-3490
3. Revelle R, Suess HE (1957) Carbon dioxide exchange between atmosphere and ocean and the
question of an increase of atmospheric CO2 during the past decades. Tellus 9:18–27
4. Jähne B (1982) Dry deposition of gases over water surfaces (gas exchange). In: Flothmann
D (ed) Exchange of air pollutants at the air-earth interface (dry deposition). Battelle Institute,
Frankfurt
5. Jähne B (2009) Air-sea gas exchange. In: Steele JH, Turekian KK, Thorpe SA (eds) Encyclo-
pedia ocean sciences. Elsevier, Boston, pp 3434–3444
6. Jähne B, Haußecker H (1998) Air-water gas exchange. Annu Rev Fluid Mech 30:443–468
7. Reichardt H (1951) Vollständige darstellung der turbulenten geschwindigkeitsverteilung in
glatten leitungen. Z Angew Math Mech 31:208–219
8. Peng TH, Broecker WS, Mathieu GG, Li Y-H, Bainbridge A (1979) Radon evasion rates in the
Atlantic and Pacific oceans as determined during the geosecs program. J Geophys Res
84(C5):2471–2487
9. Roether W, Kromer B (1984) Optimum application of the radon deficit method to obtain
air–sea gas exchange rates. In: Brutsaert W, Jirka GH (eds) Gas transfer at water surfaces.
Reidel, Hingham, pp 447–457
10. Watson AJ, Upstill-Goddard RC, Liss PS (1991) Air-sea exchange in rough and stormy seas
measured by a dual tracer technique. Nature 349(6305):145–147
11. Liss PS, Merlivat L (1986) Air-sea gas exchange rates: introduction and synthesis. In: Buat-
Menard P (ed) The role of air-sea exchange in geochemical cycling. Reidel, Boston,
pp 113–129
12. Deacon EL (1977) Gas transfer to and across an air-water interface. Tellus 29:363–374
13. Frew NM (1997) The role of organic films in air-sea gas exchange. In: Liss PS, Duce RA (eds)
The sea surface and global change. Cambridge University Press, Cambridge, pp 121–171
14. Jacobs C, Nightingale P, Upstill-Goddard R, Kjeld JF, Larsen S, Oost W (2002) Comparison of
the deliberate tracer method and eddy covariance measurements to determine the air/sea
192 B. Jähne
transfer velocity of CO2. In: Saltzman E, Donelan M, Drennan W, Wanninkhof R (eds) Gas
transfer at water surfaces. Geophysical Monograph, vol 127. American Geophysical Union
15. Jähne B (1980) Zur Parametrisierung des Gasaustauschs mit Hilfe von Laborexperimenten.
Dissertation, Institut für Umweltphysik, Fakultät für Physik und Astronomie, Univ.
Heidelberg, http://d-nb.info/810123614. IUP D-145
16. Jähne B (1987) Image sequence analysis of complex physical objects: nonlinear small scale
water surface waves. In: Proceedings of 1st international conference on computer vision,
London, pp 191–200
17. Jähne B (1991) New experimental results on the parameters influencing air-sea gas exchange.
In: Wilhelms SC, Gulliver JS (eds) Air-water mass transfer, Selected papers from the 2nd
international symposium on gas transfer at water surfaces, Minneapolis, 11–14 Sep 1990.
ASCE, pp 582–592
18. Jähne B, Heinz G, Dietrich W (1987) Measurement of the diffusion coefficients of sparingly
soluble gases in water. J Geophys Res 92(C10):10,767–10,776
19. Keeling RF (1993) On the role of large bubbles in air-sea gas exchange and supersaturation in
the ocean. J Marine Res 51:237–271
20. King DB, Bryun WJD, Zheng M, Saltzman ES (1995) Uncertainties in the molecular diffusion
coefficient of gases in water for use in the estimation of air-sea exchange. In: Jähne B,
Monahan E (eds) Air-water gas transfer, Selected papers, 3rd international symposium on
air-water gas transfer. AEON, Hanau, pp 13–22
21. McKenna SP, Bock EJ (2006) Physicochemical effects of the marine microlayer on air-sea gas
transport. In Gade M, Hühnerfuss H, Korenowski GM (eds) Marine surface films: chemical
characteristics, influence on air-sea interactions, and remote sensing. Springer, Berlin/
Heidelberg, pp 77–91
22. Sündermann J (ed) (1986) Landolt-Börnstein, vol V 3c, Oceanography. Springer, Heidelberg
23. Torgersen T, Top Z, Clarke WB, Jenkins WJ (1977) A new method for physical limnology-
tritium-helium-3 ages results for Lakes Erie, Huron, and Ontario. Limnol Oceanogr
22:181–193
24. Tschiersch J, Jähne B (1980) Gas Exchange trough a rough water surface in a circular
windtunnel; Wave characteristics under limited and unlimited fetch. In: Broecker HC, Hasse
L (eds) Berichte aus dem Sonderforschungsbereich 94 Meeresforschung – Symposium on
capillary waves and gas exchange, Trier, 2–6 July 1979, number 17. Univ. Hamburg, pp 63–70
25. Wanninkhof R (1992) Relationship between wind speed and gas exchange over the ocean.
J Geophys Res 97:7373–7382
26. Woolf D, Leifer I, Nightingale P, Rhee T, Bowyer P, Caulliez G, de Leeuw G, Larsen S,
Liddicoat M, Baker J, Andreae M (2007) Modelling of bubble-mediated gas transfer: funda-
mental principles and a laboratory test. J Marine Syst 66:71–91
Books and Reviews
Banner ML (ed) (1999) The wind-driven air-sea interface, electromagnetic and acoustic sensing,
wave dynamics and turbulent fluxes, proceedings of the symposium Sydney, Australia, 11–15
January 1999. University of New South Wales, Sydney
Bengtsson LO, Hammer CU (eds) (2001) Geosphere-biosphere interactions and climate.
Cambridge University Press, Cambridge
Borges AV, Wanninkhof R (2007) Fifth international symposium on gas transfer at water surfaces.
J Mar Syst 66(1–4):1–308
Brasseur GP, Prinn RG, Pszenny AA (eds) (2003) Atmospheric chemistry in a changing world.
An integration and synthesis of a decade of tropospheric chemistry research. Springer, Berlin
Brutsaert W, Jirka GH (eds) (1984) Gas transfer at water surfaces. Reidel, Dordrecht
Csanady GT (2001) Air-sea interaction, laws and mechanisms. Cambridge University Press,
Cambridge
Cussler EL (2009) Diffusion – mass transfer in fluid systems, 3rd edn. Cambridge University
Press, Cambridge
Danckwerts PV (1970) Gas-liquid reactions. MacGraw-Hill, New York
Davies JT (1972) Turbulence phenomena. An introduction to the eddy transfer of momentum,
mass, and heat, particularly at interfaces. Academic, New York/London
Davies JT, Rideal EK (1963) Interfacial phenomena, 2nd edn. Acadamic, New York
Dobson F, Hasse L, Davis R (eds) (1980) Air-sea interaction: instruments and methods. Plenum,
New York
Donelan MA, Hui WH, Plant WJ (eds) (1996) The air-sea interface, radio and acoustic sensing,
turbulence and wave dynamics, proceedings of the symposium, Marseille, France, 24–30 June
1993. Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami,
Florida
Donelan MA, Drennan WM, Saltzman ES, Wanninkhof R (eds) (2002) Gas transfer at water
surfaces. American Geophysical Union, Washington, DC
Emerson SR, Hedges J (2008) Chemical oceanography and the marine carbon cycle. Cambridge
Favre A, Hasselmann K (eds) (1978) Turbulent fluxes through the sea surface, wave dynamics, and
prediction, proceedings of the symposium, Marseille, 1977. Plenum, New York
Fogg PGT, Sangster J (2003) Chemicals in the atmosphere: solubility, sources, and reactivity.
Wiley, Chichester
Gade M, Hühnerfuss H, Korenowski GM (eds) (2005) Marine surface films: chemical
characteristics, influence on air-sea interactions and remote sensing. Springer, Berlin
Garbe CS, Handler RA, Jähne B (eds) (2007) Transport at the air sea interface – measurements,
models and parameterizations. Springer, Berlin
Gulliver JS (2007) Introduction to chemical transport in the environment. Cambridge University
Press, Cambridge
Jähne B, Monahan EC (eds) (1995) Air-water gas transfer – selected papers from the third
international symposium on air-water gas transfer. AEON Verlag & Studio Hanau. http://
d-nb.info/946682526
Kantha LH, Clayson CA (2000) Small scale processes in geophysical fluid flows, vol 67, Interna-
tional geophysics series. Acadamic, San Diego
Komori S, McGillis W (eds) (2011) Gas transfer at water surfaces, selected papers from the 6th
international symposium. Kyoto University Press, Kyoto
Kraus EB, Businger JA (1994) Atmosphere-ocean interaction, vol 27, 2nd edn, Oxford
monographs on geology and geophysics. Oxford University Press, New York
Liss PS, Duce RA (eds) (2005) The sea surface and global change. Cambridge University Press,
Cambridge
Robinson IS (2010) Discovering the ocean from space – the unique applications of satellite
oceanography. Springer, Heidelberg
Sarmiento JL, Gruber N (2006) Ocean biogeochemical dynamics. Princeton University Press,
Princeton
Soloviev A, Lukas R (2006) The near-surface layer of the ocean, vol 31, Atmospheric and
oceanographic sciences library. Springer, Dordrecht
Wanninkhof R, Asher WE, Ho DT, Sweeney C, McGillis WR (2009) Advances in quantifying air-
sea gas exchange and environmental forcing. Annu Rev Mar Sci 1:213–244
Wilhelms SC, Gulliver JS (eds) (1991) Air-water mass transfer – selected papers from the 2nd
international symposium on gas transfer at water surfaces, Minneapolis Minnesota, September
11–14, 1990. American Society of Civil Engineers, New York
Zeebe RE, Wolf-Gladrow DA (2001) CO2 in seawater: equilibrium, kinetics, isotopes, vol 65,
Elsevier oceanography series. Elsevier, Amsterdam
Chapter 9
Sediment–Water Interfaces, Chemical Flux at
Louis J. Thibodeaux and Joseph Germano
Glossary
Benthic boundary layer A slow moving water layer above the sediment.
Bioturbation transport A chemical mobility process driven by the presence of
macrofauna and macroflora residing near the interface.
Chemical flux The basic term that quantifies chemical mobility across an
interface with units of mass per area per time (kg/m2/s)
Chemical mobility A general term used to denote the idea that chemicals do
move from place to place.
Interface A real or imaginary plane which separates water from
sediment.
Mass transfer rate The chemical flux times the area perpendicular to its
direction of movement (kg/s).
Sediment surface layers A series of distinctive mud layers occupying thickness
of several centimeters depth.
Transport model One of several concepts for describing a chemical
mobility process, and the associated formula or algo-
rithm needed to describe it mathematically (a.k.a., the
flux expression).
L.J. Thibodeaux (*)
Cain Department Chemical Engineering, Louisiana State University, Baton Rouge,
LA 70820, USA
J. Germano
Germano & Associates, Inc., 12100 SE 46th Place, Bellevue, WA 98006, USA
196 L.J. Thibodeaux and J. Germano
Numerous individual transport processes which mobilize chemicals on either side

of the interface have been studied. However, a consistent theoretical framework
connecting the processes across the interface that correctly quantifies the overall
flux remains elusive. This occurs because two fundamentally different individual
flux relationships are needed to represent the two very different transport
mechanisms needed for quantifying the numerous chemical, biological, and physi-
cal processes ongoing at this unique locale. The two basic types of transport
processes are the chemical potential driven and the media advection driven. Several
theoretical modeling approaches exist for combining these, but all have problematic
conceptual features, which will be reviewed. By generalizing flux continuity across
the interface, which is the fundamental basis for arriving at the well-known and
accepted two-resistance theory, the “interface compartment model” is presented
and offered as a unifying theory describing advection-driven and potential-driven
transport across the sediment–water interface.
Introduction
All theories for chemical transport across the interface originate from the nineteenth
century with the Ohm–Kirchhoff laws of electrical currents and potentials. Lewis and
Whitman [1] used an analogous electrical flux approach for deriving the chemical
potential–driven flux across a gas–liquid interface. It is presumed that most of
the significant, individual chemical transport processes on both sides of the
sediment–water interface, which influence the flux of geochemicals as well as
the anthropogenic ones, have been discovered. Many have been reported and are the
subject of several reviews. They are the result of biological, chemical, and geophysical
processes, and most have been verified based on observations in the field and/or in the
laboratory. Some have been thoroughly studied while others have not. As
a consequence, there are well-developed descriptions for several processes as well as
many theoretical equations for the flux. A unified theory is proposed for connecting flux
across the interface.
Photographs taken of the interface region, obtained using a sediment profile
camera [2, 3] are displayed in Figs. 9.1–9.4. (Figures 9.1–9.4 A collection of color
images of the sediment–water interface. These are selected photos taken by Joseph
Germano over a time period of 28 years. Four categories are presented.) Fig. 9.1
contains images of the interface and a small sampling of the wide range of effects
caused by various macrofauna. Figure 9.2 shows images of the interface being acted
upon by submerged aquatic plants with leafy parts in the water column above and
holdfasts below. Figure 9.3 has images which indicate an interface under the
influence of low oxygen and/or chemical pollutant stresses. Finally, Fig. 9.4
shows the particle advection process. These are but a few glimpses of the character
9 Sediment–Water Interfaces, Chemical Flux at 197
a b c
Deep-sea octocoral (Kophobolemnon) Mound of clam fecal pellets effectively Bryozoans, polychaetes and a clam
suspension feeding in benthic boundary changing sediment grain-size and siphon can be seen projecting above
layer. San Francisco Deep Ocean Disposal transport properties. Elliott Bay Disposal the sediment-water interface. San
Site, 2850 meters. Site, Puget Sound, 102 meters. Diego Harbor, 10 meters.
d e f
Infaunal deposit-feeding polychaetes Deep sea colonial foram colony with Dense assemblage of Ampelisca abdita
aerating sediment to depth. Puget commensal amphipods. Mid-Atlantic tubes. Jamaica Bay, Long Island, 5 meters.
Sound, 70 meters. continental slope off Virginia, USA,1600 meters.
g h i
Subsurface excavations caused by ghost Deposit-feeding polychaetes are visible in Transected polychaete burrow at depth
shrimp (Calianassa) visible at sediment the sediments below the nudibranch at the demonstrates how animals in the "anoxic
surface. Pearl Harbor, Hawaii, 12 meters. surface. Nearshore embayment in Unalaska, subsurface sediments" create their own
Aleutian Islands, 36 meters. oxygenated mileu. Peconic Estuary, Long
Island, NY, 6 meters.
Fig. 9.1 Animal–sediment–fluid relationships (9 photos)
and forms of this important global environmental interface which occupies the
largest plane area on Earth. It separates the fluid water–dominated media above and
the underlying solid, fluid water–saturated zone below. Figure 9.5 is a conceptual
illustration of the interface and its adjoining regions, modified from the original by
a b
Subsurface infaunal burrows can be seen beneath the These sandy clays support a dense assemblage of the
fronds of Caulerpa prolifera on the sediment surface. invasive Caulerpa racemosa. Coastal embayment off
Coastal embayment off Sicily, 14 meters. Sicily, 8 meters.
c d
Mixed assemblage of red algae with a frond of Laminaria Fronds of seagrass can be seen above these silty fine
pushed below the sediment surface by the camera prism. sands. Sinclair Inlet, Puget Sound, 4 meters.
Sinclair Inlet, Puget Sound, 5 meters.
Fig. 9.2 Submerged aquatic vegetation (4 photos)

a b
Records of past anoxic events are preserved in the White aggregations of Beggiatoa colonies can
layers of these clays. Caspian Sea, 750 meters. be seen on the surface of these anoxic sediments.
Nearshore embayment in Unalaska, Aleutian
Islands, 32 meters.
c d
No visible RPD can be seen in these organically enriched A diffusional layer of oxidized sediment can be seen
sediments. Hackensack River, New Jersey, 3 meters. on the surface of these riverine sediments. Hackensack
River, New Jersey, 4.5 meters.
Fig. 9.3 Polluted sediments (4 photos)

a b
Deposit-feeding infaunal taxa such as maldanid polychaetes Subsurface methanogenesis can also periodically release
can loosen sediment fabric and eject particles into the gas and sediment particles into the overlying water.
benthic boundary layer. Long Island Sound, 15 meters. Saginaw River , 3 meters.
Fig. 9.4 Examples of particle advection (2 photos)
Santschi et al. [4]. It is an idealized model of the previous images showing the
location of the interface plane and some of the processes occurring in the adjoining
bulk media phases or compartments.
At any interface locale, it is expected that a combination of individual pro-
cesses on either side may control the net flux. There may be a dozen or more
individual processes, identifying the key ones on either side and coupling them in
a logical fashion so as to understand the overall process and quantify the effective
flux is an ongoing challenge. As the photographs show, each sediment–water
interface has unique characteristics so that the flux is expected to be highly
variable from locale to locale. One goal of this chapter is to develop a unified
theory for combining the individual processes on the water-side to those on the
sediment-side and to obtain the appropriate algorithm for the across-media or
interphase flux. The substances of concern are geochemicals such as nitrogen,
silica, carbon, and lead and anthropogenic chemicals such as polychlorinated
biphenyls, naphthalene, ibuprofen, and caffeine. The development also is appli-
cable to aquasols, nanoparticles, and other identifiable particles moving across the
sediment–water interface.
This chapter will first review the available transport processes theories. Several
significant individual processes will be listed and summarized. Based on the
mechanisms that drive the transport, each will be placed into one of two categories
of flux equation types. The theoretical arguments supporting the use of two
categories will be covered. Then, a model development section will begin with
hv
atmosphere
air-water
interface hv gas exchange water

CO2, H2O2 nutrients (volatile forms)
primary production
adsorption onto
flow in suspended particles flow out
coagulation
of particles
settling
desorption
benthic boundary
filtering layer
resuspension
of particles of particles
sediment -water
interface interstitial - water

transport
a) diffusion
b) biological irrigation
particle mixing
by organisms
Fig. 9.5 Transport processes near the sediment–water interface. This figure is a slightly modified
version of the original produced by Santschi et al. [4]. It is a classical graphical illustration of the
interface plane. It is an idealized model of the interface region showing some of the processes
occurring in the water column and the bed surface layers
a literature review and summary of the various theoretical approaches proposed for
combining the individual processes arriving at the net flux across the interface. The
algebraic forms of the flux equations will be given. Finally, the concepts behind the
proposed interface compartment (IC) theory will be presented and the flux equation
derived. Included will be a discussion of the IC theory in relation to the existing ad
hoc protocols in use.
The methods section will describe the modeling approach, the chemicals used in
numerical simulations, and the individual processes selected for performing the flux
calculations. Including flux calculations in the theoretical section is necessary
because it extends and amplifies the IC model theory by providing a layer of reality
to accompany the mathematical formulas. In addition, it provides a quantitative
means of comparing it to the approaches being used.
Transport Process Theories
There are numerous individual chemical transport processes operating on both

sides of the sediment–water interface, and they have been the subject of literature
reviews [4–6] and monographs [7–12]. Two types of equations are commonly
used that reflect process mechanisms. They are the chemical potential and media
advection; one or the other type will be applicable to each process as they are
presented and discussed. The chemical potential type will be considered first.
The molecular diffusive transport process derived from Fick’s first law of
diffusion is the most basic and ubiquitous process [13–15]. The integrated equation
is a chemical potential–type process. It contains a concentration difference term
that reflects the chemical potential between two locations in space. It also contains
an effective diffusion coefficient for the porous medium and the path length. It is
applicable to the transport of all solutes and to Brownian particles in bed sediment
as well as on the water-side of the interface. On the water-side of the interface, the
mass transfer coefficient often replaces the diffusion coefficient and path length
quotient [14].
At this juncture, it is well appreciated by the reader that chemical flux, such as
molecular diffusion, is a function of the concentration difference or gradient and
a kinetic transport parameter. Chemical reactions and phase partitioning can and do
occur in the layers on either side of the interface, which affects the magnitude of the
respective concentrations and hence the flux. However, applying the transient,
reactive-diffusion equation to the chemical species of interest in each layer is
beyond the scope of this article. For the sake of clarity, the tactical analytical
approach taken in this manuscript portrays the sediment–water interface region so
as to isolate and focus only on the transport processes. Therefore, the reader should
note that the two layers of interest are assumed to be very thin, void of chemical
reactions within (i.e., degradation, oxidation/reduction, polymerization, etc.), and
have constant concentration differences across them. In other words, the chemical
species are conserved, and a steady-state flux is occurring in the defined
sediment–water interface region.
The term Brownian diffusion is conventionally applied to very small particles
that respond to the kinetic motions of the surrounding solvent molecules. The
transport of colloids, also termed aquasols, in the water on either side of the
interface is quantified by a chemical potential–type transport process. These are
present in the form of dissolved organic carbon (DOC) particles or inorganic
particles with sorbed chemical fractions. The bioturbation transport process of
particles in the bed as well as in the adjoining porewater is driven by the presence
of macrofauna. It is consistent with the chemical potential–type transport mecha-
nism in that the randomness of a collection of macrofauna-driven particle and fluid
motions mimics the molecular kinetic mechanism of diffusion on a larger physical
scale. For this reason, it is termed a biodiffusion process and is treated as such
mathematically [9, 14, 16–19]. Bioturbation has also been depicted as a convective
transport process. This typically involves use of bed turnover rate and nonlocal
particle movement rates. However, the potential-based biodiffusion also has theo-
retical merit in that the range of macrofauna sizes and transport lengths tends to
approach a Gaussian distribution, indicating that the process can be described
by a diffusion type of equation. In addition, the biodiffusion model has an extensive
database of field- and laboratory-measured biodiffusion transport coefficients that
convective transport lacks for this process [19].
The two generic forms of the chemical potential–type flux equations are

F ¼ ðD=hÞ Cs Ci (9.1)
and

F ¼ K Ci Cw ; (9.2)
where F (kg/m2/s) is the flux, D is the diffusion coefficient (m2/s), h is the path
length (m), K is the convective mass transfer coefficient (m/s), and C (kg/m3)
represents the concentration in water for the sediment bed (s), the interface (i),
and the water column (w) beyond the benthic boundary layer, respectively.
The media convective–type rate equation for chemical flux reflects a transport
mechanism driven by the directed motion of a bulk media. Several types operate
across the interface region. In-bed porewater convection moves solutes and fine
particles in both directions. The porewater flow direction responds to hydraulic
pressure differences across the bed layers. These can be long-range pressure
differences such as in-bank and water column head differences or localized pres-
sure differences generated by the flowing water column as it encounters local
bottom roughness such as sand waves, mounds, etc. [20]. In either case, the
chemical flux is the product of the effective Darcian porewater velocity and its
aqueous concentration.
Solid particles moving through the water-side benthic boundary layer are also
a media convective–type flux. The primary ones are particle deposition onto the bed
from the water column and resuspension from the bed surface. These transport
mechanisms are initiated and maintained by the action of the flowing water. The
fluid-generated shear stress at the bed surface drives particle movements into
suspension as well as change the particle deposition probability [21, 22]. However,
particles and so-called marine snow are also formed in the water column [23], and
others originate as wind-blown dust on the sea surface, etc.; all types are deposited
onto the bed. The physics of cohesive and noncohesive sediment transport employs
complex algorithms for their estimation [24, 25]. For the purpose of this manu-
script, a deposition velocity and a resuspension velocity will be used to characterize
the respective processes. In each case, the flux is equal to a velocity–concentration
product so the generic forms of the media convective–type equation are
F ¼ nw Cw (9.3)
and
F ¼ np Cp ; (9.4)
where nw and np (m/s) are the effective media velocities of water and particles
perpendicular to the interface plane, and Cw and Cp are the media concentrations.
The above review covers the most well-known, characterized, and quantified
individual processes. Several other processes have been observed and described.
One termed the “benthic cannon” is dramatic and appears in Fig. 9.4. As shown, the
organism responsible for this phenomenon (a maldanid polychaete) is seen
injecting a spray of fine particles from its burrow into the lower portion of the
water column. In a similar mechanism, gas bubbles generated within the bed can
also move to the interface and emerge, likewise injecting fine particles into the
water-side boundary layer (Fig. 9.4). The role of the nepheloid layer, made up of
submerged aquatic vegetation and other macrofauna activities that enhance and
attenuate chemical transport processes, remains to be studied and quantified. Due to
the lack of sufficient information on these and other individual processes, they
cannot be included in numerical simulations at this time.
Several computational studies have been performed aimed at comparing
aspects of various individual processes. The most comprehensive of these is
a study of trichlorobiphenyl (TCP) for nine in-bed transport mechanisms [26].
Individual processes were ranked by characteristic times-of-recovery of TCP in
freshwater riverine bed sediment. It was concluded that in high-energy
environments, sediment transport was likely the dominant sediment-side TCP
transport process, while in low-energy environments, bioturbation was likely to
dominate the movement rate of TCP in the upper layer of the bed. Singh et al. [21]
developed a framework for a comprehensive mathematical model for fine and
cohesive sediment transport to be combined with contaminant transport models in
rivers, lakes, and estuaries; they observed that much research needs to be done
before truly realistic chemical exchange models for the sediment–water interface
will be available for practical use. They further noted that sediment transport and
chemical transport must be meshed in development of a comprehensive model. In
assessing the soluble release process of polychlorinated biphenyls (PCBs) from
bed sediment in three North American rivers, Thibodeaux et al. [27] evaluated
five individual transport processes by comparing the magnitudes of the mass
transfer coefficients. However, such individual process studies do not completely
address the interconnections that result in across-media transport at the
sediment–water interface. Comprehensive studies of the interconnections of
transport processes are lacking. A study limited to solute transport of
polychlorinated biphenyls in the Hudson River highlights the importance of
connecting processes across the interface plane [28, 29]. It was found that during
active in-bed bioturbation, the transport resistance on the water-side benthic
boundary layer is significant and that both of these processes regulate the PCB
flux from the bed.
The number and complexity of the biological, chemical, and geophysical pro-
cesses cooperating to drive chemical flux across the interface is daunting, and
sorting out the cause-and-effect factors is confusing without the aid of theoretical
guidance. Ad hoc approaches are in use, but all have theoretical shortcomings and
may not, therefore, extend into areas outside of the data set. A robust theory that can
accommodate the types of various individual chemical transport processes on either
side of the interface and connect them in a logical and transparent procedure is
needed for several reasons. In the first place, there appear to be none available.
Second, having one will lead to much better understanding of the overall situation
related to transport and will provide a hypothesis for interpretation of flux data from
both laboratory and field measurements. Third, modelers of aquatic chemodynamic
processes need a theory-based procedure for connecting chemical movement
between the adjoining bulk-phase compartments based on first principles.
Theories and Model Development
G. S. Ohm (1787–1854) found that the electric current (I) is directly proportional to
the difference in voltaic potential between the ends of a conductor (V) and the
proportionality constant is R1, where R is the resistance. G. R. Kirchhoff
(1824–1887) extended Ohm’s law. By analogy, the Ohm–Kirchhoff laws were
applied by Whitman [30] and Lewis and Whitman [1] to chemical flux (F) across
a gas–liquid interface plane. The emf potential, V, was replaced by chemical
concentration potential such as (Cs Ci), and the result is the well-known resis-
tance-in-series (RIS) law for chemical mass transfer. When applied to the solute at
the sediment–water interface the flux equation is

1 1
F ¼ ðCs Cw Þ= þ ; (9.5)
B D
where B and D (m/s) are the water-side and sediment-side mass transport
coefficients, respectively. Because of chemical-to-water partitioning, with coeffi-
cient Kd (L/kg), the Cs = ws/Kd, where ws (g/kg) is chemical loading on sediment
solids and D = Db r Kd/h, where Db (m2/s) is the biodiffusion transport coefficient,
r (kg/L) particle density, and h (m) bed depth. This result has been verified using
field data [29, 31].
Resistance-in-series is applicable only to potential-driven processes. It is often
time referred to as Ohm’s law [32, 33] and has become a shortcut modeling
approach. Misapplication occurs, for example, when media velocity is treated as
a potential-type transport coefficient and likewise added in the resistance fashion to
obtain the overall resistance. It has been misapplied for atmospheric deposition of
gases and particles where both potential and convective processes resistances are
summed [34, 35]. Hybrid applications that involve converting the chemical
flux water column
B S R
Interface
compartment
D h
sediment bed
Fig. 9.6 The essence of the interface compartment. The bubble indicates the location of the IC. It
is a plane surface so the line depicts one edge. The large arrow to the left depicts the chemical flux
and direction. Transport begins in the sediment bed, crosses the interface, and emerges into the
water column. This net flux is a result of four individual processes. The B process is solute
transport through the water-side benthic boundary layer. The S process is particle deposition (i.e.,
settling) onto the bed surface and R is particle resuspension from the bed surface. On the sediment
side, the D denotes a diffusive process across a distance h. This illustration is used in deriving the
IC model equation (i.e., Eq. 9.8)
potential resistance to an equivalent media convective type for use with the
adjoining bulk-phase concentrations appear in multimedia compartment (MMC)
models [33, 36] for the sediment–water interface. As presented in the next section,
a return to first principles embodied in the Ohm–Kirchhoff and Lewis–Whitman
laws will mend the problem and produce a unified theoretical construct that
accommodates both types individually and in unison.
Environmental models contain multiple phases so interfaces are necessary. The
sediment–water interface is depicted as an idealized concept in Fig. 9.5. It is an
imaginary plane within the complex transition zones separating bulk water phase
from bulk sediment phase as depicted in the Figs. 9.1 to 9.4. The “interfacial
compartment” is defined based on the following assumptions and concepts: (a) it
is a two-dimensional surface containing no mass that separates the adjoining bulk
phases, (b) the chemical flux direction is perpendicular to the surface plane, (c) the
net entering and departing fluxes are equal, (d) a hypothetical aqueous chemical
concentrations with mass per volume units (mg/m3) is assumed, and (e) because
solid (i.e., particle) phases exist on one side, chemical equilibrium is assumed to
exist at the interface plane for estimating the equivalent aqueous concentration.
A mass balance on the interface plane is performed. Because of assumption,
a steady-state equation results and yields a simple algebraic relationship for the
interface compartment concentration. It in turn yields a single flux equation
containing the individual process mass transfer coefficients and the bulk chemical
concentrations in the adjoining media compartments.
Although there are many individual processes occurring in the region of the
interface, for demonstration purpose, only four will be used in the following
derivation of the interfacial compartment model. This approach simplifies the
mathematics of the theoretical procedure while maintaining the essence of the
concept. The combination of chemical potential–type and media convective–type
transport mechanisms used is shown in Fig. 9.6. The double-tipped arrows on either
side of the interface are the chemical potential type. They represent respectively the
solute transport across the benthic boundary layer with coefficient B (m/s) and
diffusive transport within the upper sediment layers with coefficient D/h (m/s),
where the layer thickness is h (m). The single-tipped arrows depict the media
convective–type flux equations. They represent particle deposition or settling
with coefficient S (m/s) and particle resuspension with transport coefficient R (m/
s). The assumed chemical movement pathway and direction of flux F (kg/m2/s) is
depicted by the large arrow. The sediment bed porewater concentration, CS (kg/m3),
in equilibrium with the bed load fraction, wa (mg/kg), is at position h, and the bulk
water column concentration is CW (kg/m3) and is located at the edge of the benthic
boundary layer. The overall chemical potential driving force is the concentration
difference, CS CW.
A steady-state Lavoisier mass balance for the interface compartment requires
that the flux from the sediment-side to the interface, Fsi, equals that departing the
interface on the water-side, Fiw. It is
Fsi ¼ ðD=hÞðCs Ci Þ ¼ BðCi Cw Þ þ RCi SCw ¼ Fiw (9.6)
This result yields the concentration in the interface compartment:
Ci ¼ ððD=hÞCs þ ðS þ BÞCw Þ=ðD=h þ B þ RÞ (9.7)
Combining the two equations yields flux between the compartments in terms of
the bulk-phase concentrations:
Cs ð1 þ R=BÞ CW ð1 þ S=BÞ
FIC ¼ (9.8)
h=D þ 1=B þ Rh=BD
This result is the interface compartment model (IC) flux; it is consistent with the
traditional potential flux in that it takes the form of Eq. 9.5 when the convective
parameters S and R are set to zero. It is the opinion of the authors that the above
procedure is the correct one. However, alternative flux relationships have been
proposed for the across-interface flux based on various assumptions and
methodologies. Two commonly used approaches appear below.
Invoking the RIS concept directly by mimicking the form of Equation 9.5 is one
approach used [32]. In its derivation, the water-side conductances are summed and
then inverted to obtain the overall water-side resistance which is then added to the
sediment-side resistance. This procedure yields
Cs Cw
FRIS ¼ (9.9)
h=D þ 1=ðB þ R SÞ
for the RIS flux equation. The multimedia compartment or box model approach is
a hybrid [33]. Its derivation starts by decomposing chemical potential flux into two
individual convective-type flux components. It then uses the Lavoisier mass balance
for summing the individual fluxes but assumes all are driven by the bulk compart-
ment concentrations. The multimedia compartment (MMC) flux equation is
Cs Cw
FMC ¼ þ ðRCs SCw Þ (9.10)
h=D þ 1=B
The flux equation for the RIS and MMC models reduce to Eq. 9.5 when S and
R are set to zero.
At this juncture, it is clear that three very different algebraic algorithms based on
as many approaches are available for estimating the flux across the sediment–water
interface. In all cases, the fluxes are linear in relation to the bulk media
concentrations and contain the appropriate conductance. Only in the case of the
RIS model must the bulk media concentrations be equal for a zero flux. For both the
IC model and the MMC model equations, a simple algebraic proof shows that
positive, nonzero bulk media concentrations can yield a zero flux. This is a more
realistic outcome because in nature it is possible to have a situation where a mix of
conductance produces a zero net flux and the bulk media concentration is unequal.
Numerical simulations using the theoretical models in flux calculations provide
a realistic and quantitative means demonstrating these and other outcomes. The
methods used and the results of the numerical flux calculations are presented and
discussed in the next section.
Simulation Methods and Results
Although only four individual transport processes were used in the development of
the three theoretical models presented in the previous section, nine individual
transport processes will be used in the numerical simulation. This is done in order
to realistically mimic and highlight the most significant process typically present in
the sediment–water interface region. The four on the water-side include solute
transport in the benthic boundary layer, particle resuspension from the bed surface,
particle deposition from the water column onto the surface, and colloid Brownian
transport through the benthic boundary layer. The five on the bed-side are colloid
Brownian diffusion in the porous bed, Darcian water advection into and out the bed,
solute molecular diffusion in porewater, particle biodiffusion, and porewater
biodiffusion. Altogether, there are five chemical potential–type flux expressions
and five media advection type. The types and categories of the processes plus the
base case numerical values of the parameters are summarized in Table 9.1. The top
four lines represent the water-side transport coefficients and the remaining six
represent the bed-side transport coefficients. The large numerical difference in the
PCB versus BZ transport coefficients are due to partitioning for the particle-
associated processes. For details on how the bed and water column transport
Table 9.1 Transport coefficients (m/day)

Name, location, and typea Benzene PCB
Solute, water-side mass transfer coefficient, cp. 0.32 0.30
Particle resuspension, ma. 5.72E-5 1.24
Particle deposition, ma. 2.0E-9 4.32E-5
Colloid, water-side, ma. 2.31E-7 4.99E-3
Colloid, bed-side, cp. 2.31E-4 2.31E-4
Darcian velocity into bed (), ma. 4.35E-4 3.99E-4
Darcian velocity from bed (+), ma. 1.18E-3 1.08E-3
Solute diffusion in bed, cp. 7.48E-4 6.85E-4
Particle biodiffusion, cp. 1.14E-5 0.25
Porewater biodiffusion, cp. 2.0E-4 2.0E-4
a
cp chemical potential, ma media advection–type flux
coefficients are calculated, see Chapters 10 through 13 in the Handbook of Chemi-

cal Mass Transport in the Environment [37].
Data available in the literature on several North American rivers and lakes with
bed sediment and water column contaminated with organic chemicals were used.
Several studies [27, 29, 38–41] provide the necessary physical, chemical, and
biological data and information needed for estimating the bed and water column
transport coefficients used in the calculations. Typical bed and water column
characteristics at 25 C and 3 m water column depth were used. These
characteristics were bed porosity, 75%; bulk density, 572 kg/m3; sediment layer
thickness for active transport, 0.05 m; fraction organic carbon in bed solids, 50 g/
kg; dissolved organic carbon (DOC) in the porewater, 50 g/m3; particle biodiffusion
coefficient, 2E-6 m2/day; porewater biodiffusion coefficient, 2E-5 m2/day; colloid
Brownian diffusivity, 1.61E-5 m2/day; Darcian water convection into bed, 4E-4 m/
day, and out, 1.1E-3 m/day; Peclet number = 1; water column suspended particle
concentration, 0.005 kg/m3; particle deposition velocity, 4E-4 m/day; particle resuspen-
sion velocity, 1E-4 m/day; and colloid benthic boundary layer transport coefficient,
2.3E-4 m/day.
The porewater chemical concentrations are separated into dissolved and particle-
bound DOC fractions and the fractions transported separately, the dissolved as
solute molecular diffusion and the DOC as Brownian particle diffusion. The
chemical equilibrium phase distribution partition coefficient is used to relate the
solute and DOC-bound concentrations. For characterizing the physical properties of
the bed, the New Bedford Harbor estuary site was used [41]. The chemical
2,4,2’,4’-tetrachlorobiphenyl was used as the PCB. Its partition coefficient was
21.6 m3/kg and that used for benzene (BZ) was 0.001 m3/kg. The tabulated
molecular diffusivity in water for each was used [14]. For the benthic boundary
layer, solute transport coefficients were based on those from the Hudson River; they
were 0.32 and 0.30 m/day for BZ and PCB, respectively. These two chemicals
represent the extremes of hydrophobic properties typically encountered in
contaminated bed sediments. In addition, they also represent the extremes of
numerous soluble and particle-phase geochemicals.
Table 9.2 PCB flux (g/m2∙d) increasingly active processes

Active processes
Cs =.065, Cw =.020 Ci(g/m3) IC RIS MMC
1. Mol. diff., bed porosity 0.1% .020 4.3E-9 4.3E-9 5.0E-8
2. Mol. diff., bed 78% porosity .021 3.1E-5 3.1E-5 3.18-5
3. Water advection into bed, Pe = 10 .0196 1.3E-4 2.8E-3 4.1E-4
4. Mild resuspension, Pe = +10, DOC on. .0195 4.5E-4 3.3E-4 2.4E-3
5. Aggressive resuspension .00418 4.5E-4 3.4E-4 8.1E-2
6. “Storm event” resuspension .00013 4.5E-4 3.4E-4 3.2
7. Mild resusp., mild in-bed biodiff. .0468 9.4E-3 9.0E-3 1.1E-2
8. Mild resusp., aggressive biodiff. .0648 1.5E-2 1.5E-2 1.6E-2
9. “Storm” resusp., aggressive biodiff. .0413 2.1 1.4 3.2
The data appearing above were used with the appropriate algorithms and
formulations to estimate the numerical values of the nine transport coefficients
[14]. Water (porewater in the bed and in the column above) is the continuous phase
across the interface. The transport coefficients use chemical concentrations in water
for flux calculations with both the chemical potential–type and media
advection–type equations. A summary of typical numerical values of the nine
transport coefficients appear in Table 9.1 for BZ and PCB. Several of the base-
case transport parameters were perturbed in doing numerical simulations to cover
their expected range of variation.
The results of the first numerical study appear in Table 9.2. Calculated fluxes for
the PCB using the three theoretical models, the interface compartment (IC) model,
resistance-in-series (RIS) model, and multimedia compartment (MMC) model,
appear. In addition, the IC model interface concentration, Ci (g/m3), is given.
This simulation uses a porewater concentration of 0.065 g/m3 and water column
0.020 for a chemical potential gradient driving the PCB flux from the bed to the
water column for positive flux values with unit g/m2∙d. The negative values denote
fluxes directed into the bed. The first numerical study was to assess the role of
transport aggressiveness or intensity on the flux.
The first three simulations (i.e., 1, 2, and 3) represent molecular diffusion and in-
bed water advection. The flux varies from low positive to negative. Low bed
porosity will produce a low flux; all models have essentially the same values for
simulation 2. Porewater advection in the opposite direction is sufficient to reverse
flux direction as shown in simulation 3. This reverse in flux behavior is present in all
three models, but flux numbers are different for each. This is expected because the
algebraic forms are different (see Eqs. 9.8, 9.9, and 9.10).
These IC model simulations reveal an interconnection between the media
convective–type erosion process on the water-side and the chemical
potential–type diffusive process in the bed (see simulations 4, 5, and 6). In the
absence of biodiffusion, which is the case for these three simulations, the
flux for the IC and RIS models remains unchanged with increasing particle
resuspension aggressiveness. This occurs because the in-bed transport
processes are slow; they are combined porewater advection and molecular
diffusion. The resistance-in-series law is operating properly in the IC and RIS
models, where the slowest process controls. However, an equivalent resistance-
in-series functionality is absent in the MMC model, and the flux tracks particle
resuspension directly, resulting in substantially higher fluxes. The generally
higher MMC model flux values are due to the porewater solute concentration
because it, rather than the interface concentration, is the driver and is always
numerically larger.
More interesting process interconnections are revealed in simulations 7, 8, and 9.
In these cases, in-bed particle and porewater biodiffusion are active as well as
particle resuspension. The flux increases for both the IC and RIS model simulations
as they track the level of biodiffusion; however, the MMC model does not change
much because it is already high. This behavior by the IC and RIS models shows that
an active bed-side process must be present to provide chemical mass readily
available and in the upmost layer in order for erosion to be an effective transport
process. The rapid biodiffusion provides the mass while the slower molecular
diffusion and water advection processes cannot. Finally, all three models give high
and approximately equal flux for the most aggressive resuspension and biodiffusion
transport coefficients (see simulation 9).
To summarize, the simulation study starts with mild in-bed passive molecular
diffusion and a very low-porosity sediment layer. The flux is very low. As one
moves down the table, level of transport aggressiveness increases. The final one is
for aggressive particle biodiffusion and hydraulic flows that result in aggressive
particle resuspension. The flux is very high in this case. The variation in flux from
low and high is approximately 109. The flux numbers for the IC and RIS models are
similar in magnitude while the MMC is consistently higher and, in some cases,
much higher. Its flux behavior seems to track the particle resuspension process in
aggressiveness. Clearly, Eq. 9.10 supports this behavior.
The flux behavior of BZ is somewhat different; no tabular data are provided. The
behavior for simulations 1, 2, and 3 are similar to the PCB ones. Low porosity
yields low flux, and into bed water advection can reverse the weak diffusion-driven
flux. Being less hydrophobic, BZ displays limited sorption to their surfaces. For this
reason the in-bed particle transport is not a significant chemical mobilization
process for BZ. For simulations 2 through 9, all three models give essentially the
same numerical flux values. All models reflect no particle process dependence,
and all display only chemical potential flux–driven behavior patterns. Theoreti-
cally, in the absence of particle processes, all three models are equal and become
identical to Eq. 9.5.
The second numerical study was on the in-bed concentration gradient difference
polarity and the effect on the direction of the flux. Equations 9.8, 9.9, and 9.10 show
each model has a different mathematical dependence on the porewater and water
column concentration. In the above simulations, the concentration difference was
set to simulate chemical transport from the bed, so for PCB, the concentration
differences as Cs were =.065 and 0.02 g/m3, and for BZ, they were 20 and 2.0 g/m3.
So as to further test the models for realistic flux behavior characteristics, the
+.10
+.09
+.08 Multimedia
Compartment
+.07
+.06
+.05
F (g/m2*d)
+.04
Interface
+.03 Compartment
+.02
+.01
0
Resistance-in-series
−.01
−.02
−.03
0 .02 .04 .06 .06 .04 .02 .00
Cw (g/m3) Cw = Cs Cs (g/m3)
Fig. 9.7 PCB flux across the sediment–water interface. A positive flux, F, is from the bed to the
water column and negative is from water to the bed. The concentration differences between bed
and water column used in the calculations are displayed on the ordinate. The progression of
concentration difference values was devised to force the chemical potential out the bed, on the left,
and into the bed on the right so as to drive the flux accordingly. At the vertical dotted line, bed and
water concentrations are equal. Only for the RIS model is the flux zero. The three lines depict the
theoretical behavior of the PCB flux for the three models: IC, RIS, and MMC
chemical potential gradient range and the direction were reversed; the flux results
for reversing the gradient direction appear in Figs. 9.7 and 9.8. In each figure, the
flux is on the vertical axis and the imposed concentration difference on the
horizontal axis. Each line in Fig. 9.7 represents a model from top to bottom; they
are MMC, IC, and RIS.
Figure 9.7 presents aspects of the flux behavior. The vertical axis displays the flux.
Positive numbers represent the PCB moving from the bed to the water column.
Negative numbers represent the PCB moving from the water column to the bed.
The zero flux indicates no net chemical movement in either direction. The horizontal
axis displays chemical concentrations in the water column and the porewater. It is an
unusual axis in that it has zero on each end and maximum in the center.
The horizontal axis consists of two sections divided vertically by a dotted line.
+0.04
Multimedia
Compartment
+0.03
Interface
+0.02
Compartment
+0.01
0.0
−0.01
−0.02
Resistance
−0.03 in-series
0 10 20 10 0
Cw (g/m3) Cw = Cs Cs (g/m3)
Fig. 9.8 Benzene flux across the sediment–water interface. A positive flux, F, is from the bed to the
water column and negative is from water into the bed. The concentration differences between bed
and water column used in the calculations are displayed on the ordinate. The progression of
concentration values were devised to force the chemical potential out the bed, on the left, and
into the bed on the right, so as to drive the flux accordingly. The two lines depict the theoretical
behavior of the BZ flux for the three models: IC, RIS, and MMC. The IC and MMC flux values
overlap
The concentration is a maximum at the position of the dotted line. To the left side of
the vertical line, the porewater concentration in the bed, CS, is held constant at a value
of 0.065 g/m3 and the water column concentration, CW, is varied from 0.0 to 0.065 g/
m3 as shown. Under these conditions, the concentration difference tends to move the
PCB from the bed to the water column and produce a positive flux. It does for all three
models. Only for the RIS model does the flux equal zero for the condition where CS =
CW = 0.065. Due to the particle resuspension process, the PCB flux result for both the
IC and MMC models remain positive and nonzero when moving from the bed to the
water column. In the next simulation, concentration levels in bed porewater versus
the water column concentration will be reversed.
Consider concentration conditions to the right of the dotted line in Fig. 9.7. As
shown on the horizontal axis, the water column concentration, CW, is held constant
at 0.065 g/m3 and the porewater concentration, CS, is reduced from 0.065 down to
0.0 g/m3. The flux for the RIS model starts at zero and goes negative as the PCB
moves into the bed. The other model results display a positive flux at the start and
both trends downward. The MMC model shows a steep decline and reaches zero flux
at CS of about 0.01 g/m3. The IC flux declines to the same value as well, and this
occurs because the imposed concentration gradient encourages PCB movement from
the water column to bed.
The point of the above flux versus concentration study for PCBs was to show
the different behavior patterns produced by the three models. The horizontal
concentration axis was contrived to force the flux to range in magnitude from
positive to negative and therefore cover all possible conditions to be encountered
in nature. The IC model is theory-based, and so it characterizes the correct PCB
flux behavior. Hopefully, future experiments will be able to verify or refute this
behavior. The RIS model predicts lower flux values while those for the MMC
model are much higher. All three models behave according to the algebraic forms
of 9.8, 9.9, and Eqs. 9.10 for the IC, RIS, and MMC, respectively. Both the IC and
the MMC models have a flux inflection; this suggests they are better quantitative
representations of the overall flux process. Crossing the Cw = Cs line and
interchanging concentration represents a switching of processes; it is in the
algebra (see Eqs. 9.8 and 9.10).
A somewhat different behavior occurs for BZ. The results are displayed in
Fig. 9.8 where the graph is constructed similar to that for PCB. The RIS and MMC
flux values are identical and appear as a single diagonal line that starts positive,
goes to zero, and then ends negative. The IC model starts at a slightly lower
positive flux. It decreases then goes through a slope change at CS = CW. This is an
algebra-driven inflection point in the IC model; it occurs at CS = CW = 20 g/m3
(see Eq. 9.8). As CS decreases toward zero, the flux goes into a steep decline with
increasing negative values. It parallels the behavior of the RIS and MC models but
with slightly higher flux values. Beyond the IC model, flux is higher than the
others. The particle transport process is low, and the MMC model takes the
same algebraic form as the RIS. However, the IC model reflects the correct
theoretical approach and displays a very different behavior pattern. Being water
soluble, BZ has minimal particle association. However, the particle processes
enter as ratios (see Eq. 9.8). This has the effect of delaying the zero flux. It occurs
at CS of 11.0 g/m3.
The flux results for the IC versus RIS and MMC models using this specific
simulation with BZ and PCB are different numerically. They also display different
behavior patterns as the imposed concentration gradient condition is changed to
encourage chemical movement from the bed to a condition of chemical movement
into the bed. Presumably, the IC model result is the correct one since it is theory-
based. However, at this juncture, it is a hypothesis in need of testing against
laboratory and field experimental measurements.
Significance for Aquatic Environments
There are numerous individual transport processes on both sides of the interface that
are driven by biological, chemical and geophysical phenomena. Some processes
work in parallel while others work in series, forming a connected network of
processes moving anthropogenic substances and geochemicals across the interface.
The interface compartment (IC) concept and associated mathematical model is
developed and presented as the appropriate theoretical approach for understanding
the overall process and quantifying the resulting net flux. For aquatic researchers, it is
a tool with several uses. It provides a testable hypothesis and a means of interpreting
rate data based from measurements in the laboratory or field. It provides
a mathematical rate equation for individuals making numerical flux estimates.
Finally, the derivation provides a protocol for obtaining one additional model equa-
tion for use by chemical fate modeler’s connecting mobility across the interface that
separates the adjoining bulk sediment and water compartments.
Future Directions
The sediment–water interface is the largest plane surface on Earth. Understanding

and quantifying chemical and particle mobility across this semipermeable interface
is relevant to the work of a broad community of aquatic researchers. The interface
compartment model provides the basic theory for connecting chemical flux across
the sediment–water interface. Correctly quantifying the net chemical, particle, and
aquasol exchange rates across this plane is a key factor for understanding the fate
of numerous natural and anthropogenic substances on Earth and aid in assessing
the ecological significance. The potential impact of the IC model’s further devel-
opment and use in the environmental and geosciences fields may be a key
contributing factor. Life-forms residing on both sides, in the water column or the
surface sediment layers, depend on oxygen and nutrient fluxes. The bed is a source
or a sink of soluble and particulate carbon compounds depending on the
chemodynamics of the specific locale. The bed is a sink for chemical pollutants
entering the aquatic system but, later as conditions improve in the water column, it
becomes the source. These are just a few examples of the types of possible uses
which require theoretically sound and verified science-based tools. As outlined
above, further work is needed on the interface compartment concept before it is
accepted as a unified theory and modeling protocol.
Bibliography
1. Lewis WK, Whitman WG (1924) Principles of gas absorption. Indust Eng Chem
16:1215–1220
2. Rhoads DC, Germano JD (1982) Characterization of benthic processes using sediment profile
imaging: An efficient method of remote ecological monitoring of the seafloor (REMOTS™
System). Mar Ecol Prog Ser 8:115–128
3. Rhoads DC, Germano JD (1986) Interpreting long-term changes in benthic community

structure: a new protocol. Hydrobiologia 142:291–308
4. Santschi P, Hohener P, Benoit G, Brink MB (1990) Chemical processes at the sediment-water
interface. Mar Chem 30:269–315
5. Duursma EK, Smies M (1982) Sediments and transfer at and in the bottom interfacial layer. In:
Kullenberg G (ed) Pollutant transfer and transport in the sea, vol II. CRC Press, Boca Raton,
pp 101–137
6. Krantzberg G (1985) The influence of bioturbation on physical, chemical and biological
parameters in aquatic environment – a review. Environ Pollut (Ser A) 39:99–122
7. DiToro DM (2001) Sediment flux modeling. Wiley, New York
8. Kullenberg G (ed) (1976) Pollutant transfer and transport in the Sea-II. CRC Press, Boca Raton
9. Boudreau BP, Jorgensen BB (eds) (2001) The benthic boundary layer. Oxford University
Press, New York
10. McCave IN (ed) (1976) The benthic boundary layer. Plenum, New York
11. Windom HL, Duce RA (1976) Marine pollutant transport. Lexington, Lexington
12. Tenore KR, Coull BC (eds) (1980) Marine benthic dynamics. University of South Carolina
Press, Columbia
13. Fanning KA, Manheim FT (eds) (1982) The dynamic environment of the ocean floor.
Lexington Books, Lexington
14. Thibodeaux LJ (1996) Environmental chemodynamics. Wiley, New York
15. Lerman A (1979) Geochemical processes water and sediment environments. Wiley, New York
16. Boudreau BP (1997) Diagenetic models and their implementation. Springer, Berlin
17. Schink DR, Guinasso NL Jr (1975) Modeling the influence of bioturbation and other processes
of CaCO3 dissolution at the sea floor. In: Andersen NR, Malahoff A (eds) The fate of fossil fuel
CO2 in the oceans. Plenum, New York, pp 375–399
18. Berner RA (1980) Early diagenesis. Princeton University Press, Princeton
19. Thibodeaux LJ, Matisoff G, Reible DD (2010) Bioturbation and other sorbed-phase transport
processes in surface soils and sediment. In: Thibodeaux LJ, Mackay D (eds) Handbook of
chemical mass transport in the environment. CRC Press, Boca Raton (Chap 13)
20. Thibodeaux LJ, Wolfe JR, Dekker TJ (2010) Advective porewater flux and chemical transport
in bed-sediment. In: Thibodeaux LJ, Mackay D (eds) Handbook of chemical mass transport in
the environment. CRC Press, Boca Raton (Chap 11)
21. Singh VP, Reible DD, Thibodeaux LJ (1988) Mathematical modeling of fine sediment
transport. Hydrol J IAH 11:1–3
22. Lick W (2009) Sediment and contaminant transport in surface waters. CRC Press, Boca Raton
23. Lohmann R, Dachs J (2010) Deposition of dissolved and particle-bound chemicals from
surface ocean. In: Thibodeaux LJ, Mackay D (eds) Handbook of chemical mass transport in
the environment. CRC Press, Boca Raton (Chap 17)
24. Yang CT (2003) Sediment transport. Krieger, Malabar
25. DePinto JV, McCulloch RD, Redder TM, Wolfe JR, Dekker TJ (2010) Deposition and
resuspension of particles and associated chemical transport across the sediment-water inter-
face. In: Thibodeaux LJ, Mackay D (eds) Handbook of chemical mass transport in the
environment. CRC Press, Boca Raton (Chap 10)
26. Reible DD, Valsaraj KT, Thibodeaux LJ (1991) Chemodynamic models for transport of
contaminants from sediment beds. In: Hutzinger O (ed) The handbook of environmental
chemistry, part F, vol 2. Springer, Berlin, pp 186–228
27. Thibodeaux LJ, Reible DD, Valsaraj KT (2002) Non-particle resuspension chemical transport
from stream beds. In: Lipnick RL, Mason RP, Phillips ML, Pittman CU Jr (eds) Chemicals in
the environment, vol 806, ACS symposium series. American Chemical Society, Washington,
DC, pp 130–149
28. Thibodeaux LJ, Valsaraj KT, Reible DD (2001) Bioturbation–driven transport of hydrophobic
organic contaminants from bed sediment. Environ Eng Sci 18:215–223
29. Erickson MJ, Turner CL, Thibodeaux LJ (2005) Field observation and modeling of dissolved
fraction-sediment-water exchange coefficients for PCBs in the Hudson River. Environ Sci
Technol 39:549–555
30. Whitman WG (1923) The two-film theory of gas absorption. Chem Metall Eng 29:146–148
31. Thibodeaux LJ, Bierman VJ (2003) The bioturbation-driven chemical release process. Environ
Sci Technol 1:253A–258A
32. Monteith JL, Unsworth M (1990) Principles of environmental physics. Butterworth-
Heinemann, Oxford
33. Mackay D (2001) Multimedia environmental models. Lewis, Boca Raton
34. Slinn WGN (1978) 4-Wet and dry removal processes. In: NRC (ed) The tropospheric transport
of pollutants and other substances to the oceans. National Academy of Sciences, National
Academy Press, Washington, DC
35. Trapp S, Matthies M (1998) Chemodynamics and environmental modeling. Springer, Berlin
36. Van de Meent D (1993) Simple box: a generic multimedia fate evaluation model. RIVM
Report No. 6727200001. Bilthoven
37. Thibodeaux LJ, Mackay D (eds) (2010) Chapters 10,11,12,13 &17. In: Handbook of chemical
mass transport in the environment. CRC Press, Boca Raton
38. Thoma GJ, Koulermos AC, Valsaraj KT, Reible DD, Thibodeaux LJ (1991) The effect of pore-
water colloids on the transport of hydrophobic organic compounds from bed sediments. In:
Baker RA (ed) Organic substances in water, vol 1, Humics and soils. Lewis, Boca Raton
39. Valsaraj KT, Thibodeaux LJ, Reible DD (1997) A quasi-steady-state pollutant flux methodol-
ogy for determining sediment quality criteria. Environ Toxicol Chem 16:391–396
40. Savant SA, Reible DD, Thibodeaux LJ (1987) Convective transport within stable river
sediments. Water Resour Res 23:1763–1768
41. Thibodeaux LJ, Reible DD, Bosworth WS, Sarapas LC (1990) A theoretical evaluation of the
effectiveness of capping PCB contaminated New Bedford Harbor bed sediments. Final Report.
HSRC. Middelton Library, Louisiana State University, Baton Rouge
Chapter 10
River Fate and Transport
Zhen-Gang Ji
Glossary
Advection The horizontal transport by flows that move patches of

material around but do not significantly distort or dilute
them.
Biodegradation The breakdown of a compound by enzyme-mediated trans-
formation primarily due to bacteria, and to a lesser extent,
fungi.
Dispersion The mixing of water properties in rivers.
Henry’s law A law which states that at a given temperature, the solubility
of a gas is proportional to the pressure of the gas directly
above the water.
Hydrograph A graph showing time variation in flow rate or stage (depth)
of water in a river.
Hydrolysis The reaction of a chemical with water in which splitting of
a molecular bond occurs in the chemical and there is forma-
tion of a new bond with either the hydrogen component (H+)
or the hydroxyl component (OH) of a water molecule.
Manning equation An empirical formulation relating velocity (or flow rate)
depth, slope, and a channel roughness coefficient in a river.
Mineralization The process by which a dissolved organic substance is
converted to dissolved inorganic form.
Z.-G. Ji (*)
Minerals Management Service, 381 Elden Street, Herndon, VA 20170, USA
220 Z.-G. Ji
Nonpoint SOURCE A pollution source that cannot be traced to a specific spot.

Photolysis The transformation of a compound that results directly from
the adsorption of light energy.
Point source A pollution source that comes from a specific identifiable
source such as a pipe.
Residence time The time required by a particle to cross a river reach.
River A naturally flowing waterbody.
Volatilization The process representing a chemical substance entering the
atmosphere by evaporation from water.
Rivers are naturally flowing waterbodies. Small rivers are also called streams or
brooks. Rivers are a watershed’s self-formed gutter system and usually empty into an
ocean, lake, or another river. This chapter describes the characteristics of rivers and the
fate and transport in rivers. The mathematical description of river processes and the
modeling of rivers are also described here.
Rivers are complex and dynamic. A river often acts as a sink for contaminants
discharged along the river, such as effluents from wastewater treatment plants that
discharge nutrients, heavy metals, and/or pathogens into the river. Rivers may also act
as sources of contaminants in the watershed, depending on the time of the year or the
section of the river. The health of a river is directly linked to the health of the surrounding
watershed. The water quality in a river will deteriorate, if the watershed condition
deteriorates. Via rivers, pollutants can travel hundreds or even thousands of kilometers
and cause environmental problems in a waterbody that is located far away from the
sources. The common designated uses of a river include aquatic life support, water
supply, and recreation activities (such as swimming, fishing, and boating).
Introduction
The rivers and their tributaries, normally occupying less than a few percent of the total
drainage basin, are the conduits of the river basin. They are like a gutter system and
transport water, nutrients, sediment, and toxicants downstream (often to an estuary or
a large lake). Compared with lakes and estuaries, the most distinct characteristic of
a river is its natural downstream flow. Lakes typically have much smaller flow
velocities than rivers. Flow velocities in estuaries, though their magnitudes can be
comparable to the ones in rivers, are tidally driven and can be in either direction
(downstream or upstream).
The origins of contaminants can be divided into point and nonpoint sources.
Point source pollution comes from a specific, identifiable source such as a pipe.
10 River Fate and Transport 221
Nonpoint source pollution cannot be traced to a specific spot. Point sources include
wastewater treatment plants, overflows from combined sanitary and storm sewers,
and industry discharges. Nonpoint sources include runoffs from urban, agriculture,
and mining areas. Point and nonpoint sources have caused a wide range of water
quality problems and the deterioration of the ecological state in rivers. Leading
pollutants and stressors in the USA include [1]:
1. Pathogens (bacteria)
2. Siltation
3. Habitat alterations
4. Oxygen-depleting substances
5. Nutrients
6. Thermal modifications
7. Toxic metals
8. Flow alterations
Pathogens are the most common pollutant affecting rivers and streams in the USA.
Pathogen pollution is a major public health problem especially in the use of river water
for water supply and the consumption of fish and shellfish harvested in rivers and
estuaries. Bacteria commonly enter surface waters in inadequately treated sewage,
fecal material from wildlife, and runoff from pastures, feedlots, and urban areas.
Sediment siltation is one of the leading environmental problems in rivers. The filling
of river channels, harbors, and estuaries by sediments brings a high cost to society. The
condition of a river’s watershed greatly affects the amount of sediment delivered into
the river. The sediment sources vary among rivers, and even within a particular river,
from year to year. Extreme events, such as hurricanes, can produce dramatic changes in
the amounts and types of sediments that are delivered into a river. The vulnerability of
a river to sediments and contamination reflects a complex combination of upstream
flows, land use, and land-management practices. The vast majority of river sediments is
discharged during only 10% of the year (36 days), and 90% of the year represents
a very small amount of the sediment load [2]. Low flow rates usually result in net
deposition conditions. High flow rates may cause net erosion in upstream reaches and
net deposition in downstream reaches or in the estuary into which the river flows.
Often water quality is defined in terms of concentrations of the various dissolved
and suspended substances in the water, for example, temperature, salinity, dissolved
oxygen, nutrients, phytoplankton, bacteria, heavy metals, etc. The distribution of these
substances can be calculated by a mathematical model. Based on the principle of
conservation of mass, the concentration change can be represented simply in a one-
dimensional form [3]:

@C @C @ @C
¼ U þ D þSþRþQ (10.1)
@t @x @x @x
where
C = substance concentration
t = time
222 Z.-G. Ji
x = distance
U = advection velocity in x direction
D = mixing and dispersion coefficient
S = sources and sinks due to settling and resuspension
R = reactivity of chemical and biological processes
Q = external loadings to the aquatic system from point and nonpoint sources
The changes of concentration C in Eq. 10.1 are determined by the following:
1. The hydrodynamic processes control the water depth (D), the advection
(represented by the U term), and mixing (represented by the D term).
2. The size and properties of sediment (or particular organic matter) affect the
settling and resuspension (represented by the S term).
3. The chemical and biological reactions of pathogens, toxics, and/or nutrients are
represented by the R term.
4. External loadings from point and nonpoint sources are included by the Q term.
Fate and Decay
Contaminants in rivers include nutrients, organic toxicants, heavy metals, and

pathogens. If no degradation reactions occurred in Nature, every single contaminant
discharged in the past would still be polluting the environment. Fortunately, natural
purification processes dilute, transport, remove, and degrade contaminants. It is
essential to understand the kinetics of reactants and to describe them mathemati-
cally. This section summarizes the fate and decay of contaminants and their
mathematical formulations.
The fate and transport of contaminants are controlled by two factors: their
reactivity and their hydrodynamic transport. Reactivity includes:
1. Chemical processes
2. Biological processes
3. Bio-uptakes
Transport in a river, which will be discussed in the next section, includes three
mass transport processes:
1. Advection of water current
2. Diffusion and turbulent mixing within the water column
3. Deposition and resuspension on the water-sediment bed interface
Mathematical Formulations
How long contaminants remain in a waterbody depends on the nature of the compound.
Most chemicals undergo chemical or biological decay. Some chemicals are conservative
and do not undergo these types of reactions, even though it is very difficult to find a truly
conservative chemical in Nature. The fate and decay of a contaminant represent the
gradual decrease in the amount of a substance in a river, as the result of various sink
processes, including chemical and biological transformation, or dissipation/deposition to
other environmental systems.
Although reaction kinetics in aquatic systems can be described in numerous
ways, the form for a single reactant is generally expressed as:
dC
¼ R ¼ kCm (10.2)
dt
where
m = the order of reaction
k = rate constant of the m-order reaction
In natural waters, the commonly used forms of Eq. 10.2 are with m = 0, 1, and 2.
Zero-order reactions: A zero-order reaction (m = 0) represents irreversible
degradation of a reactant that is independent of the reactant concentration. The
solution to Eq. 10.2 is:
C ¼ C0 kt (10.3)
where C0 = the initial concentration at t = 0. In this case, a plot of concentration

versus time should yield a straight line with a slope of k, as shown in the left panel
of Fig. 10.1. Zero-order reactions have their reaction rates determined by some
factor other than the concentration of the reacting materials.
First-order reactions: First-order reactions (m = 1) have their reaction rates
proportional to the concentration of the reactant and are most commonly used in
describing chemical and biological reactions. For first-order reactions, the solution
to Eq. 10.2 is:
C ¼ C0 ekt (10.4)
Equation 10.4 indicates that for first-order reactions, reactant concentration

decreases exponentially with time. In this case, a plot of logarithm concentration
versus time should yield a straight line with a slope of k, as shown in the middle
panel of Fig. 10.1. Most of the reactions found in the environment can be conve-
niently expressed by a first-order approximation without much error. Examples of
first-order reactions include biochemical oxygen demand in surface waters, death
and respiration rates for bacteria, and production reaction of algae.
Second-order reactions: For second-order reactions (m = 2), the solution to
Eq. 10.2 is:
1 1
¼ þ kt (10.5)
C C0
224 Z.-G. Ji
Zero-order First-order Second-order
t
C= lnC +k
C– =ln 1/C
0
C– = k
C, 1/C
0 k
C, lnC
t
k
0 kt
k 1/C
C
C= C=C
C 0 /(1+
0 e −k kt)
t
t t t
Fig. 10.1 Left panel: concentration versus time for zero-order reaction. Middle panel: concentra-
tion and logarithm concentration versus time for first-order reaction. Right panel: concentration
and inverse concentration versus time for second-order reaction
Therefore, if a reaction is indeed second-order, a plot of inverse concentration of

C (1/C) with time should yield a straight line with a slope of k (the right panel of
Fig. 10.1). Equation 10.5 can also be expressed as
C0
C¼ (10.6)
1 þ kC0 t
which reveals that, similar to the first-order reaction, the resulting concentration of
a second-order reaction also decreases and approaches zero as time increases.
Processes Affecting Fate and Decay
The fate and decay of contaminants can result from physical, chemical, and/or
biological reactions. In addition to sorption and desorption, processes that can
significantly affect the fate and decay processes include:
1. Mineralization and decomposition
2. Hydrolysis
3. Photolysis
4. Biodegradation
5. Bioconcentration
6. Volatilization
Most decay processes are expressed as first-order reactions. The first-order decay
coefficients for individual processes are additive and can be linearly superimposed
to form a net decay coefficient:
kd ¼ km þ kh þ kp þ kbd þ kbc þ kv (10.7)
where
kd = net decay coefficient
km = mineralization coefficient
kh = hydrolysis coefficient
kp = photolysis coefficient
kbd = biodegradation coefficient
kbc = bioconcentration coefficient
kv = volatilization coefficient
In modeling studies, either the net degradation coefficient or the individual
coefficients can be specified.
Mineralization and Decomposition
Mineralization is the process by which a dissolved organic substance is

converted to dissolved inorganic form. Mineralization makes nutrients, such
as nitrogen and phosphorus, available for a fresh cycle of plant growth.
Bacteria decompose organic material to obtain energy for growth. Plant
residue is broken down into glucose that is then converted to energy:
energy released
C6 H12 O6 þ O2 ! CO2 þ H2 O (10.8)
In water quality models, the term “mineralization” often represents the

process by which dissolved organic matter is converted to dissolved inorganic
form, and thus includes both heterotrophic respiration of dissolved organic
carbon and mineralization of dissolved organic phosphorus and nitrogen.
Hydrolysis
Hydrolysis is the reaction of a chemical with water, in which splitting of

a molecular bond occurs in the chemical and there is formation of a new bond
with either the hydrogen component (H+) or the hydroxyl component (OH) of
a water molecule. This involves ionization of the water as well as splitting of the
compound hydrolyzed:
RX þ H2 O ! ROH þ HX (10.9)
Essentially, water enters a polar location on a molecule and inserts itself, with an
H+ component going to one part of the parent molecule and an OH component going
to the other. The two components then separate. The concentration of hydrogen and
hydroxide ions, and therefore pH, is often an important factor in assessing the rate of
a hydrolysis reaction. Hydrolysis is a major pathway for the degradation of many toxic
organics.
226 Z.-G. Ji
Photolysis
Photolysis is the transformation of a compound that results directly from the

adsorption of light energy. Compounds that absorb sunlight may gain sufficient
energy to initiate a chemical reaction. Some of these photochemical reactions result
in the decomposition or transformation of a substance.
The energy of light varies inversely with its wavelength. Longwave light lacks
sufficient energy to break chemical bonds. Short wave light (x-rays and gamma
rays) is very destructive. Fortunately for life on earth, this type of radiation largely
is removed by the upper atmosphere. Light near the visible spectrum reaches the
earth’s surface and can break the bonds of many organic compounds, which can be
important in the decay of organic chemicals in a water system.
The basic characteristics of photolysis are:
1. Photolysis has two types of energy absorption: direct photolysis and indirect
photolysis. The direct photolysis is the result of direct absorption of sunlight by
the toxic chemical molecule. Indirect photolysis is the result of energy transfer to
the toxic chemical from some other molecule that has absorbed the sunlight.
2. Photolysis is the destruction of a compound activated by the light energy and is an
irreversible decay process.
3. Products of photolysis may remain toxic and the photolysis process does not
necessarily lead to detoxification of the system.
4. The photolysis coefficient in Eq. 10.7 is usually a function of the quantity and
wavelength distribution of incident light, the light adsorption characteristics of
the compound, and the efficiency at which absorbed light produces a chemical
reaction.
Biodegradation
Biodegradation is the breakdown of a compound by enzyme-mediated transforma-

tion, primarily due to bacteria, and to a lesser extent, fungi. Although these types of
microbial transformations can detoxify and mineralize toxics, they can also activate
potential toxics. The rate of biodegradation can be very rapid, which means that
biodegradation is often one of the most important transformation processes
in rivers.
Even though the biodegradation process is largely mediated by bacteria, the
growth kinetics of the bacteria is complicated and is not well understood. As
a result, toxic models often assume constant decay rates rather than modeling the
bacteria activity directly. The first-order decay rate is commonly used. Biodegra-
dation rate is influenced by water temperature and can be represented by an
Arrhenius function:
kb ¼ kb20 yðT20Þ (10.10)

where
kb = biodegradation rate
kb20 = biodegradation rate at 20 C
T = water temperature in C
y = temperature correction factor
The effect of the Arrhenius function is that a higher temperature will cause
a faster chemical reaction rate. It gives a quantitative relationship between the
reaction rate and its temperature.
Biodegradation rate is also related to the contaminant concentration and can be
expressed by a typical Michaelis–Menten formulation:
c
kb ¼ kbmax (10.11)
c þ c1=2
where
kbmax = the maximum biodegradation rate
c = the contaminant concentration
c1/2 = half saturation (Michaelis) constant.
The combination of the above two formulations yields
c
kb ¼ kmax yðT20Þ (10.12)
c þ c1=2
where kmax = maximum decay rate due to biodegradation. Equation 10.12 combines
the effects of contaminant concentration and water temperature on the biodegrada-
tion process.
Volatilization
Volatilization represents a chemical substance entering the atmosphere by evaporation

from water. Volatilization is often treated as an irreversible decay process, because of
its mathematical similarities to these decay processes. However, volatilization is
actually a reversible transfer, in which the dissolved concentration in water attempts
to equilibrate with the gas phase concentration in the overlying atmosphere. Equilib-
rium occurs when the partial pressure exerted by the chemical in water equals the
partial pressure of the chemical in the atmosphere.
Henry’s law states that, at a given temperature, the solubility of a gas is
proportional to the pressure of the gas directly above the water. Volatilization is
often treated similarly to surface oxygen exchange, where the volatilization flux is
proportional to the difference between the chemical concentration in water and the
saturation concentration, as:
228 Z.-G. Ji
Fv ¼ kv ðcw cws Þ (10.13)
where
Fv = volatilization flux
kv = transfer rate
cw = dissolved concentration of the chemical in water
cws = saturation dissolved concentration of the chemical in water
Equation 10.13 indicates that the chemical enters the water when the chemical in
the water is unsaturated (cw < cws) and the chemical leaves (volatizes from) the
water when the chemical in the water is oversaturated (cw > cws). The saturation
dissolved concentration is dependent upon the atmospheric partial pressure and
Henry’s law constant for the chemical. The transfer rate, kv, depends on the
properties of the chemical as well as the characteristics of the waterbody and the
atmosphere, including the molecular diffusion coefficient of the chemical in the
water and in the atmosphere, the temperature, the wind speed, the current velocity,
and the water depth.
Transport in a River
Rivers have distinct hydrodynamic characteristics that are different from those of
lakes or estuaries. This section focuses on the following:
1. River flow and the Manning equation
2. Advection and dispersion processes in rivers
River Flow and the Manning Equation
The flow rate of a river is the volume of water that passes a cross section of the river
in a unit of time, which is usually expressed in cubic meters per second (cms) or
cubic feet per second (cfs) and is calculated as:
Q¼AV (10.14)
where
Q = Flow rate in cms or cfs
A = Area through which the water is flowing in m2 or ft2
V = Average velocity in the downstream direction in m/s or ft/s
The river flow can generally be separated into two components:
1. Base flow
2. Storm flow
Base flow is composed largely of groundwater effluent and sustains river flow
during dry weather periods. Storm flow is from the runoff during or shortly after
a precipitation event. The water from base flow is the precipitation that percolates into
the ground and flows slowly through a long path before reaching the river, whereas the
water from storm flow is the precipitation that reaches the river shortly after precipita-
tion through runoff. In addition to base flow from the groundwater and the storm flow
from the runoff, point sources, such as wastewater treatment plant discharges and
tributaries to the river, also contribute to a river flow.
A hydrograph is a graph showing time variation in flow rate or stage (depth) of
water in a river. As sketched in Fig. 10.2, the river flow is composed of the storm
flow and the base flow. After the beginning of a rainfall, the storm flow from runoff
starts to increase and reaches its peak some time after the peak rainfall. There is
a time lag between the two peaks. The rising limb is the portion of the hydrograph to
the left of the peak of the storm flow, which shows how long the river takes to reach
its peak flow rate after a rainfall event. The receding limb is the portion of the
hydrograph to the right of the peak, which shows how long the river takes to return
to the base flow.
In addition to flood events, low flow conditions are also important characteristics of
a river. When there is no precipitation contributing to the storm flow, and the base flow
from groundwater is low, the river experiences low flow conditions. Low flow results
in less water available for dilution of pollutants from point sources, causing high
pollutant concentrations in the river. Therefore, point source discharges during low
flow conditions have the most significant impact on the river’s water quality, since the
discharge may constitute a larger percentage of river flow. For instance, wastewater
discharges to the Blackstone River can account for up to 80% of the total river flow in
summer [4].
A hydrodynamic model based on momentum and continuity equations is often
used to calculate flow velocity, flow rate, and water depth in a waterbody. A simpler
approach to calculate these parameters is to use the Manning equation, which is an
empirical formulation relating velocity (or flow rate), depth, slope, and a channel
roughness coefficient in a river. The Manning equation was derived by curve-fitting
data measured in rivers and channels. The equation is:
2 1
Q R3 S2
V¼ ¼ (10.15)
A n
where
V = mean flow velocity in m/s
Q = flow rate in m3/s
A = cross-sectional area in m2
R = hydraulic radius in m
S = slope of the channel bed in m/m
n = Manning roughness coefficient
230 Z.-G. Ji
Fig. 10.2 A storm Rainfall Time lag

hydrograph of a river
Rising limb
St
or
m
Flow rate
flo Receding limb
w
Base
flow
Time
The hydraulic radius is defined as:
A
R¼ (10.16)
P
where P is the wetted perimeter in m, which is the length of contact of the water
with the channel in m, measured in a direction normal to the flow. The Manning
roughness coefficient, n, represents the channel roughness that contributes to the
dissipation of flow energy. Table 10.1 shows a range of n values for various
channels and rivers.
Originally developed in the 1880s, the Manning equation is still widely used in
hydraulic calculations with reasonable accuracy today. In hydrodynamic modeling, the
Manning equation may serve the purpose of giving a quick estimation of flow
conditions in a river. However, the Manning equation is an empirical formulation
that may not reflect actual conditions of a river.
Advection and Dispersion in Rivers
Advection refers to horizontal transport by flows that move patches of material

around but do not significantly distort or dilute them. In rivers, advection often
represents the primary transport process of pollutant in the longitudinal direction.
Dispersion is the mixing of water properties. In rivers, a prominent feature is the
longitudinal dispersion: the transport and spreading of pollutants downstream from
a point source. When a tracer is released into a river, two distinct processes control
the tracer transport:
1. Flow advection carries the tracer away from the releasing point.
2. Turbulence dispersion spreads out and dilutes the tracer concentration.
Table 10.1 Values of the Manning roughness coefficient, n, for various channels and rivers [5]
Manning roughness
Type of channel coefficient (n)
Smooth concrete 0.012
Ordinary concrete lining 0.013
Earth channels in best condition 0.017
Straight unlined earth canals in good condition 0.020
Natural rivers and canals 0.020–0.035
Mountain streams with rocky beds and rivers 0.040–0.050
with variable sections and some vegetation along banks
Alluvial channels without vegetation 0.011–0.035
Water depth
Mean velocity
Downstream
Channel bed
Fig. 10.3 Velocity vertical profile in a channel
Mathematically, the above two processes are represented by the first and second
terms on the right-hand side of Eq. 10.1, respectively. Advection results in the
pollutant’s moving downstream, while longitudinal mixing leads to spreading or
smearing in the longitudinal dimension. Lateral and vertical mixing processes deter-
mine how long it takes for a pollutant to be completely mixed across a river. The
dominant transport process in rivers is the advection due to river flow. Flow velocity
controls a river’s residence time, the time required by a particle to cross a river reach.
The dispersion process in rivers is often less important in the transport of pollutants.
The effect of dispersion may be ignored in analyzing a continuous pollutant load to
a river. Fig. 10.3 is a velocity vertical profile in a channel. In small rivers, however, the
turbulence generated by bed friction is strong, and the depth is generally small,
resulting in rivers that are often well mixed vertically.
To illustrate the longitudinal dispersion in a river, an idealized dye release experi-
ment is shown in Fig. 10.4, in which Panel A gives the plain view of the dye transport
in the river and Panel B presents the lateral-averaged dye concentration along the river.
In the river, a line source of constant concentration is instantaneously released at time
t = 0, and the longitudinal velocity has parabolic variation across the river. As shown in
Panel A, the advection process transports the dye downstream, and the dispersion
process spreads the dye and reduces the maximum concentration. Dye travels down-
stream faster in the middle of the river than near the banks. As a result, the line source
released at t = 0 becomes approximately a parabolic shape at t = t1 and t = t2. The
concentration profiles at t1 and t2 in Panel A also reflect the random fluctuations of
turbulence activities in the river. Because of variations in flow velocity across the
232 Z.-G. Ji
Velocity profile Line source release
Left bank
Width
Spreading by dispersion
Downstream
Right bank
a
Concentration at t = 0, t = t1, and t = t2
Lateral-averaged
concentration
b Downstream distance
Fig. 10.4 Advection and dispersion processes in a river. Panel A gives the plain view of dye
transport in the river. Panel B presents the lateral-averaged dye concentration along the river
river, dye spreads both along and across the river by dispersion. The laterally averaged
dye concentration in Panel B also indicates that the velocity shear and turbulent
dispersion contribute to the concentration spreading along the river.
Impacts of River Flow on Water Quality
Water quality processes can be highly dependent on river flow conditions. The time
that a pollutant remains within a section of a river is called residence time. The flow
velocity and the length of the river section determine the residence time. River flow
affects water quality in a river in several ways:
1. Dilution. A large volume of flow dilutes concentrations of pollutants that are
discharged into the river.
2. Residence time. High flow velocity reduces the residence time and affects the
amount of material that can be produced or degraded in the river section.
3. Mixing. High flow velocity increases mixing in the river, enhances the assimila-
tive capability of the river, and reduces pollutant concentration gradients.
4. Erosion. High flow can erode bed material and destabilize the benthic
environment.
The impact of pollutant loadings to a river is largely determined by the magnitudes

of the loadings and the flow rate. Rapid transport of pollutants by high flow results in
a short residence time and often causes minimal water quality problems. Conversely,
slow transport of pollutants by low flow results in a long residence time and can lead to
water quality problems, such as oxygen depletion and eutrophication. Channel alter-
ation and watershed disturbance can lead to abnormally high flow rates for a given
amount of rain and amplify the impact of floods. Watershed disturbance can also
increase sedimentation and harm aquatic biota in a river.
In temperate regions, seasonally high flow typically occurs during the periods
of snowmelt in early spring and spring rains, whereas seasonally low flow
normally occurs in summer and early fall. The river flow affects the concentration
and distribution of water quality variables. Generally, point sources have a larger
impact on a river during low flow (dry weather) conditions due to less water
diluting the pollutants. Low DO concentrations and high algal growth in a river
often occur during low flow periods and hot weather conditions. The combination
of low flow, minimum dilution, and high temperature often makes summer and
early fall the critical periods for evaluating the impact of point sources (such as
wastewater treatment plants).
In contrast, nonpoint sources can bring large amounts of pollutants from the
watershed into a river during high flow (wet weather) conditions. It is important to
examine both point and nonpoint sources in both high and low flow conditions.
Point sources of nutrients often cause algal blooms in rivers during low flow
conditions, while nonpoint sources may increase nutrient concentrations and tur-
bidity following periods of wet weather events. Municipal discharges, agriculture
runoff, and urban runoff are among the most common sources of impairment to
rivers.
In the study of the Blackstone River, for example, Ji et al. [2] reported that
a discharge from a wastewater treatment plant was the dominant point source of
contaminants and had significant impact on the sediment contamination in the river.
However, this point source alone is still insufficient to account for the total metal
concentrations in the river. Nonpoint sources and the processes of sediment depo-
sition and resuspension are also important factors that control the concentrations of
sediment and toxic metals.
Dissolved oxygen is essential to river ecosystems. Processes controlling DO
spatial distribution in a river include:
1. Oxidation of the biochemical oxygen demand (BOD): BOD is used to represent
all sinks of dissolved oxygen, such as the oxidation of carbonaceous and
nitrogenous organic matter, the benthic oxygen demand, and the oxygen utilized
by algal respiration.
2. Reaeration of DO from the atmosphere: In addition to atmospheric reaeration,
DO produced by photosynthesis and DO contained in incoming flows are also
major oxygen sources.
3. Transport due to the river flow: Advection and diffusion processes enhance DO
mixing and reaeration within a river.
234 Z.-G. Ji
The pioneering work by Streeter and Phelps [6], who developed the first water
quality model to describe the oxygen depletion in the Ohio River, is useful for
understanding DO processes in a river. It can be described in a first-order reaction
equation:
dC
U ¼ kd B þ ka ðCs CÞ (10.17)
dx
where x = distance
U = advection velocity in x direction
C = DO concentration
B = BOD concentration
Cs = saturated dissolved oxygen concentration
kd = deoxygenation rate constant of BOD
ka = first-order reaeration rate constant of DO
By assuming that BOD has a first-order degradation reaction with a decay rate
constant of kr, the solution to Eq. 10.17 is the famous Streeter-Phelps equation:
kd L0 kr x=U
C ¼ Cs e eka x=U
ka kr (10.18)
ðCs C0 Þeka x=U
A schematic representation of the Streeter-Phelps equation is shown in Fig. 10.5,

describing a DO sag curve in a river. The DO sag curve gives DO longitudinal
variation as the result of oxygen depletion and recovery, after a BOD load is
discharged into a receiving river. Between the discharge point (x = 0) and the
critical distance (x = xc), oxidation exceeds reaeration (i.e., kd B > ka ðCs CÞ in
Eq. 10.17) because of high BOD concentrations and a small DO deficit (= CsC).
Oxygen in the river is consumed faster than it is resupplied. The DO concentration
decreases to a minimum Cmin at a critical distance xc (or critical time tc = xc/U).
This position is the critical location where the lowest DO concentration occurs, and
the oxidation rate and reaeration rate are equal. After passing the critical location,
reaeration exceeds oxidation (i.e., kd B<ka ðCs CÞ in Eq. 10.17) because of
a low BOD concentration and a high DO deficit. Thus, oxygen in a river increases
gradually. Further downstream, the rate of supply exceeds the utilization rate,
resulting in a full recovery of the DO concentration.
River Modeling
The two primary reasons to conduct river modeling are:

1. To better understand physical, chemical, and biological processes
Fig. 10.5 DO sag curve
Dissolved oxygen (DO)

in a river
Saturated DO (Cs)
Critical point River velocity U

C0
Cmin
0 Distance, x
xc, tc
2. To develop models capable of realistically representing rivers, so that the models

can be used to support water quality management and decision making
Water quality management needs to understand key processes affecting envi-
ronmental problems in order to evaluate management alternatives. Examples of
such environmental problems are:
1. Thermal pollution due to power plant discharges
2. Sedimentation in harbors causing siltation and high dredging costs
3. Eutrophication due to excessive nutrient loadings
4. Low dissolved oxygen conditions caused by wastewater discharges
5. Accumulation of toxic materials in the sediment bed
Models play a critical role in advancing the state-of-the-art of hydrodynamics,
sediment transport, and water quality, and of water resources management. Because
of their requirements for precise and accurate data, models also ultimately contrib-
ute to the design of field data collection and serve to identify data gaps in
characterizing waterbodies. Models are used to analyze the impact of different
management alternatives and to select the ones that result in the least adverse
impact to the environment.
Models are often used to improve the scientific basis for theory development, to
make and test predictions, and to clarify cause-and-effect relationships between
pollutant loadings and the receiving waterbody. Models are often used to evaluate
and test potentially expensive water quality management alternatives prior to their
implementation. The cost of a hydrodynamic and water quality modeling study is
usually a small fraction of the implementation cost. Models can simulate changes in
an ecosystem due to changes in internal and/or external conditions, such as water
elevation variations or increased external pollutants. These simulations predict posi-
tive or negative changes within the river ecosystem due to the management actions,
such as improved sewage treatment or reduced agricultural runoff. These simulations
are obviously far more cost-effective than testing expensive management actions on
a trial-and-error basis, thus making models a useful tool for water quality
management.
236 Z.-G. Ji
In the past decades, hydrodynamic and water quality models have evolved from
simplified one-dimensional, steady-state models, such as the legendary QUAL2E
model [7], to complex three-dimensional, time-dependant models of hydrodynam-
ics, sediment, toxics, and eutrophication. Three-dimensional modeling has matured
from a research subject to a practical engineering tool. Over this same period,
computational requirements for realistic three-dimensional modeling have changed
from supercomputers, to high-end workstations, and then to personal computers.
These advanced three-dimensional and time-dependant models, which can also
be readily applied for one- and two-dimensional problem settings, provide
a powerful computational tool for sediment transport, water quality, eutrophication,
and toxic chemical fate and transport modeling studies. These advanced models
often include several coupled submodels for different physical, chemical,
biological processes in surface waters, such as:
1. Hydrodynamic model
2. Wind wave model
3. Sediment model
4. Toxic model
5. Eutrophication model
6. Sediment diagenesis model
7. Submerged aquatic vegetation (SAV) model
As an example, Fig. 10.6 illustrates the major components of the Environmental
Fluid Dynamics Code (EFDC) model [8]. In addition to computational modules,
these advanced models tend to evolve into complex software systems, comprising
many tools and sources of information. They may contain components for grid
generation, data analysis, preprocessing, postprocessing, statistical analysis,
graphics, and other utilities.
Transport in rivers is often dominated by the processes of advection and
dispersion. One-, two-, and three-dimensional models have been developed to
describe these processes. Study objectives, river characteristics, and data avail-
ability are key factors determining model applicability. In river studies, 1D and
steady-state models are commonly used, such as the QUAL2E model [7]. If a river
is wide enough to have significant lateral variations or deep enough to develop
vertical stratifications, 2D (and even 3D) models may be needed to simulate sediment
and toxicant transport in the river. For instance, sediment transport within
a meandering river is very complex. The velocities are faster at the outer bank and
slower at the inner bank. The lateral velocity difference directly influences the
sediment transport. There might be erosion occurring along the outer bank and
deposition occurring on the inner bank. Using a 1D model to represent the river is
equivalent to treating sediment transport as being uniform across the river,
eliminating the effect of river meandering on sediment transport and vertical
stratifications. A 1D model represents the entire cross section of the river as being
either net depositional or net erosional.
One-dimensional models, such as the widely used QUAL2E model [7], are
traditionally applied to river modeling. For most small and shallow rivers, these
Atmospheric data
Temperature, Atmospheric
Wind Solar radiation
precipitation, cloud deposition
Volatilization Water N2
Evaporation surface NH3
Hydrodynamic
Wind Wave Hydrodynamic
boundary data &
Model Model
inflow/outflow
Current Nutrients
surface elevation Nutrient Do
Wave mixing boundary & Algae
height temperature loading data Macroalgae
salinity
bottom stress
Sorption/
desorption
Sediment boundary Sediment
Eutrophication
& loading data Transport
Model
Model
Toxicants Sorption/
Sediment
desorption
Settling
Toxic Resuspension/
Model settling
Toxic boundary &
loading data Nutrient fluxes
Sediment oxygen
demand
Decay
Sediment Bed Model

Sediment Diagenesis Model
Burial
Fig. 10.6 Major components (submodels) of the EFDC model
1D models are often adequate to simulate hydrodynamic and water quality pro-
cesses. In 1D models, water surface elevation, velocity, and discharge vary only in
the longitudinal (along-the-river) direction and are constants in the lateral (across-
the-river) direction. This approach provides a simplified mathematical description
of river flows.
Rivers with a steep bottom slope often have a relatively high velocity and
a shallow water depth, and are characterized by gravel, cobbles, and rocks in
the riverbed. Coarse sands and finer particles are washed out by the high velocity.
238 Z.-G. Ji
The dominant gradient of water quality constituents is along the river in the
direction of flow. A 1D laterally and verticallyaveraged model is thus appropriate
for describing water flow and the transport of sediment and toxic chemicals.
Rivers with a moderate bottom slope result in a low-velocity waterway, often
characterized by a sediment bed consisting of a mixture of fine-grained cohesive
particles and fine sands. The dominant gradient of water quality constituents in
this kind of river is in the direction of the flow and a 1D model may still be
adequate. One-dimensional models are limited in their ability to capture the
complexity of natural rivers. The assumption that the characteristics of the river
are uniform both vertically and laterally may not be valid for wide, deep rivers. In
this case, the 1D approach may fall short of describing the river processes.
Transport in these rivers can have significant gradients either laterally or verti-
cally. In this case, a 2D or 3D model is needed to provide a better representation of
the river. Ji et al. [4] gave an example of modeling hydrodynamics, sediment
transport, and toxics in a small, shallow river.
Future Directions
The fate and transport of contaminants in rivers are complicated processes that
include physical transport and chemical and biological kinetics. Contaminants in
a river may be the result of either past or present disposal practices. Shutting off the
sources does not always solve the problem (e.g., DDT persists many years).
Consequently, it is essential that mathematical models for assessing contaminants
are accurate and reliable. In the past decades, significant progress has been made in
numerical model development, data collection, and computer software and hard-
ware. These developments have helped mathematical models to become reliable
tools for environmental management and engineering applications.
“Modeling is a little like art in the words of Pablo Picasso. It is never completely
realistic; it is never the truth. But it contains enough of the truth, hopefully, and
enough realism to gain understanding about environment systems” [9]. Water
quality management increasingly depends upon accurate modeling. This depen-
dency is further amplified by the adoption of the watershed-based approach to
pollution control. Models enable decision-makers to select better, more scientifi-
cally defensible choices among alternatives for river water quality management.
In many cases, the models are used to evaluate which alternative will be most
effective in solving a long-term water quality problem. The management decisions
require the consideration of existing conditions, as well as the projection of
anticipated future changes of the water system. In these applications, the river
models not only need to represent the existing conditions, but also have to be
predictive and give conditions which do not yet exist. Models are also used to
provide a basis for economic analysis, so that decision makers can use the model
results to evaluate the environmental significance of a project as well as the cost-
benefit ratio.
Three key factors have contributed to the great progress in the modeling of
rivers:
1. Better understanding and mathematical descriptions of physical, chemical, and
biological processes in rivers
2. Availability of fast and efficient numerical schemes
3. Progress in computer technology
The powerful, yet affordable computers in combination with fast numerical
algorithms have enabled the development of sophisticated 3D hydrodynamic and
water quality models. These advanced models contain very few simplifying
approximations to the governing equations. Personal computers (PCs) have evolved
rapidly to become the standard platform for most engineering applications (with the
exception of very large scale problems). PCs represent the most widely used computer
platform today. Models developed on a PC can be transformed to other PCs without
much difficulty. The relatively low prices of PCs also make modeling more cost-
effective. Due to the rapid advances in computer technology, PCs are now widely
used in river modeling studies.
Bibliography
Primary Literature
1. USEPA (2000) National water quality inventory: 1998 report to Congress. EPA 841-R-00-001.
US Environmental Protection Agency, Office of Water, Washington, DC
2. CSCRMDE (1987) Sedimentation control to reduce maintenance dredging of navigational
facilities in estuaries. Report and symposium proceedings. Committee on sedimentation control
to reduce maintenance dredging in estuaries, National Academy Press, Washington, DC
3. Ji Z-G (2008) Hydrodynamics and water quality: modeling rivers, lakes, and estuaries. Wiley,
Hoboken, 676 pp
4. Ji Z-G, Hamrick JH, Pagenkopf J (2002) Sediment and metals modeling in shallow river.
J Environ Eng 128:105–119
5. Chow V (1964) Handbook of applied hydrology, a comparison of water-resources technology.
McGraw Hill, New York
6. Streeter HW, Phelps EB (1925) A study of the pollution and natural purification of the Ohio
river. III: factors concerned in the phenomena of oxidation and reaeration. Bulletin Number 146,
US Public Health Service
7. Brown LC, Barnwell TO (1987) The enhanced stream water quality models QUAL2E and
QUAL2E-UNCAS: documentation and user manual. EPA/600/3-87-007. US Environmental
Protection Agency, Athens, Georgia
8. Hamrick JM (1992) A three-dimensional environmental fluid dynamics computer code: theo-
retical and computational aspects. The College of William and Mary, Virginia Institute of
Marine Science, Special Report 317, p 63
9. Schnoor JL (1996) Environmental modeling: fate and transport of pollutants in water, air, and
soil. Wiley, New York
240 Z.-G. Ji
Books and Reviews
Blumberg AF, Mellor GL (1987) A description of a three-dimensional coastal ocean circulation

model. In: Heaps NS (ed) Three-dimensional coastal ocean models, coastal and estuarine science,
vol 4. American Geophysical Union, Washington, DC, pp 1–19
Bowie GL, Mills WB, Porcella DB, Campbell CL, Pagenkopf JR, Rupp GL, Johnson KM, Chan
PWH, Gherini SA (1985) Rates, constants, and kinetics formulations in surface water quality
modeling, 2nd edn. USEPA, Environmental Research Laboratory, Athens. EPA/600/3-85/040
Casulli V, Cheng RT (1992) Semi-implicit finite difference methods for three-dimensional
shallow water flow. Int J Numer Meth Fl 15:629–648
Cerco CF (1999) Eutrophication models of the future. J Environ Eng 125(3):209–210
Chapra SC (1997) Surface water-quality modeling. McGraw-Hill, New York, 844 pp
Chapra SC, Canale RP (1998) Numerical methods for engineers, with programming and scientific
applications. McGraw-Hill, New York, 839 pp
Di Toro DM (2001) Sediment flux modeling. Wiley, New York
Fischer HB, List EJ, Imberger J, Brooks NH (1979) Mixing in inland and coastal waters.
Academic, New York, 483 pp
Gill AE (1982) Atmosphere-ocean dynamics. Academic, New York, 662 pp
Hutchinson GE (1957) A treatise on Limnology. In: Geography, physics and chemistry, vol I.
Wiley, New York, p 1015
Ji Z-G (2004) Use of physical sciences in support of environmental management. Environ Manage
34(2):159–169
Ji Z-G (2005) Water quality models: chemical principles. In: Water encyclopedia, vol 2, Water
quality and resources development. Wiley, New Jersey, pp 269–273
Ji Z-G (2005) Water quality modeling-case studies. In: Water encyclopedia, vol 2, Water quality
and resources development. Wiley, New Jersey, pp 255–263
Martin JL, McCutcheon SC (1999) Hydrodynamics and transport for water quality modeling.
Lewis, Boca Raton
Morel F (1983) Principles of aquatic chemistry. Wiley, New York, 446 pp
Park K, Kuo AY, Shen J, Hamrick JM (1995) A three-dimensional hydrodynamic-eutrophication
model (HEM3D): description of water quality and sediment processes submodels. The College
of William and Mary, Virginia Institute of Marine Science. Special Report 327, 113 pp
Schumm SA (1977) The fluvial system. Wiley, New York
Thomann RV, Mueller JA (1987) Principles of surface water quality modeling and control. Harper
and Row, New York
USEPA (1994) Water quality standards handbook, 2nd edn. US Environmental Protection Agency,
Office of Water, Washington, DC, EPA-823-B-94-005b
USEPA (1998) Bacteria water quality standard status report. US Environmental Protection Agency,
Office of Water, Washington, DC
USEPA (2000) Nutrient criteria technical guidance manual: rivers and streams. EPA-822-B-00-
002. Office of Water, Office of Science and Technology, Washington, DC
Wezernak CT, Gannon JJ (1968) Evaluation of nitrification in streams. J Sanit Eng Div ASCE
94(SA5):883–895
Wool AT, Ambrose RB, Martin JL, Corner EA (2003) Water quality analysis simulation program
(WASP), Version 6: Draft users manual. Available at: http://www.epa.gov/athens/wwqtsc/
html/wasp.html
Ziegler CK, Nesbitt B (1994) Fine-grained sediment transport in Pawtuxet river, Rhode Island.
J Hydraul Eng 120:561–576
Ziegler CK, Nesbitt B (1995) Long-term simulation of fine-grained sediment transport in large
reservoir. J Hydraul Eng 121:773–781
Chapter 11
Lake and Reservoir Fate and Transport
of Chemicals
Heinz G. Stefan, Xing Fang, and John S. Gulliver
Glossary
Dimictic lake A lake that has two complete mixing (circulation) periods per year
(one in spring after the ice melts and another in fall before the ice
forms).
Hyporheic flow Flow in a region beneath and along a stream bed. It is
characterized as mixing of shallow groundwater and surface
water which is important to sedimentary oxygen uptake.
Lentic “Stagnant” waterbody as opposed to “flowing”.
Limnology Science or the study of inland waters, e.g., lakes and reservoirs.
Oxythermal Variable to define fish habitat in inland waters using dissolved
parameter oxygen (DO) and water temperature limits, e.g., TDO3 – tem-
perature at DO = 3 mg/L.
Particulate Tiny subdivisions of solid matter suspended in a gas or liquid,
also known as particulate matter (PM) or fine particles.
Residence time The mean amount of time that water or a substance would stay or
“reside” in a lake or reservoir. Hydraulic residence time is equal to
lake volume divided by outflow rate. The residence time of
H.G. Stefan (*)
St. Anthony Falls Laboratory, Department of Civil Engineering,
University of Minnesota,
55414 Minneapolis, MN, USA
X. Fang
Department of Civil Engineering, Auburn University, 36849-5337 Auburn, AL, USA
J.S. Gulliver
Department of Civil Engineering, University of Minnesota, Minneapolis, MN, USA
242 H.G. Stefan et al.
a substance is equal to the quantity of a substance in volume

divided by the change of a substance in volume over time through
various lake’s removal mechanisms (outflow or flushing, settling,
and chemical and biological reactions).
Sedimentation The process or tendency for particles in suspension to settle out
of the fluid in which they are entrained and come to rest against
a barrier (e.g., lake or river bed).
Solute Material dissolved in water of a lake.
Stratification The formation of horizontal layers (strata) in which water
temperature and concentration of substances are different along
depth of a lake and reservoir.
“Eutrophication” is originally used to describe aging process whereby a lake is

transformed from a lake to a marsh to a meadow (fill the lake with sediments).
“Cultural eutrophication” occurs when the lake aging process is quickened or
accelerated by excess nutrients from human activities [1]. Understanding of the
fate and transport of water quality constituents in lakes and reservoirs is essential to
sustaining water quality and fish habitat in these inland waters. Constituent is used
generically and does not necessarily mean a polluting substance, e.g., dissolved
oxygen (DO) is a relatively benign variable. The fate of a constitute typically
depends on its transport (movement) through an inland water system (lake or
reservoir) and on sources, sinks, chemical and biological reactions, and other decay
mechanisms (e.g., settling). When sediment input is more than sediment outflow or
nutrients are more than demands of aquatic plants, a lake or reservoir becomes not
sustainable and the aging process of a lake is accelerated. The study of fate and
transport of a substance in a lake or reservoir is to qualitatively and quantitatively
account for mass balance of the substance through boundaries of and within the
waterbody. Using the principle of the conservation of mass to investigate mass
balance is not a new topic, but closely examining mass balance of various water
quality constituents in inland waters was only started a few decades ago, and
sustainability of aquatic systems is a relative new topic to researchers, water
resources managers, and the public. Due to waterborne pathogens as one of the
prime causes of disease, civil engineers began to plan, design, and construct urban
water and wastewater systems in the late nineteenth century, and then the water
quality management processes or models from streams to lakes and reservoirs
emerged. In the United States (U.S.), Rivers and Harbors Act in 1899 to Federal
Water Pollution Control Act in 1972 (subsequently amended and called the Act Clean
Water Act) promoted studies on water quality in receiving waters. Sustainability is
the capacity to endure. For humans, sustainability is the potential for long-term
maintenance of well-being, which has environmental, economic, and social
11 Lake and Reservoir Fate and Transport of Chemicals 243
dimensions. Sustainability in lakes and reservoirs involves how biological systems

remain diverse and productive over time and how designated uses (e.g., water supply,
recreation, fish and wildlife, etc.) endure over time. Lake and reservoir fate and
transport involve understanding and maintaining healthy aquatic ecosystems and
environments that provide vital goods and services to humans and other organisms.
Inland waterbody provides suitable habitat for survival and reproduction of
desirable fish, shellfish, and other aquatic organisms. Sustaining fish habitat in
lakes and reservoirs is a major resource management issue. Freshwater fish habitat
is constrained by physical, chemical, and biological attributes that relate to water
quality, food supply, and human interference [2]. In lakes, water temperature and
DO concentration are two of the most significant water quality parameters affecting
survival and growth of fishes [2, 3] and sustainability of lake aquatic ecosystem.
An increase of atmospheric greenhouse gases is projected to cause climate
warming, which would alter water temperature and DO characteristics in lakes.
These changes are in turn expected to have an effect on indigenous fish populations.
Understanding impacts of future climate changes on water quality and fish habitat in
lakes and reservoirs is essential to promoting sustainability of valuable but limited
water resources for humanity.
Introduction
“Sustainability” of lakes and reservoirs relates/depends on the transport and trans-

formation of materials into, within, and out of these “lentic” waterbodies. Typi-
cally, lakes are formed by natural, i.e., geophysical and hydrogeological processes,
whereas reservoirs or impoundments are typically man-made. There is a vast
amount of literature on physical, chemical, and biological structure and function
of lakes. There is a lot of literature on the engineering of dams and other hydraulic
structures, e.g., by Zipparro and Hasen [4], USBR [5], Novak [6], and Press [7],
but less on water quality processes in the impoundments they create [8–10].
Decisions on water releases from reservoirs and associated water quality down-
stream are difficult to make because of overlapping and contradictory objectives
such as water storage, flood protection, hydropower generation, sediment transport,
recreational water use, and water quality [11].
The study of lakes probably goes back to the scientists of the antiquity with
a keen sense of observation such as Plinius. Thieneman in Germany and Birge and
Juday in the USA (Wisconsin) were early scientific explorers of the physics,
chemistry, and biology of lakes. “Limnology” is the discipline under which
early investigators of lakes conducted their research, but geophysicists, biologists,
engineers, and people with other diverse backgrounds have contributed signifi-
cantly. The two volume book by Hutchinson [12] was an early and thorough
compilation of the state of scientific knowledge on lakes. An overview of North
American Limnology was compiled by Frey [13]. Books, e.g., by Lerman [14],
Wetzel [15], Horne and Goldman [16], Lampert and Sommer [17]; conference
proceedings of the International Society of Theoretical and Applied Limnology
(SIL); and journal articles, e.g., in Limnology and Oceanography, Hydrobiologia,

Water Resources Research, and Aquatic Sciences, have summarized the progressively
larger body of knowledge. Physical processes in lakes and their quantitative analysis
are described by Henderson-Sellers [18]. Ecosystems in lakes and reservoirs and their
modeling are described by Straskraba and Gnauck [19].
Reservoirs were created by man-made dams thousands of years ago in the
Middle East and in China to provide water for irrigation of agricultural lands
[20]. This practice provided the food source that sustained major cultures.
Reservoirs connected by aqueducts to major cities supplied drinking water even
in relatively dry regions. Eventually, sediment accumulation in reservoirs as well as
natural lakes made it painfully obvious that water-filled depressions in the land-
scape that receive inflows from a watershed are not sustainable forever. Sedimen-
tary rocks give testimony to this fact on geological timescales.
While sediment transport into reservoirs received some early attention, e.g., in
China [21], the limnology of reservoirs has become of interest recently because of
fishery issues in the reservoirs (e.g., Columbia River) and water quality and
ecological issues in the tailwaters below dams (e.g., Colorado River) [11, 22].
The need to better understand reservoir processes and reservoir operations on the
environment became obvious. Information is typically presented at meetings of the
major water-resources-oriented societies such as American Geophysical Union
AGU, the American Society of Civil Engineers (ASCE), American Water
Resources Association (AWRA), International Water Resources Association
(IWRA), American Society of Limnology and Oceanography (ASLO), SIL, and
International Commission of Large Dams (ICOLD). Symposium proceedings were
published by AGU [9] and ASCE [10], and a book on reservoir limnology was
written by Thornton et al. [8]. Introductions to the modeling of lakes and reservoirs
were given by Imberger in Fischer et al. [23] and Gray [24].
Although often lumped together, lakes and reservoirs differ significantly in several
respects: (a) the water levels in a reservoir can be controlled by an outlet structure and,
therefore, often varies more widely and at shorter timescales as water demand dictates,
whereas the water level of a lake depends on natural outflow, evaporation, and
precipitation on the water surface, and groundwater; (b) a reservoirs has typically
a shorter hydraulic residence time than a natural lake because a river or stream is
flowing through it; (c) horizontal water quality gradients in a reservoir can be far
greater than in natural lakes (except the very largest lakes) because of the flow through
a reservoir.
Even though entire cultures collapsed because water reservoirs for the irrigation
of farm fields could not be sustained, construction of large dams for the purpose of
creating new reservoirs has continued to the present (Three Gorges on the Yangtze
River in China). Reservoirs have been created now not only for water storage but
also to provide hydropower, flood protection, and sufficient water depth in navigable
waterways. Small retention and flow retarding structures have been built in large
numbers in cities and agricultural areas and in suburban developments for erosion
control, to retain stormwater runoff for infiltration and groundwater recharge, and
for urban stormwater management, i.e., flow reduction and water quality improve-
ment [25, 26].
Sustainability of lakes and reservoirs is a big issue. Topographic depressions that

lead to the formation of lakes in the landscape have been created naturally by a variety
of geophysical processes. Large numbers of lakes were created in the northern USA
during the last major glacial period only 10,000 year ago. Processes, resulting in over
70 lake types, are reviewed in Hutchinson’s 1957 book [12]. However, all lakes
and reservoirs are eventually filled by sediments because they are at a low point of
a landscape. Prolonging their lifespan while maintaining high water quality and water
use potential is the goal of many efforts.
Inputs, Outputs, and Transformations of Materials

in a Lake or Reservoir
The fate and transport of a water quality constitute in a lake and reservoir is the
study of mass balance of a substance. Mechanistic water quality studies and models
are based on the conservation of mass; that is, within a finite volume of water, mass
is neither created nor destroyed [1]. A mass balance equation quantitatively
accounts for all transfers (fluxes) of matter across the system’s boundaries (e.g.,
water surface and lake bottom) and all transformations occurring within the system
(a lake or reservoir) during a finite period.
Accumulation ¼ loading transport reactions
The movement of substances through the volume, along with water flow (surface
water and groundwater), is termed transport. Mass of substances is gained or lost by
transformations or reactions of the substances within the volume. Chemical and
biological reactions either add mass by changing another constituent into the
substance being studied or modeled or remove mass by transforming the substance
into another constitute [1].
The concept of water and material residence times is intimately linked to the
“sustainability” of lakes and reservoirs. As lentic waterbodies, they experience
periods of low flow and “flushing.” In those times, “deposition” of inorganic or
organic particulate matter becomes a crucial and defining process. Material fluxes
that can be used to develop material budgets for lakes and reservoirs are crucial for
the understanding of how lake chemistry and biology function. Material fluxes and
budgets can be for total suspended solids (TSS), total dissolved solids (TDS),
specific nutrients such as phosphate (PO4), ammonia and nitrate (NO3), silica
(SiO2), sulfate (SO4), organic matter in terms of total (C) or cell count, dissolved
oxygen (DO), or subcategories of nutrients such as total Kjeldahl N (TKN) or
dissolved reactive phosphorus or orthophosphorus (P).
ATMOSPHERE
WATER VAPOR HEAT ATMOSPHERIC LOADING
WINDSHEAR
PHYSICAL FRAME: TEMPERATURE, LIGHT, MIXING, FLOW, WIND, TURBIDITY
EXTERNAL PRIMARY CONSUMERS

WASTE LOADS HERBIVORES
ZOOPLANKTON
FEEDING
DETRITUS SECONDARY CONSUMERS

EXCRETION
DEAD BIOTIC TERTIARY CONSUMERS
& DEATH
SUBSTANCES CARNIVORES
RESUSPENSION
DECOMPOSITION
SETTLING
GRAZING
EXCRETION PRIMARY PRODUCERS
DECOMPOSERS ABIOTIC SUBSTANCES LIGHT
PHYTOPLANKTON
BACTERIA REMINERALIZATION BASIC INORGANIC & UPTAKE
MACROPHYTES
FUNGI ORGANIC COMPOUNDS
NANOPLANKTON
LEACHING SETTLING
SEDIMENT
Fig. 11.1 Material flow in a lake or reservoir with emphasis on biological processes (From [27])
If a material budget for a substance is developed, and the control volume is the
entire lake or reservoir, the material fluxes can be divided into the following
components:
1. External material inputs to a lake or reservoir by runoff from the watershed, i.e.,
by rivers, small streams, storm sewers, drainage ditches, and “overland” flow
from surrounding watersheds
2. External material inputs from the atmosphere through the lake or reservoir
surface in form of wet (rain) or dry deposition
3. Material flux to or from the lake or reservoir sediments by organic or inorganic
particle deposition or resuspension
4. Material flux of solutes (dissolved substances) into or from the pores of lake or
reservoir sediments
5. Material transformations by chemical and biological processes (kinetics) within
the various layers of a lake or reservoir
The “overland” flow is storm runoff flowing over the ground surface toward
a channel or a waterbody when rainfall has exceeded infiltration capacity and depres-
sion storage capacity. Figure 11.1 gives a schematic of fluxes across the upper and
lower boundaries of a lake or reservoir and internal transfer processes with emphasis
on fundamental biological processes, and Fig. 11.2 is an example of biochemical
interactions for various phosphorus components within a lake or reservoir.
External Material Inputs to a Lake or Reservoir

by Runoff from the Watershed
Besides the water, material inputs to lakes and reservoirs come in particulate or
dissolved form as TSS or TDS. Besides suspended particles, there can also be
4
ZOOPLANKTON
ALGAE
3
1 DRP 2
4
PP DOP 5 5
5
2
6 7 MACROPHYTES
SEDIMENT
Fig. 11.2 Schematic of phosphorus components and interactions in a lake (From [28]).
DRP dissolved reactive phosphorus, DOC dissolved organic phosphorus, PP particulate phospho-
rus. (1) Adsorption/desorption, (2) assimilation, (3) grazing by zooplankton, (4) regeneration,
(5) sedimentation, (6) sediment release/recycling, (7) regeneration (From [27])
a significant bed load component in rivers and streams. These inputs are often
detrimental to water quality. Suspended inorganic sediments cause undesirable
turbidity and fill lakes and reservoirs when they settle out, reducing the useful life
expectancy of reservoirs. This is one of the main arguments against the construction
of dams and impoundments. Since there are no known economical methods to
remove sediment from reservoirs, water storage in impoundments becomes unsus-
tainable, until incoming and outgoing sediment loads balance. This is a huge
problem for reservoirs on sediment-laden rivers. On the Yellow River in China,
the life expectancy of a reservoir can be just a few decades. The Colorado River
also has significant sedimentation problems. Bed load carried by inflowing rivers
contributes to the loss of storage capacity, mostly by the formation of deltas in
reservoirs. Diverting clay-laden turbid inflow from Lake Chico, AK, to the nearby
Mississippi River increased light penetration and photosynthesis sufficiently to
restore a dysfunctional ecosystem and bass fishery [29].
External materials (substances) that are fully dissolved in water and are not
reactive are not likely to accumulate long-term in lakes or reservoirs. An example is
road salt which is not only a conservative substance but also highly soluble [30].
Snowmelt water from roads in northern cities of the USA will first flow to and
accumulate near the bottom of a lake because the salinity makes the water slightly
denser than freshwater, some of it will flow into the pores of the lake sediments and
the groundwater table underneath, and some will get mixed into the lake and flushed
out if there is an outflow of water from the lake. Depending on the annual salt
loading and the hydraulic residence times, each lake receiving road salt runoff will
reach an equilibrium concentration, which may be high if the loading rate and the
residence times are high. Novotny and Stefan [30] give examples for Twin Cities’
lakes (7–104 t of Cl input per year and 3–14 years hydraulic residence time). The
associated seasonal maximum Cl concentrations at the end of the winter for these
lakes were projected to be 107 and 387 mg/L, compared to a chronic water quality
standard of 230 mg/L. Road salt accumulation is therefore considered to be fully
reversible in lakes but not in groundwater because hydraulic residence times can be
several order of magnitudes larger.
Nutrients such as phosphate, ammonium, or nitrate also come from watersheds,
especially where fertilizers are used for agriculture. They are dissolved in lake or
reservoir water in their original form, but they are reactive, and once they are
incorporated in plant mass, they are particulates that precipitate to the lake or
reservoir bottom. There have been literally thousands of studies of water quality
in runoff from watersheds. Studies have progressed from estimates of annual yields
of sediment, phosphorus, or nitrate from agricultural and urban watersheds to
computer simulations using fairly detailed models such as SWAT [31] and
SWMM [32, 33]. Heat transport in urban runoff has been simulated in models
such as SNTEMP [34] for natural watersheds and MINUHET [35] in urban
watersheds. Stormwater management is being studied to improve the quality of
the runoff into lakes and reservoirs. Reducing the inflow of undesirable nutrients or
contaminants to lakes or reservoirs is the goal of many TMDL (Total Maximum
Daily Load) studies.
External Material Inputs from the Atmosphere

and Material Release to the Atmosphere
The term “precipitation” in water quality studies may be used to describe the
external material deposited by wet (rainfall) and dry (dustfall) processes on water
surface of a lake or reservoir. The dry dustfall is a continuous process, while wet
precipitation (rain, snow, sleet, etc.) is an intermittent process. Precipitation
contributes chemical substances to surface waters and should be considered
(not simply ignored) in analyzing surface water quality [26]. Precipitation
impurities or chemical depositions will vary from one location to another and one
time period to another, for example, chloride ion concentrations are highest along
coastal areas.
Acid rain has harmed lakes in Europe and the USA. Sulfate emissions from the
burning of coal were identified as the cause. Another example of atmospheric input
is phosphorus loading of Lake Superior, where one third of the total P input was
found to be from the atmosphere [36]. On the other hand, Lake Superior cleansed
itself of PCB (polychlorinated biphenyl) that had accumulated by transferring it to
the atmosphere in gaseous form (volatilization) [37].
Volatilization and gas exchange between atmosphere and water are controlled
by mechanisms or processes at either sides of the air–water interface. Several gas
transfer theories [1], e.g., Whitman’s two-film theory and surface renewal models
[38], and many empirical formulas through data synthesis were developed and used
to quantify gas transfer process. Air–water equilibrium concentrations of gases
dissolved in water are quantified by Henry’s law [1]. Resistance to gas transfer for
many gases (e.g., O2, N2, CH4, H2S, and SO2) is in the liquid phase [1]. Oxygen
reaeration is an important process to surface water quality control and modeling and
was studied from the early stage of water quality engineering and management
research development and applications. Various formulas of oxygen reaeration in
rivers, streams, through hydraulic structures, in standing waters, and estuaries were
developed, e.g., summarized by Chapra [1] and Gulliver [38].
Particle Settling and Resuspension in a Lake or Reservoir
Organic or inorganic particle deposition is the most significant process that impairs
sustainability of lakes and reservoirs. Particles can settle out from the water column
when they are no longer kept in suspension; inorganic clay or silica particles settle
relatively fast; algae, which are almost neutrally buoyant, settle in dimictic lakes
after they have died in fall. The mechanics of particle settling in lakes and reservoirs
are very complicated because settling is the net effect of falling and resuspension in
an intermittently turbulent or laminar environment. Settling is driven by the density
difference between the solid particle and the water. If the two processes are
parameterized by a particle fall velocity and a vertical turbulent resuspension
coefficient, their interaction can be used to describe profiles of particle distributions
in lentic waterbodies [39].
Resuspension of particles in lakes and reservoirs can be an important process
that affects water quality, e.g., turbidity, in lakes and reservoirs. Resuspension of
sediment particles on the shoreline of a lake or reservoir can be observed when large
enough waves are present. In fact, shorelines may be devoid of fine sediments
because their resuspension has carried them to deeper and deeper water by turbidity
currents. This process of sediment redistribution in lakes is responsible for the
sorting of sediment beds such that shorelines exposed to wind and waves may be
devoid of fine sediments, while very fine particulates, particularly organic particles,
are settling to the deepest part of a lake [14, 40]. Because resuspension facilitates
the redistribution of once deposited materials in a lake, it is an important process.
The effect of wind on resuspension in shallow settling basins was investigated by
Sheng and Lick [41] and modeled by Rodney and Stefan [42] among others.
Erdmann et al. [43] provided data from Duluth harbor.
Because of adsorption and desorption of solutes on particle surfaces, particles
also become carriers of substances. These interactions are well known, and
descriptions can be found in textbooks, e.g., Thomann and Mueller [44], Clark
[45], and Weber and DiGiano [46]. For example, a substantial portion of the TP
load to a lake or reservoir can be adsorbed to particles. A seminal article on particle
transport in lakes was written by Weilenmann et al. [47].
Flux of Solutes into or from the Pores of a Lake

or Reservoir Sediment Bed
The best example of this type of flux is DO transfer from the benthic boundary layer
to the surfaces of sediment bed particles, where it is used by microbial biofilms or in
chemical reactions. The linkage between DO uptake rates, sometimes referred to
sedimentary oxygen demand (SOD), and the flow over, through and below the
sediment surface, have been established by field studies (Boynton et al. 1981) and
laboratory experiments [48]. Sediment topography and surface waves have been
shown to enhance physical transfer rates by “hyporheic flow” between the
overlying water and the pore system significantly [49–53]. Connections to the
microbial populations have been investigated experimentally, e.g., by Steinberger
and Hondzo [54] and O’Connor and Hondzo [55], and modeled, e.g., by Nakamura
and Stefan [56] and Higashino et al. [57]. Transfer of solutes other than DO, such as
nitrate, iron, and sulfate, at the sediment–water interface have also been studied by
Arnon et al. (2007) and Thumdrup et al. [58]. Transfer rates of solutes over a wide
range of flow and sediment conditions have been scaled, i.e., linked to dimensionless
parameters by O’Connor and Harvey [59]. Transfer of colloids and organic particles
into sediment pore systems have been the subject of studies by Packman et al. (2000)
and Huettel et al. (1996).
In the long-term, materials incorporated or adsorbed to particles on a lake or
reservoir bed may be buried in the sediments, or they may be recycled. The
recycling process of phosphate is well understood and depends on chemical
reactions with iron [60]. The process has been modeled in detail, e.g., by Nakamura
(1994), and as a seasonal bulk process by Lorenzen et al. [61]. The annual recycling
of phosphate from the lake or reservoir sediment is referred to as “internal nutrient
loading” and can sustain multiple annual algal blooms [62] or blooms over long
periods of time because the fertilizer is essentially recycling. The biogeochemical
and transport processes associated with the transfer and use of dissolved substances
in the sediment pore system are very numerous and complex and have been
described, analyzed, and modeled in books, e.g., by Boudreau and Joergensen
[63], DiToro [64], Clark [45], and Zheng and Bennett (1995). An experiment on
sulfate uptake in lake sediments is described by Manous et al. [65].
Material Transformations by Chemical and Biological Processes

(Kinetics) Within the Water Layers of a Lake or Reservoir
Biological processes by which materials in a lake are transformed include plant

growth by photosynthesis (primary productivity), grazing by zooplankton, and
predation by fish and higher-order animals. Respiration, death, and microbial
decomposition affect all organic materials. Chemical processes include oxidation
and reduction, and adsorption and desorption round out the menu of possible
interactions. Models of lake or reservoir water quality [66, 67] include algorithms
that quantify the material fluxes associated by these processes. Chemical and
biological kinetics used for this purpose can be simple zero or first order, or can
be higher order and/or more complex, e.g., Michaelis Menten, Monod, Haldane
kinetics, or processes in series. Figure 11.1 gives a very simple flowchart for the
principal biological processes in a lake.
Information on the chemical kinetics of lakes can be found in the limnological
literature; the literature on chemical kinetics, e.g., by Stumm and Morgan [68, 69],
Gobas and McCorquodale [70], Brezonik [71], Weber and DiGiano [46], and
Gulliver [38]; and the water quality modeling literature, e.g., by Thomann and
Mueller [44], Chapra [1], Zison et al. [72], and Bowie et al. [73].
Studies of lakes and reservoirs have progressed to ever finer detail and are often
highly specialized. Earlier studies of the physics of lakes and reservoirs dealt with
readily identifiable phenomena such as the seasonal temperature stratification, wind-
driven Langmuir circulation, and the thermal bar which are all described in classical
textbooks [15, 16]; more recent studies rely on field measurement techniques that
record temperature and turbulent velocities at high temporal and spatial resolution, as
well as transparency, specific conductance, and other parameters. Similarly, the study
of lake chemistry has progressed from dissolved oxygen and nutrients to subspecies of
nutrients, toxic substances, and endocrine disruptors [74, 75]. Lake ecology used to
concentrate on phytoplankton, zooplankton, and fish, but has evolved to include
detailed measurements and analyses of macrophytes, periphyton, and microbial
components (see, e.g., Scheffer [76]).
Lake and Reservoir Manipulations to Affect Fluxes of Materials
Manipulation of water quality and lakes and reservoirs has become a subject of
considerable interest, and many different tools have been tried. Cooke et al. [77]
give an overview of techniques used in lakes. Engineering approaches have been
described by Reckhow and Chapra [78]. Examples of simulation model
applications to guide lake or reservoir management decisions have been given by
Henderson-Sellers [79].
Destratification by air bubble release at the lake bottom was an early tool [80]
and has been modeled by Fast [81], Goossens [82], Zic and Stefan [83], and Zic
et al. [84]. Instead of air, pure oxygen has been used in deeper lakes. Air bubble
plume can cause phosphorus recycling from the lake sediments [85].
The US Army Corps of Engineers in Vicksburg, MS, has developed and applied
various techniques to mix stratified reservoir water near dam faces with large
propellers before its release through outlet structures, in order to improve the
water quality in the tailwater downstream from reservoirs.
Mixing (aeration) of selective layers has been attempted and practiced, particu-
larly for aeration of selective layers in stratified lakes [86, 87]. Hydraulic jet mixing
systems are energetically attractive [88–90]. To prevent internal phosphorus load-
ing of eutrophic lakes, various sediment treatment techniques have been developed
LIGHT
3
WATER
VOLUME
NUTRIENTS INFLOW 1 SURFACE OUTFLOW
2 4
SEDIMENT
SURFACE MIXED LAYER
ALGAE MIXED LAYER
NUTRIENTS
2 4
THERMOCLINE
HYPOLIMNION HYPOLIMNETIC HYPOLIMNETIC

6
NUTRIENTS OUTFLOW
6 7
SEDIMENTS
Fig. 11.3 Schematic of lake manipulation (processes 1–7) for eutrophication (water quality)
control (From Henderson-Sellers [79])
(alum, RIPLOX). Herbicides have been applied to kill phytoplankton [91] or

macrophytes selectively. Biomanipulation by influencing the food chain is an
elegant technique that has had some success. Fish stocking and fish management
in lakes are very common in some states where recreational fishing is popular.
Lake manipulation techniques (Fig. 11.3) to control the excessive growth of
algae are designed to change one or more of the following:
1. The inflow of water, nutrients (mostly P), or sediments to a lake
2. The loss of algal populations by settling or grazing or outflow from a lake
3. The availability of light to an algal population of a lake
4. The nutrients in the photogenic surface layer of a lake
5. The transfer of nutrients from the hypolimnion to the surface layer of a lake
6. The concentration of nutrients in the hypolimnion
7. The release of nutrients from the sediments
Material Residence Times, Stratification, Mixing,

and Currents in a Lake or Reservoir
Residence Times
Hydraulic residence time (years) is a measure of the flushing effect by flow through
a lake or reservoir. It can be defined as a lake’s or reservoir’s volume (m3) divided
by the volumetric outflow rate (m3/year). Some definitions use inflow instead of
outflow rates. Hydraulic residence times range from a few days or weeks for
10
‘EUTROPHIC’
Non - acceptable
PHOSPHOROUS LOADING (gTPm–2 yr–1)
loading
1.0
Acceptable
loading
0.1
‘OLIGOTROPHIC’
eutrophic
mesotrophic
oligotrophic
0.01
0.1 1.0 10 100
OUTFLOW PER UNIT SURFACE AREA (myr–1)
Fig. 11.4 Annual phosphorus loading tolerance of a lake as a function of annual hydraulic
flushing rate (outflow per unit surface area) (From [92])
shallow river impoundments, to several years for quaternary lakes and large
reservoirs, to hundreds of years of the world’s largest lakes. Hydraulic residence
time determines how long different chemical or biological processes can work on
the removal or generation of different materials/substances. Lake and reservoir
ecosystems depend on hydraulic residence times.
Hydraulic flushing or residence time determines the impact of a nutrient input to
a lake. Lake or reservoir eutrophication is in many instances linked to phosphorus
supply to a lake. The famous Vollenweider [92] diagram reproduced in Fig. 11.4
shows annual flushing rate (m/year) = annual outflow per lake surface area on the
horizontal axis and annual volumetric P loading rate (g/m3/year) on the vertical
axis. The lines are boundaries between eutrophic, mesotrophic, and oligotrophic
lakes, i.e., very high, intermediate, and low primary (plant) productivity. It can be
seen that the same trophic state can be achieved with a high P loading rate when the
flushing rate is higher.
The rate of removal of a material/substance/organism, e.g., by settling or by
volatilization or by uptake, can be used in addition to the flushing rate to
define a material residence time. If material is removed, the material residence time
will be shorter than the hydraulic residence time. For first-order processes,
the following relationship holds:
(material residence time)1 ¼ (hydraulic residence time)1

þ (settling rate coeff.)
þ (removal rate coeff.)
þ (volatilization rate coeff.)
Stratification or Vertical Mixing of a Lake
For water quality in lakes or reservoirs, it makes a large difference if a lentic

waterbody is density (temperature) stratified or not, in addition to the flushing effect
by flow through the lake or reservoir. Temperature stratification is driven mostly by
solar irradiation of the water surface and seasonal temperature varia-
tions. Stratification can occur during the day and be absent during the night, it
can be intermittent for several days at a time, it can be seasonal in northern
temperate climate regions with one or two full mixing events, or permanent if a
saline water layer forms in a lake. Correspondingly, lakes are referred to as
well-mixed, polymictic, monomictic, dimictic, or meromictic. Stratification and/
or mixing depend on the physical processes that produce them.
The density of freshwater changes ever so slightly with temperature or salinity
(concentration of dissolved substances). A lake or reservoir can become density
stratified whenever heat or any dissolved substance is added or removed from portions
of its water. Because a lake or reservoir is a lentic or standing waterbody, its water will
tend to arrange itself in layers of increasing density from top to bottom. The
differences are typically very slight, but significant. The stable equilibrium of
a density-stratified lake is easily reinforced or upset by many physical mechanisms
such as daily heating through the water surface by solar radiation, inflows of warmer or
colder water, or wind action on the water surface. These processes and their effects on
lake stratification have been studied extensively.
There are two major processes that produce lake mixing: convective mixing
induced by density instability, i.e., denser water on top of lighter water; or forced
mixing, e.g., by wind or inflows (jet mixing). Density instabilities can be caused by
cooling of surface waters (cooler water is denser above 4 C) or by dissolved (e.g.,
road salt) or suspended (e.g., sediment) materials in inflowing waters. Wind mixing
depends on the shear stress exerted by wind on the water surface. That shear causes
not only clearly visible “progressive” surface waves with amplitudes dependent on
shear stress but also less noticeable and relatively slow “standing” waves on the
lake surface. In a stratified lake, internal waves are produced by wind and contribute
to mixing at different depths when they break on the sloping beaches, invisible to
the eye below the water surface.
Criteria for lake or reservoir mixing have been developed. One of the easiest to
use is the “lake geometry ratio” AS0.25/Hmax [93]. It is the ratio of the fourth root of
the lake surface area AS (in m2) divided by the maximum lake depth Hmax (in m).
In the north-central USA, lakes with a geometry ratio greater than 4 are polymictic;
those with a ratio smaller than 1 are dimictic. The lake geometry ratio depends on
lake bathymetry only and is climate independent. That makes it easier to use than
other criteria, e.g., the Wedderburn number which requires knowledge of the wind
shear stress (Imberger in Fischer et al. [23]).
Density- or Wind-Driven Currents in a Lake or Reservoir
When water temperature or dissolved or suspended material content makes

a portion of the water in a lake denser than its ambient water, that water will always
tend to sink lower in a stratifies lake until it reaches a layer of equal density.
This sinking has a huge effect on water quality because it redistributes material at
a rate that is much faster than molecular diffusion.
An organized sinking flow is a density current or turbidity current if suspended
sediment makes the water heavier than its surroundings. Turbidity currents are
sediment-laden underwater rivers that can occur unnoticed because they take place
below a quiescent water surface. Only a plunge line visible on the water surface
by the accumulation of floating wood or a color contrast between clear lake water
and turbid inflow may give away the occurrence of a significant process. Reservoirs
in the Yellow River Basin are famous for their turbidity currents, and the routing
and the release of these turbidity currents through bottom outlets at a downstream
dam site have become management tools to delay the deposition and filling of the
Yellow River reservoirs by the fine loess sediments. Plunging flows and density
currents in reservoirs have been investigated and modeled among others by
Imberger [23], Alavian et al. [94], Akiyama and Stefan [95–98], Johnson et al.
[99, 100], Farrell and Stefan [101, 102], and Fang and Stefan [103]. A density
current of saline snowmelt water into an ice-covered lake was described by Ellis
et al. [104].
When the inflow to a lake or reservoir is warmer than the lake, it will spread out
on the surface of the lake [105–109].
Wind induces horizontal exchange flows in multibasin lakes [110–113] that can
transport salinity, nutrients, or contaminants from basin to basin. Wind also controls
vertical mixing of a lake by progressive waves and standing waves (seiche) both on
the lake surface and internally, if a lake is density (temperature) stratified. Wind
also has a determining effect on the depth of the surface mixed layer [114, 115].
In nutrient-rich (eutrophic) lakes, wind can become the determinant/controlling
factor for algal growth [116]. Wind controls the wind-driven circulation,
stratification, and mixing dynamics in lakes, and thereby the fate and transport of
organic and inorganic materials as well as the growth of algae [116] and rooted
vegetation (macrophytes) in them [117]. Wind sheltering by trees, buildings, or
topography can severely hinder wind access and thereby reduce wind shear on the
water surface, wave action, and vertical turbulent mixing by waves, turbulent
eddies, and Langmuir circulation. The development of the atmospheric boundary
layer on a lake surface and its effects on surface mixing and water quality is often
ignored and only partially understood and predictable [118].
Simulation of Water Quality in a Lake or Reservoir
Water quality models are often used as tools to guide decisions on lake management
and reservoir operation. Without them, decisions have to be made based on
experience. Water quality in a lake or reservoir is best simulated by a deterministic
model. Physical modeling attempted by the USEPA and the Army Corps of
Engineers did not work out because scaling laws for the physics, chemistry, and
biology in lakes or reservoirs were incompatible. The simulation models use
mathematical descriptions and rely on basic principles, such as conservation of
mass, heat, and mass transfer equations, and well-known relationships for chemical
and biological kinetics. A review of basic principles of model formulations can be
found, e.g., in textbooks by Thomann and Mueller [44], Chapra [1], and Clark [45],
and in a monograph by Stefan, Ambrose, and Dortch [119] but also in some user
manuals.
A review of major available hydrological and water resources simulation models
was given by Wurbs [120]. One- or multidimensional models of lake or reservoir
water quality, e.g., WASP (EUTRO04) [66], CE-QUAL-R1 [121], CE-QUAL-W2
[67], DYRESM [122], and MINLAKE [123], are composed of submodels that
analyze and project the hydrodynamics in a lake or reservoir, the heat transfer to
obtain temperature distribution and density stratification, sediment particle dynamics,
chemical kinetics, biological kinetics and particle–solute interaction. Numerous
boundary conditions have to be specified at the water surface or the sediment–water
interface. These models require considerable data input and can therefore be tedious to
use. Training on the use of some of these models is available from the custodian
organizations or from consultants. Information on some of these models can be found
on the internet under the name of the model.
Empirical models, e.g., to relate nutrients to lake eutrophication [27, 92, 124], are
popular and widely used tools to guide nutrient management in lakes. However, the
simplifications necessary in empirical models can lead to outliers and high variance
in the results of these models. Empirical models are typically zero-dimensional, i.e.,
they do not resolve spatial differences and assume well-mixed conditions; they are
usually applicable at long (monthly or annual) timescales. Sometimes they apply
only to a portion, e.g., the surface mixed layer, of a lake (Fig. 11.3).
One-dimensional water quality models of stratified lakes describe water quality

C(z,t) as a function of depth (z) below the water surface and of time (t). They can be
quite sufficient for smaller lakes and work well even at short, e.g., daily, timescales.
They do not give water quality changes in the longitudinal direction of reservoirs.
Two-dimensional reservoir water quality models have to resolve the flow
through the waterbody and can show longitudinal water quality gradients. Three
zones are often found by two-dimensional models [8]: a shallow inflow region where
velocities are initially high, sediment settles out, and deltas form; a productive zone
where nutrients in the inflow combined with sunlight lead to high primary productivity,
i.e., phytoplankton growth; and finally a quiescent region where algae are consumed or
settle out and water quality is increased. Two-dimensional (width-integrated) models
where water quality parameters C(x, z, t) are functions of distance from the dam or
inflow, depth below the water surface, and time are very appropriate for reservoirs.
Such models can handle stratified flows such as density and turbidity currents, at least in
their developed stages away from the plunge point. A schematic of 0-, 1-, 2-, and
3-dimensional model resolution is shown in Fig. 11.5.
Three-dimensional models are used to simulate the most demanding and com-
plex waterbodies. A 3-dimensional model developed by HYDROQUAL has been
applied to the tidal Chesapeake Bay, San Francisco Bay, and the Upper Mississippi
River impoundments and Lake Pepin in the Twin Cities area of Minnesota. Over 30
water quality constituents have been simulated at daily or shorter timescales and for
multiyear periods. Another 3-dimensional hydrodynamic and water quality model
ELCOM-CAEDYM was developed and applied by Bruce et al. [125] to simulate
the role of zooplankton on C, N, and P cycling in Lake Kinneret, Israel. Other
applications of this model have been described by Gal et al. [126], Leon et al. [127],
and Robson and Hamilton [128]. An introduction and CE-ROM for 3-dimensional
hydrodynamic and water quality lake modeling was offered by Edinger [129].
As an example of a water quality model, the structure and use of the MINLAKE
model will be described in more detail. The original MINLAKE model [123]
simulates the processes shown in Fig. 11.6. A flowchart is shown in Fig. 11.7. It
was applied, e.g., by West and Stefan [130–132], to several lakes in the Twin Cities,
Minnesota, and elsewhere, and modified by Fang et al. [133] to MINLAKE96. This
deterministic, process-oriented, dynamic and one-dimensional (vertical)
year-round lake water quality model was later modified for water temperature
and dissolved oxygen (DO) simulations in deep oligotrophic lakes in Minnesota,
and then called MINLAKE2010 [134].
Year-Round Water Temperature and Ice-Cover Model
The MINLAKE numerical simulation model for water temperature profiles in lakes
solves the one-dimensional, unsteady heat transfer equation
CONTROL
VOLUME
0-D C(t)
MIXED CELL
CONTROL
VOLUME
1-D C(z,t)
STRATIFIED LAKE
CONTROL
VOLUME
x 2-D C(x,z,t)
STRATIFIED RESERVOIR
z WITH VERTICAL AND
HORIZONTAL GRADIENTS
CONTROL
VOLUME
y x 3-D C(x,y,z,t)
z THREE-DIMENSIONAL LAKE
Fig. 11.5 Schematic of the spatial resolution of lake or reservoir water quality models. Control
volumes are shown as shaded elements. Concentration of a water quality constituent is designated
by (C) (From Stefan, Ambrose, and Dortch [119])

@T 1 @ @T Hw
¼ Kz A þ (11.1)
@t A @z @z rcp
where T ( C) is the water temperature, t (day) is the time, A (m2) is the horizontal
area of a lake as a function of depth z (m), KZ (m2 day1) is the vertical turbulent
heat diffusion coefficient, rcp (J m3 C1) represents heat capacity per unit
volume of water and is equal to the density of water (r) times heat capacity of
water (cp), and Hw (J m3 day1) is the heat source or sink strength per unit
volume of water. Solar radiation absorption in the water column contributes to
the heat source term. Heat exchange between the atmosphere and the water
during the open-water season (7) is treated as a source/sink term for the topmost
SURFACE
HEAT
WIND MIXING EXCHANGE LIGHT
WATER
VOLUME
SURFACE
NUTRIENTS
ENTRAINMENT MIXED EUPHOTIC OUTFLOW
LAYER ALGAE
ZONE
SEDIMENT
DENSITY
CURRENT
INTERLAYER
THERMOCLINE
INTERFLOW DIFFUSION
PARTICLE
SETTLING
NUTRIENTS
SEDIMENTS
Fig. 11.6 Schematic of the variables and processes simulated by the 1-D MINLAKE model [123]
water layer in a lake. For the open-water season, the computational scheme and
the determination of source and sink terms have been discussed, e.g., by Edinger
et al. [135], Ford and Stefan [136], Harleman [137], Hondzo and Stefan [138,
139], among others. Equation 11.1 is solved numerically using an implicit finite
difference scheme and a Gaussian elimination method, e.g., for time steps of 1
day and water layer thicknesses of 1 m. The model uses a stacked layer system to
represent a lake and its environment in the open-water and winter ice-cover
periods. Besides the water layers, the lake sediments, the ice cover, and the snow
cover are included in the model by separate submodels (Fig. 11.8).
Climate conditions and variations over seasons are driving forces of seasonal
variations of water temperature in a lake. Weather parameters needed are: daily
average air temperature, dew point temperature, wind speed, sunshine percentage,
total daily solar radiation, and precipitation (rainfall and snowfall). They are used as
model input parameters to calculate heat fluxes across the water surface or ice and
snow cover in winter. If solar radiation has not been measured, it is calculated [141].
Compared to the regional water temperature model for the open-water season
[139], the year-round water temperature simulation model has been expanded
significantly by simulating ice and snow covers above the water and including the
heat exchange between each water layer and its adjoining sediments.
The heat flux across the sediment–water interface is treated as a contribution to
the source/sink term for each water layer from the water surface to the lake bottom
[142]. The direction of the heat flux between the sediment and the water reverses
frequently on shorter, e.g., daily, timescales [143]. Heat can transfer into or out of
the lake sediment during both the open-water season and the winter ice-cover
period. The lake sediments not only provide seasonal heat storage but also add
significant thermal inertia to the water column. Sediment heat fluxes are most
important in shallow lakes and during winter ice-cover period.
INPUT
HYDROLOGIC MODEL
INFLOW
QUALITY- INITIAL METEOROLOGICAL
QUANTITY CONDITIONS DATA MORPHOMETRY
DATA
HYDRODYNAMIC PROCESSES
HEAT WATER
BALANCE TEMPERATURE
CONSERVATION WIND MIXING CONVECTIVE

MIXING
OF MASS MODEL MODEL MIXING MODEL
DENSITY
WATER OUTFLOW GROUNDWATER
CURRENT
BALANCE MODEL MODEL
MODEL
BIOCHEMICAL COMPONENTS
NON-LIMITING AVAILABLE NITRATE AMMONIUM DISSOLVED

NUTRIENTS PHOSPHORUS NITROGEN NITROGEN OXYGEN
LIMITING INTERNAL INTERNAL

PHOSPHORUS LIGHT NITROGEN
NUTRIENTS
BIOMASS PHYTOPLANKTON ZOOPLANKTON
INTERNAL
SUSPENDED BOTTOM
SOURCES DETRITUS
SEDIMENT SEDIMENT
AND SINKS
OUTPUT
PREDICTED LAKE QUALITY PREDICTED OUTFLOW QUALITY
Fig. 11.7 Flowchart and major model sections and components in the original MINLAKE model
[123]
Meteorological forcing
T = 0°C TA Open water season Ice cover season TA T = 0°C

HSN HE HC
HSN HA HBR HE HC U
U
Snow
Ice
Water
HSED HSED
Sediment Sediment
Sediment
Temperature profile Temperature profile

in summer in winter
Fig. 11.8 Schematic of a stratified lake with a 10-m sediment layer showing heat transfer
components and water/sediment temperature profiles in the open-water season and in the ice-
cover season [140]
The snow and ice thickness submodels developed by Gu and Stefan [144] have
been used with some modifications. The complete set of equations for the ice- and
snow-cover simulation has been summarized by Fang and Stefan [142]. During the
ice-cover period, the model simulates ice thickness and sediment temperature
profiles (heat conduction equation) first, then determines the heat source/sink term
HW in (Eq. 11.1), and finally solves the heat transfer (Eq. 11.1) to obtain water
temperature profiles below the ice. At the air/snow interface (or air–ice interface if
snow is absent), the net heat flux from the atmosphere into or out of the snow/ice
cover is calculated. Contributions of heat flux are made by solar radiation (HSN),
evaporation (HE), and convection (HC). Snow thickness is determined from snow
accumulation (precipitation), followed by compaction and snow melting. The model
simulates melting of snow by surface heat input (convection, rainfall, solar radia-
tion), melting within the snow layer due to internal absorption of short
wave radiation, and transformation of wetted snow to ice when cracks in the ice
cover allow the water to migrate on the ice surface [133]. In the model, ice growth
occurs at the ice/water interface and at the ice surface [133]; ice decay occurs and is
simulated at the snow/ice interface, the ice/water interface, and within the ice layer.
Ice formation on small freshwater lakes generally occurs on a calm, cold night.
Rising winds and daytime heating may subsequently break up this cover until
calm and cold conditions occur again and the ice cover forms a second time
[145]. A process-descriptive algorithm which replaced previous empirical and
lake-size-dependent criteria to predict the date of ice formation is incorporated in
the model [146]. The ice-formation algorithm uses a full heat budget equation to
estimate surface cooling, quantifies the effect of forced connective (wind) mixing,
and includes the latent heat removed by ice formation. The algorithm has a fine
spatial resolution near the water surface where temperature gradients before
freeze-over are the greatest. Inverse temperature stratification occurs in the
near-surface water several hours before ice formation. The ice-formation
algorithm is combined with the year-round temperature model and was tested
previously against observations in Ryan Lake and eight other Minnesota lakes for
multiple (9–36) years. The difference between the simulated and observed (per-
manent) ice-formation dates was less than 6 days for all lakes studied. Water
temperature profiles are typically predicted with an RMSE on the order of 1 C.
The water temperature and ice-cover model can be run in a continuous mode
over many simulation years covering sequences of both open-water seasons and
ice-cover periods. The model uses a stacked layer system (Fig. 11.1); lake sediment
and water layers during the summer open-water season, and additional ice-cover
and snow-cover layers for winter. A lake or reservoir is divided into well-mixed
horizontal layers having typical thickness of 1 m.
Year-Round Dissolved Oxygen Model
The one-dimensional, deterministic, unsteady year-round dissolved oxygen trans-

port equation [147], which is the basis of the DO model, is

@C 1 @ @C Sb @A
¼ AKZ
@t A @z @z A @z
þ PMAX Min½LChla (11.2)
1
kr yr T20 Chla kb yb T20 BOD
YCHO2
where C(z,t) is the dissolved oxygen concentration in mg L1 as a function of depth

(z) and time (t), A(z) is the horizontal area at different depths in m2, Kz is the
turbulent diffusion coefficient of DO in m2 day1, Sb is the sedimentary oxygen
demand (SOD) coefficient in mg O2 (m2 day1), Pmax is the maximum specific
oxygen production rate by photosynthesis at saturating light conditions in [mg O2
(mg Chl-a)–1 h–1], Min[L] is the light limitation determined by the Haldane
equation, Chl-a is the chlorophyll-a concentration in mg L–1, YCHO2 is the yield
coefficient which is the ratio of mg chlorophyll-a to mg oxygen, kr and kb are the
first-order decay for biochemical oxygen demand (BOD) and respiration rate
coefficient (day–1), respectively, yr and yb are the temperature adjustment coeffi-
cient for plant respiration and BOD, BOD is the biochemical oxygen demand
concentration in mg L–1, and T(z, t) is the water temperatures in C. In the
model, the oxygen transfer through the water surface (reaeration) is expressed as
ke ðCs CÞ= D zs (11.3)
It is used as an oxygen source or sink term in the topmost water (surface) layer.
Diffusive oxygen flux at the lake bottom is set equal to zero as a boundary
condition. (Sedimentary oxygen demand is treated as a source/sink term.)
Meteorological forcing
Open water season Ice cover season

DO=0mg/l O2
Fs
DO=0mg/l
Ua
Snow
Phytoplankton Ice
Phytoplankton
P R
R=0 P
(Chl-a) BOD Water
(Chl-a)
SOD
WOD
Sediment
Sediment Sediment
Do profile Do profile
in summer Oxygen source Oxygen sink in winter
Fig. 11.9 Schematic of a stratified lake showing dissolved oxygen source/sink terms and
dissolved oxygen profiles in the open-water season and in the ice-cover period [140]
For the dissolved oxygen simulations in a lake over the winter season (Fig. 11.9),
some modifications had to be made in Eq. 11.2 to account for the presence of an ice
cover and low temperatures. These modifications [142] include: (a) reaeration is
zero (ke is set equal to zero); (b) the respiration rate coefficient kr is zero; (c) the
water column oxygen demand, WOD, by detrital and other organic matter, is set to
a constant value (0.010 g O2 m3 day1), independent of the trophic status of a lake
[148]; and (d) sedimentary oxygen demand (Sb) is dependent on trophic state and
set equal to 0.226, 0.152, and 0.075 (g O2 m2 day1) for eutrophic, mesotrophic,
and oligotrophic lakes, respectively.
Water column oxygen demand in winter (WOD) is constant and very low,
whereas the biochemical oxygen demand (BOD) in summer varies as a function
of trophic status. Both WOD and BOD describe the same processes, i.e., microbial
and chemical decomposition of detrital or nonliving organic material, and
can therefore be expected to depend on trophic state. However, the very limited
database for WOD and its small value did not justify or require an adjustment of
WOD for trophic state.
Trophic state characterizes biological productivity and relates to plant density,
especially phytoplankton; nutrient availability; photosynthetic oxygen production;
and respiratory consumption. All of these characteristics can be used to measure
trophic status. In many, but not all, lakes, trophic states are closely related to
phytoplankton concentration and lake clarity. With this in mind, lakes were divided
into three trophic states by the mean annual chlorophyll-a concentrations given in
Table 11.1. Secchi depths (SD) of 1.2, 2.5, and 4.5 m were selected for eutrophic,
mesotrophic, and oligotrophic lakes using Carlson’s trophic state index [159],
respectively.
The radiation attenuation in a lake is used to quantify how much of the solar
energy reaching the water surface can penetrate through a water column to heat
water below the surface and to support photosynthesis of phytoplankton. Total
attenuation coefficient, m (m1), is given by
m ¼ mw þ mch ðChl aÞ (11.4)
where mw (m1) is the attenuation coefficient of the lake water excluding chloro-
phyll-a, mch [m1 (g m3 Chl-a)1] is the attenuation coefficient due to chlorophyll-a
[152], and Chl-a is chlorophyll-a concentration (g m3). An alternative is to estimate
total attenuation coefficient as function of Secchi depth [139] by the Eq. 11.5. The
coefficient 1.84 is empirical, and other similar values have also been proposed.
1:84
m¼ (11.5)
SD
Model Input and Output
Model parameters and coefficients are shown in Table 11.1 for the temperature
submodel and in Table 11.2 for the DO submodel of MINLAKE2010.
MINLAKE2010 model input data are read through ten input files: (1) bathymetry
data, (2) weather data (solar radiation optionally from Maxwell et al. 1998),
(3) lake-specific model parameter, (4) fixed model parameter, (5) long-term annual
and January air temperatures, (6) weather station information (elevations, longi-
tude, and latitude), (7) dimensionless sediment temperature profile [162], (8) sea-
sonal chlorophyll-a pattern file [147, 163], (9) path file directing the program to
open model input or output files, and (10) future climate data, e.g., output data of
General Circulation Models (GCM) models.
Most of the input data files are common data files for model simulations of many
lakes. Only three data files are created for a specific lake: the bathymetry data file,
the lake-specific model parameter input file, and the path file. Sediment temperature
at 10 m below sediment–water interfaces of a water layer is a model input [162].
These 10-m sediment temperatures have been related to mean annual air
temperatures [164]. Long-term air temperature in January is used to specify
whether a lake will have an ice cover [162]. The station data files, including
long-term air temperatures, were developed for 209 weather stations over the
contiguous United States of America. Therefore, in addition to three lake-specific
input files mentioned earlier, the station data file and the file with long-term air
temperatures were also updated when new weather stations were used.
Because the growth of phytoplankton is not explicitly modeled in the
MINLAKE2010 version of the model (it is in the original version by Riley and
Stefan [123]), the seasonal pattern of standing algal crops had to be specified, as
shown by the solid line in Fig. 11.10 for mesotrophic lakes. Similar patterns were
developed also for eutrophic and oligotrophic lakes [134, 147].
The model output consists of daily vertical profiles of water temperature (example
in Fig. 11.11) and DO in the water column. Three measured temperature profiles on
Table 11.1 Parameters and coefficient values in the hydrothermal model [147]
Range and Selected
Coefficients and symbols Units references value
Open-water season
Radiation absorption for bw – 0.4a 0.4
water
Sediment specific heat cpsed kcal kg–1 C–1 0.2–0.3b 0.28
Sediment thermal ksed kcal day–1 C–1 m–1 8.64–51.8b 19.25
conductivity
Radiation attenuation by mchh m2 g1 Chl-a 0.2–31.5c 20.0
Chl-a
Radiation attenuation by mw m1 0.33–1.03d 0.51
water
Sediment density rsed kg m3 1,650–2,300b 1,970
Wind sheltering Wstr – 0.01–1.0e Varies
Winter ice cover
Surface reflectivity for ice ai – 0.55f 0.55
Surface reflectivity for snow asw – 0.4–0.95g 0.80
Radiation absorption for ice bi – 0.17–0.32h 0.17
Radiation absorption for snow bsw – 0.17–0.34i 0.34
Snow compaction csw – 0.125–0.5g 0.4
Ice thermal conductivity ki kcal day1 C1 m1 45.8b 53.6
Snow thermal conductivity ksw kcal day1 C1 m1 2.16b 5.57
Ice density ri kg m3 920b 920.0
Snow density rsw kg m3 100–400g 300.0
Radiation attenuation by ice mi m1 1.6–7.0j 1.6
Radiation attenuation by msw m1 20–40k 40.0
snow
Ice latent heat of fusion li kcal kg1 80l 80.0
Snow latent heat of fusion lsw kcal kg1 80l 80.0
a
Dake and Harleman [149]
b
Carslaw and Jaeger [150]
c
Bannister [151]
d
Megard et al. [152]
e
Riley and Stefan [123]
f
Bolsenga [153]
g
Lock [154]
h
Wake and Rumer [155]
i
Scott [156]
j
Pivovarov [157]
k
Greene [158]
l
Ashton [145]
July 5, 1996, in Swan Lake (Fig. 11.11) indicate that horizontal variations of water
temperature at different depths are typically much smaller than vertical temperature
difference (stratification) from surface to bottom; therefore, one-dimensional assump-
tion in lake modeling is valid. An example of simulated versus observed water
temperatures and DO concentrations in Minnesota lakes is given in Fig. 11.12.
Table 11.2 Parameters and coefficient values in the dissolved oxygen model [134, 147]
Range and Selected
Coefficients and symbols Units references value
Independent of trophic status
BOD decay coefficient kb day1 0.02–3.4a 0.1
Respiration rate coefficient kr day1 0.05–0.5a 0.1
BOD temperature adjustment yb – 1.047a 1.047
Photosynthesis temperature yp – 1.066b 1.036
adjustment
Respiration temperature yr – 1.045c, 1.047d 1.047
adjustment
Sediment temperature adjustment ys – 1.034–1.13e 1.065
Respiration ratio YCHO2 – 0.0083f 0.0083
3 1
Water column oxygen demand WOD g m day 0.01 0.01
during winter
Dependent on trophic status
Coefficients and symbols Units Eutrophic Mesotrophic Oligotrophic
Oxygen equivalent BOD mg l1 1.0g 0.5 0.2
Chlorophyll-a Chl-a mg m3 15h 6 2
SOD during open-water Sb20 g m2 day1 1.5g 1.25 1.0
season Hmax = 24 m Hmax = 24 m Hmax = 24 m
1.95 1.75 1.2
Hmax = 13 m Hmax = 13 m Hmax = 13 m
2.2 1.9 1.25
Hmax = 4 m Hmax = 4 m Hmax = 4 m
SOD during ice-cover Sb g m2 day1 0.226 0.152 0.075
period
Hmax = 4 m = shallow lake, Hmax = 13 m = medium depth, Hmax = 24 m = deep lake [147]
a
From QUAL2E [160]
b
From [44]
c
From EUTR04 [66]
d
From MINLAKE [123]
e
From [161]
f
From [68]
g
From Fang et al. [134]
h
From Stefan and Fang [147]
Simulation of Sustainable Fish Habitat in Lakes Under

Global Warming
Sustaining fish habitat in lakes and reservoirs is a major resource management

issue. Freshwater fish habitat is constrained by physical, chemical, and biological
attributes that relate to water quality, food supply, and human interference [2].
In lakes, water temperature and dissolved oxygen (DO) concentrations are two
of the most significant water quality parameters affecting survival and growth of
fishes [2, 3]. An increase of atmospheric greenhouse gases is projected to cause
200
Chla-pattern
Percent difference from annual mean
150 Bear Head

Carlos
100
Elephant
Elk
50
Hill
Kabekona
0
South twin
Carrie
–50
White Iron
–100
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Fig. 11.10 Seasonal chlorophyll-a distribution specified as MINLAKE2010 model input for
mesotrophic lakes. Generic distribution (solid line) and field data from nine study lakes
TEMPERATURE 7-5-96
Temperature (°C)
0 5 10 15 20 25
0
2
4
6
Depth (m)
8
10
12
14
16
18
20
101 102 103 Simulated
Fig. 11.11 Example of simulated (solid line) and three measured (sites 101,102,103) temperature
profiles in Swan Lake, Minnesota
climate warming, which would alter water temperature and DO characteristics in

lakes. These changes are in turn expected to have an effect on indigenous fish
populations [165–169]. For example, cold-water species typically have physiologi-
cal optima at temperatures near 15 C and are generally not found where summer
water temperatures are higher than 20–26 C [170]. Can fish habitat, especially of
cold-water species, be sustained under global warming? Cold-water fish habitat in
lakes can be reduced in several ways: by direct warming of the water, by increased
32
30 Bear Head Elephant
Big Trout Elk
28 Blue Fish Hook
Measured Water Temperature (°C)
26 Burntside G. Wood
24 Carlos G. Stone
22 Carrie Hill
Cedar Kabekona
20
18
16
14 S. Center
Snowbank
12 L. Sand S. Twin
10 L. Trout St. Olaf
Madison Ten Mile
8
Mukooda Trout (Cook)
6 Siseebakwet Trout (S. Louis)
4 Six White Iron
2
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32
Simulated Water Temperature (°C)
20
Bear Head Elephant
18 Big Trout Elk
Fish Hook
Measured Dissolved Oxygen (mg / l)
Blue
16 Burntside G. Wood
Carlos G. Stone
14 Carrie Hill
Cedar Kabekona
12
10
8 L. Sand S. Center
L. Trout S. Twin
6 Madison St. Olaf
Mukooda Ten Mile
4 Siseebakwet Trout (Cook)
Six Trout (S. Louis)
2 Snowbank White Iron
0
0 2 4 6 8 10 12 14 16 18 20
Simulated Dissolved Oxygen (mg / l)
Fig. 11.12 Simulated versus measured temperature and dissolved oxygen (DO) in 28 Minnesota
lakes. The solid line (1:1) would indicate perfect agreement
hypolimnetic oxygen depletion during periods of stratification, and by thermocline

deepening [165, 171, 172]. Can potential refuge lakes and reservoirs be identified
and protected?
Water temperatures and DO concentrations in Minnesota lakes under several
projected climate scenarios were previously simulated for the open-water season
and the entire year [140, 173], and the results were used to estimate potential future
fish habitats [166–169, 174]. Fish habitat estimated from daily temperature and
DO profiles under past climate conditions was validated against fish observations
[175, 176]. In these studies, fish species were grouped into three thermal guilds –
cold-water, coolwater, and warm-water – that differ in the optimal temperature
range for physiological function and ecological success.
In a follow-up study, habitat for only one fish species was investigated. Cisco,
Coregonus artedi, is the most common cold-water stenothermal fish in Minnesota
lakes. The combination of a wide geographic distribution and a requirement for
cold, oxygenated water make cisco an excellent “canary in a mineshaft” species that
is a sensitive indicator of climate change.
After daily water temperature and DO were simulated under past and future
climate scenarios, two approaches were explored to determine if cisco habitat
existed in a lake. The first approach was based on separate water temperature and
DO constraints; this approach had previously been used to examine the potential
impact of future climate warming on cold-water, coolwater, and warm-water fish
species (guilds) in small lakes in the contiguous United States [166].
The second approach used a single oxythermal habitat variable to define suitable
or unsuitable fish habitat. Jacobson et al. [177] developed a generalized oxythermal
habitat variable, TDO3, called “temperature at 3.0 mg/L DO.” It is determined by
interpolating the water temperature at a benchmark oxygen concentration (i.e.,
3.0 mg/L) from vertical temperature and DO profiles in a lake. DO = 3.0 mg/L is
an oxygen concentration limit that is probably lethal or nearly so for many cold-water
species [2, 178, 179]. TDO3 has a strong connection with four cold-water taxa (lake
trout, cisco, whitefish, and burbot) in Minnesota lakes. Cisco were present in lakes
with a broad range of maximum TDO3 values, with central borders of 4.0–16.9 C
[177]. TDO3 allows to evaluate or quantify which lake is a better refuge lake for cisco
under a future climate scenario.
Simulation of Cisco Habitat by the Constant Values

Method (CVM)
Fish habitat for cold-water fish guilds (e.g., cisco) was estimated from simulated
daily water temperature (T) and DO profiles in lakes, similar to the approach
by Christie and Regier [3]. Temperature and DO criteria for survival and good
growth of a fish species were applied to the simulated year-round daily water
temperature and DO profiles, as shown schematically in Fig. 11.13. This sche-
matic figure is for a lake where cold-water fish cannot be present during much of
the open-water season. On days when the isotherm of lethal temperature (LT)
and the DO limit isopleth for a fish species intersect, the entire depth of
a stratified lake is uninhabitable. When the maximum water temperature is
lower than the lethal temperature (LT) everywhere in a lake, the isotherm for
LT will not show up. In this case, the DO survival limit becomes the only
survival criterion; this occurs especially during the winter ice-cover period, as
shown in Fig. 11.14.
GSL1 Cisco kill GSL2

GSB NSB NSL NSE GSE Area Volume
T > LT
LT
UG
GT
Depth (m)
LG
G
T
DO < 3 mg/L
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Uninhabitable Good growth Restricted growth
Fig. 11.13 Depth versus time plot of the critical isotherm and DO isopleth, showing projected
cisco kill in a seasonally stratified lake as dark shaded areas. Habitat parameters (e.g., NSL, GSL)
are discussed by Stefan et al. [166]
0
2 Madison Lake 1992 1997 1999 2006
4
Depth (m)
6
8
10
12
14
16
F M A M J J A S O N D F M A M J J A S O N D F M A M J J A S O N D F M A M J J A S O N D
Fig. 11.14 Depth versus time contours of cisco fish habitat in Madison Lake for selected years.
Thin lines are isotherms and indicate a temperature constraint for fish habitat in the upper part of
the stratified lake; bolder lines are DO isopleths that indicate a dissolved oxygen constraint for fish
habitat in the lower portion of the lake
In addition to the lethal temperature (LT) isotherm, Fig. 11.13 also gives the
upper good-growth temperature (UGGT) and the lower good-growth temperature
(LGGT) isotherms. Between these isotherms/isopleths, three fish habitats are
identified:
1. Uninhabitable space if temperature is above or DO is below the survival limit
2. Good-growth habitat if temperature is between the upper and lower good-growth
limits (i.e., LGGT < T < UGGT) and DO is above the survival limit
3. Restricted growth habitat if temperature is above the upper good-growth tem-
perature but below the survival limit (i.e., UGGT < T < LT), or if temperature is
below the lower good-growth limit of temperature (i.e., T < LGGT) and DO is
above the survival limit
Projections on growth habitat and survival parameters of three fish guilds (cold-
water, coolwater, and warm-water) in the U.S. were reported by Stefan et al. [166]
and Fang et al. [167–169].
The LT is the water temperature to which fish cannot be acclimated without

causing death. The LT = 23.4 C [170] used for the cold-water fish guild in previous
studies [166] was the mean value of LT values for ten cold-water fish species (pink
salmon, sockeye salmon, chinook salmon, chum salmon, coho salmon, brown trout,
rainbow trout, brook trout, lake trout, mountain whitefish).
The DO = 3.0 mg/L requirement for the cold-water fish guild, below which
mortality is more likely to occur or growth is impaired [179], was developed from
an available US EPA database [180]. The DO = 3.0 mg/L was for the open-water
season. In ice-covered shallow, eutrophic or mesotrophic lakes DO can drop
below 2.0 mg/L near the end of an extended ice-cover period due to bottom
(sediment) and water column oxygen demands. For ice-covered lakes, DO
criteria for fish survival could be set at lower values than 3.0 mg/L, but data
from previous studies are still inadequate to establish specific tolerance limits of
DO for fish winterkill [181]. The use of a single year-round DO criterion had no
adverse effect on projections of winterkill under a 2 CO2 climate scenario; no
winterkill was projected despite using higher DO limits [166, 182].
Figure 11.14 is an example of contour plots in Madison Lake, Minnesota, which
has a maximum depth of 18.0 m, surface area of 4.5 km2, and mean summer Secchi
depth of 0.88 m (eutrophic lake). Isotherms at the lethal temperature of 23.4 C are
drawn as thin black lines, and isopleths at the DO limit of 3.0 mg/L are drawn as
thicker red lines. The CVM projected fish kill in 42 out of 47 simulation years (1961
to 2008), summer kill in 1999 and 2006, and winterkill in 1997 in Fig. 11.14 are
samples because 2006 was the warmest year and 1997 was the coldest year in the
simulation period. No fish kill was projected in 1992, as shown in Fig. 11.14. The
maximum continuous kill days in Madison Lake were projected to last 34 days and
occurred in 2006. Madison Lake is classified as a non-cisco lake by Minnesota
Department of Natural Resources. The maximum continuous kill days in 15 lakes
studied were projected to occur either in 2006 for summer kill (with a few
exceptions) or in 1996 or 1997 for winterkill (the coldest 2 years of the simulation
period).
Simulation of Cisco Habitat by the Equation Method (EM)
Jacobson et al. [177] developed an equation for the lethal-niche boundary of adult
cisco by remapping the measured DO concentrations and temperatures from the
profiles in 16 Minnesota lakes that experienced cisco mortality in midsummer of
2006. The equation is given as
DOlethal ¼ 0:40 þ 0:000006e0:59Tlethal (11.6)
where DOlethal and Tlethal are the dissolved oxygen concentration and the water
temperature that define the lethal-niche boundary. DOlethal is the required or needed
20
18 16 Mar South Center Lake 01 Jul South Center Lake 21 Jul South Center Lake
01 Apr 2004 04 Jul 2006 23 Jul 2007
16
12 Jul 08 Jul 25 Jul
14
15 Jul
DO (mg/l)
12
10
8
6
4
2
0
0 4 8 12 16 20 24 28 0 4 8 12 16 20 24 28 0 4 8 12 16 20 24 28
Water Temperature (°C) Water Temperature (°C)
Fig. 11.15 Simulated DO concentration versus simulated temperature in South Center Lake on
selected summer days in 2006 with potential cisco mortality against the lethal-niche-boundary
curve developed by Jacobson et al. [177]
minimum DO concentration at a given temperature Tlethal for cisco to survive.

Equation 11.6 gives a DO survival limit for cisco that is not constant but instead
depends on water temperature. The Equation Method (EM), therefore, uses survival
limits for temperature and DO that are variable and related to each other.
The study by Jacobson et al. [177] leading to the EM was done specifically for
cisco, whereas the study by Eaton et al. [170] that lead to the CVM was done on the
entire cold-water fish guilds and not specifically on cisco (though cisco is a cold-
water fish).
Equation 11.6 was implemented in a fish habitat simulation program to deter-
mine potential fish mortality in 15 study lakes. For each simulation day, the
required DO concentrations, DOlethal, were computed from simulated water
temperatures in all water layers using Eq. 11.6 and compared with DO
concentrations simulated by MINLAKE2010 in the same layers. Fish kill was
assumed to occur if the simulated DO was less than the DOlethal value at all water
depths on that day. If simulated DO was larger than DOlethal in some of the water
layers, fish mortality was not assumed to occur because cisco could swim to the
water layers with suitable DO and temperature conditions.
Sample plots of simulated DO concentrations versus simulated temperatures at
the same location in four study lakes on selected summer days when cisco mortality
(kill) potential was high were developed. The lethal-niche-boundary curve
(Eq. 11.6) developed by Jacobson et al. [177] was added for reference. Examples
of results for South Center Lake are shown in Fig. 11.15. All simulated data points
in Fig. 11.15 are located to the right of the curve, indicating that adult cisco kill
could indeed occur in all water layers (depths) on the days shown [183]. Fish kill of
adult cisco was simulated to occur also on some days in four other lakes (Carrie
Lake, Hill Lake, White Iron Lake, and Madison Lake) (Fang et al. [183]), and this
result agreed with field observations because no cisco had been found in those five
lakes. This means that these lakes are not sustainable to cisco or potential other
cold-water fish species.
Projection of Sustainable Cold-Water Fish Habitat in Lakes

Under Future Climate Scenarios
Fish habitat projections under future climate scenarios were made for cisco in
Minnesota lakes using the two methods described above and three future climate
scenarios based on the output of three GCMs of the global atmosphere: CCCma
GCM 2.0, CCCma CGCM 3.1 (A1B scenario), and MIROC 3.2 (A1B scenario)
from the Special Report on Emission Scenarios of the IPCC (2001). Descriptions of
the evolving stages of the GCMs and examples of applications were given by Boer
et al. [184], McFarlane et al. [185], and Kim et al. [186, 187]. MIROC was
described by Hasumi and Emori [188]. One output of the cisco habitat simulations
was the projected number of annual cisco kill days; these were summarized and
interpreted for the different climate scenarios.
Results obtained by both the CVM and the EM showed a considerable increase
in the number of annual cisco kill days under future climate scenarios. Projections
of the number of annual kill days under all three future climate scenarios were
consistent. The GCM 2.0 scenario projected higher numbers of cisco kill days
because the future atmospheric CO2 concentration was higher for this earlier
climate scenario model. Future climate scenarios based on the two more recent
CGCMs (CCCma CGCM 3.1 and MIROC 3.2) gave almost the same numbers of
projected cisco kill days [183].
As a preliminary conclusion, 19 out of 21 simulated cisco lakes were projected
to be viable cisco refuge lakes under a warmer climate. In lakes not supporting cisco
habitat, there is a shift in cisco kill days from winter to summer.
Figure 11.16 shows the distribution of 620 current cisco lakes in Minnesota and
21 selected cisco study lakes in a coordinate system of lake geometry ratio (GR)
versus Secchi depth (SD). Under the CCCma CGCM 3.1 future climate scenario, 19
of the 21 simulated cisco lakes were projected to have no cisco kill under future
climate scenarios. There is a region on the (SD) versus (GR) plot in Fig. 11.15 with
“refuge” lakes that can support cisco habitat under future climate scenarios. Such
refuge lakes should have an SD greater than 2.5 m (mesotrophic or oligotrophic
lakes) and a GR less than 3.0 m0.5 (typically seasonally stratified, dimictic lakes).
Once refuge lakes are identified, watershed protection efforts can be initiated at
refuge lakes to prevent deterioration of water quality by anthropogenic activities.
The findings in Fig. 11.16 were confirmed by projections using the
oxythermal habitat parameter TDO3. Jacobson et al. [177] proposed a single
variable to quantify oxythermal habitat that allows for comparison across
several cold-water fish species (lake trout, cisco, lake whitefish, and burbot)
that have different requirements for cold, oxygenated water. The single
generalized oxythermal habitat variable is defined as the water temperature at
3 mg/L of DO and is called TDO3. The TDO3 can be determined by
interpolating the temperature of water at the DO concentration of 3 mg/L
from measured or simulated vertical temperature and DO profiles. When
10.0
620 cisco lakes
9.0 19 cisco lakes without kill
2 cisco lakes with kill
8.0 Geometry ratio < 3.0 m–0.5
Secchi depth > 2.5 m
7.0
Secchi Depth (m)
6.0
5.0
4.0
3.0
2.0
1.0
0.0
0.4 0.6 1.0 2.0 4.0 6.0 10.0 20.0
Lake Geometry Ratio (m–0.5)
Fig. 11.16 Distribution of 21 study lakes among 578 current cisco lakes in Minnesota projecting
no future cisco kill in 19 and potential cisco kill in two study lakes under the CCCma CGCM 3.1
future climate scenario
nonmonotonic profiles generate low oxygen concentrations with more than one
TDO3 value, the coldest TDO3 is used [177].
The oxythermal parameter TDO3 measures environmental conditions and stress
in a lake. Low values of summer TDO3 indicate excellent oxythermal habitat for
cold-water fish, i.e., fish have a wide range of temperatures available in the
hypolimnion with sufficient oxygen concentrations. High values of TDO3 indicate
poor oxythermal habitat for cold-water fish, with little or no cold water with
sufficient oxygen. Very high values of TDO3 indicate hypolimnia that are anoxic
or are found in unstratified lakes. The conclusion of the fish habitat study [183] was
that under fairly stringent selection criteria (DO > 3 mg/L and TDO3 17 C), at
least one fourth to one third of the lakes that currently have cisco populations are
projected to maintain cisco habitat under projected future climate scenarios. Many
of the “refuge” lakes are located in northeastern and central Minnesota.
Refuge lakes with TDO3 < 17 C under the future climate scenario MIROC 3.2
have Secchi depths greater than 2.5 m, lake geometry ratios less than 2.5 m,
maximum depths greater than 15 m, and surface areas less than 30 km2 [183].
Cisco refuge lakes were also ranked in narrow (2 C) bands of TDO3 values
(Fig. 11.17). AvgATDO3VB in Fig. 11.17 is the average of mean daily TDO3 values
over the fixed 31-day benchmark period (Julian day 209–239) in the simulation
period (one value for 1962–2008 period). Names and locations of lakes in each
11
TDO3 contour lines
10 Cisco lakes between TDO3 lines
13
9
8
15
Secchi Depth (m)
17
29
25
6
21
19
27
9
23
5
11
25
13
17
19
15
4
11
21
3
0
0.3 0.5 1.0 2.0 4.0 10.0
Geometry Ratio (m-0.5)
Fig. 11.17 Distribution of 620 cisco lakes in Minnesota between contour lines of AvgATDO3VB
for variable benchmark periods simulated for the MIROC 3.2 future climate scenario using Duluth
weather data
band were given to the Minnesota Department of Natural Resources for manage-
ment decisions. Lakes with the lowest TDO3 values were the best candidates for
cisco refuge lakes because a low TDO3 value is presumed to reflect the least stress
on adult cisco.
Future Directions
There are many unanswered questions regarding transport and transformation

processes that control water quality and habitat in lakes or reservoirs because
these processes are numerous and often interacting. This lack of knowledge and
information hinders meaningful decision making on lake and reservoir manage-
ment and can lead to waste of effort and resources. Examples where future research
efforts can pay sustained benefits are:
1. Continuous automated data collection in lakes and reservoirs. Existing and
future sensors of physical, chemical, and biological lake parameters will enable
the collection and recording of time series of information at high temporal and
spatial resolution. The installation of measuring stations can be expensive, but
the return will be information that can really decipher the timescales and
vertical scales of processes.
2. Development of simulation models for critical lakes as a permanent tool for

decision making. Existing models can be tailored to specific, valuable lakes and
can be used first to simulate the outcome of contemplated actions. In the long-
term, model development, modification and validation can benefit greatly when
paired with automated data collection advocated above (1).
3. Research on the intermediate-scale coupling between the atmosphere and the
hydrosphere of lakes and reservoirs. A missing link in our understanding of
lakes is their coupling to the atmospheric boundary layer. Atmospheric bound-
ary layers over lakes have been all but ignored roughness transitions from the
land to water and the transfer of momentum from wind, and the transfer rates of
gases at lake and reservoir surfaces have been measured and analyzed as if the
atmospheric boundary layer over a lake surface did not have to respond to
wakes of canopies, buildings, and hills, and adjust to the roughness transition
from land to water.
4. Research on the large-scale coupling between the atmosphere and the hydro-
sphere of lakes and reservoirs. “Large-scale” here means responses to climate
(warming). Climate/weather effects on heat transfer at a water surface and
the consequences for stratification and mixing can be estimated. However, the
intensification of the hydrologic cycle by global warming has effects on lake
and reservoir water budgets, on material import from watersheds, and material
budgets of lakes and reservoirs; it also affects internal material transport and
transformation processes, and thereby habitat in ecosystems.
5. Interactions and exchange processes at the sediment–water interface are highly
diverse and difficult to model in a comprehensive way. Flow of water in the
boundary layer can be highly intermittent, possibly density stratified, and
possibly affected by sediment settling and/or sediment resuspension. The
porous sediment surface allows for some exchange of mass and momentum
but also much damping. The pore scale is far shorter than the internal lake
modeling scale, and the chemistry of the waters in the sediment pore system
can be complex; sediment particle interactions with solutes can be significant,
and there are also microbial biofilms. At present, the fluxes of materials across
the sediment interface that are needed as boundary conditions in lake water
quality models are mostly estimated empirically. Processes in the flowing water
boundary layer above the sediment layer and in the sediment pore system need
to be better coupled to in lake models.
6. Microbial processes and solute–particle interactions in lakes may have far more
impact on water quality and material fluxes than is acknowledged in current lake
models. The microstructure of lake and reservoir stratification needs to be better
understood because it controls some of these processes.
7. Criteria used to assess fish habitat in lakes are currently fairly crude. What are
the proper timescales? Criteria need to be established for different life stages of
fish, not only adult fish. More quantitative information on the response of fish to
environmental parameters has a bearing on fish survival projections under
environmental stress such as global warming.
8. Lake and reservoir management, including goal-setting, data collection and

storage, planning, decision making, resource allocation, and implementation,
are complicated not only because lakes and reservoirs are complicated but also
the multitude of agencies and organizations that are in charge gives an adminis-
trative and organizational structure that is not optimal. A single, multidisciplinary
entity in charge of a lake or reservoir may do a better job than the compromises
developed under current administrative structures. Simplification of the organiza-
tional structure and the resulting decision-making process are appropriate.
9. One of the many management issues has to do with internal nutrient loading.
Many TMDL studies of external inputs to lakes and reservoirs accomplish little
when lake ecosystems function on the basis of internal nutrient recycling.
10. Biomanipulation is practiced in a very empirical way and needs a better basis.
11. Lake and reservoir improvement technologies need to be reviewed and can
probably be improved. Guidance manuals need to be developed for these
techniques.
Bibliography
Primary Literature
1. Chapra SC (1997) Surface water-quality modeling. McGraw-Hill, New York

2. Fry EFJ (1971) The effect of environmental factors on the physiology of fish. Academic, New
York
3. Christie CG, Regier HA (1988) Measurements of optimal habitat and their relationship to
yields for four commercial fish species. Can J Fish Aquat Sci 45:301–314
4. Zipparro VJ, Hasen H (eds) (1993) Handbook of applied hydraulics. McGraw-Hill, New
York
5. USBR (U. S. Bureau of Reclamation) (1987) Design of small dams. U. S. Govmt. Printing
Office, Denver
6. Novák P, Nalluri C (1990) Hydraulic structures. Unwin Hyman, London
7. Press H (1958) Talsperren (in German), 2nd edn. Ernst & Sohn, Berlin
8. Thornton KW, Kimmel BL, Payne FE (1990) Reservoir limnology. Wiley, New York
9. Ackerman WC, White GF, Worthington EB (eds) (1973) Man-made lakes: their problems
and environmental effects. American Geophysical Union, Washington, DC
10. Stefan HG (ed) (1980) Surface water impoundments. ASCE, New York
11. NRC (national Research Council) (1987) Glen canyon environmental studies review com-
mittee. National Academy, Washington, DC
12. Hutchinson GE (1957) A treatise on limnology. Wiley, New York
13. Frey DG (1966) Limnology in North America. The University of Wisconsin Press, Madison
14. Lerman A, Baccini P (1978) Lakes – chemistry, geology, physics. Springer, New York
15. Wetzel RG (1983) Limnology, 2nd edn. WB Saunders, Philadelphia
16. Horne A, Goldman C (1994) Limnology. McGraw-Hill, New York
17. Lampert W, Sommer U (1997) Limnoecology: the ecology of lakes and streams. Oxford
University Press, New York
18. Henderson-Sellers B (1984) Engineering limnology. Pitman Advanced Publication Program,
Boston
19. Straskraba M, Gnauck AH (1985) Freshwater ecosystems. Modeling and simulation.

Elsevier, Amsterdam
20. Rouse H, Ince S (1957) History of hydraulics. Iowa Institute of Hydraulic Research, Iowa
City
21. Chien N, Wang Z (1998) Mechanics of sediment transport. ASCE Press, Reston
22. Dortch MS (1992) Modeling water quality of reservoir tailwaters, Technical Report W-92-1.
US Army Corps of Engineers, Waterways Experiment Station, Vicksburg
23. Fischer HB (1979) Mixing in inland and coastal waters. Academic, New York
24. Gray WG (1986) Physics-based modeling of lakes, reservoirs, and impoundments. ASCE
Press, Reston
25. Herb WR, Janke B, Mohseni O, Stefan HG (2009) Runoff temperature model for paved
surfaces. J Hydrol Eng 14:1146–1155
26. Wanielista MP, Yousef YA (1993) Stormwater management. Wiley, New York
27. Hino M (1994) Water quality and its control. International Association for Hydraulic
Research, AA Balkema, Rotterdam
28. Baker LA (1984) A model for phosphorus cycling in lakes. International Memorandum No.
100, St Anthony Falls Laboratory, University of Minnesota, Minneapolis
29. Stefan HG, Dhamotharan S, Schiebe FR, Fu AY, Cardoni JJ (1989) Dynamic simulation of
turbidity and its correction in Lake Chico, Arkansas. In: Henderson-Sellers B (ed) Water
quality modeling, vol IV. CRC Press, Baco Raton
30. Novotny EV, Stefan HG (2010) Projections of chloride concentrations inurban lakes receiv-
ing road de-icing salt. Water Air Soil Pollut 211:261–271
31. Gassman PW (2007) The soil and water assessment tool: historical development,
applications, and future research direction. Trans ASABE 50(4):1211–1250
32. James W, Huber W, Rossman L et al (2005) Water systems models: user’s guide to SWMM.
CHI, Guelph
33. Rossman L (2008) Storm Water Management Model. User’s manual Version 5.0. US Environ-
mental Protection Agency, Cincinnati
34. USGS (2008) SNTEMP stream temperature model, http://smig.usgs.gov/cgi-bin/SMIC/
model_home_pages/model_home?selection=sntemp
35. Herb WR, Mohseni O, Stefan HG (2009) Simulation of temperature mitigation by
a stormwater detention pond. J Am Water Resour Assoc 45:1164–1178
36. PLUARG (1978) Environmental management strategy for the great lakes system. Final
Report of the Pollution from Land Use Activities Reference Group (PLUARGE) to the
International Joint Commission. Windsor, Ontario
37. Hornbuckle KC, Jeremiason JD, Sweet CW, Eisenreich SJ (1994) Seasonal variations in
air-water exchange of polychlorinated biphenyls in Lake Superior. Environ Sci Technol
28:1491–1501
38. Gulliver JS (2007) Introduction to chemical transport in the environment. Cambridge Uni-
versity Press, Cambridge
39. Dhamotharan S, Gulliver JS, Stefan HG (1981) Unsteady one-dimensional settling of
suspended sediment. Water Resour Res 17:1125–1132
40. Johnson TC (1980) Sediment redistribution by waves in lakes, reservoirs and embayments.
In: Stefan HG (ed) Proceeding of international symposium on surface water impoundements.
ASCE, Reston, Minneapolis, pp 1307–1317
41. Sheng YP, Lick W (1979) The transport and resuspension of sediments in a shallow lake.
J Geophys Res 84:1809–1826
42. Rodney MW, Stefan HG (1987) Conceptual model for wind-generated sediment resuspen-
sion in shallow ponds. In: Proceedings of the national symposium on mining, hydrology,
sedimentology and reclamation. University of Kentucky, Lexington, pp 263–269
43. Erdmann JB, Stefan HG, Brezonik PL (1994) Analysis of wind and ship induced sediment
resuspension in Duluth-Superior harbor. J Am Water Resour Assoc 30:1043–1053
44. Thomann RV, Mueller JA (1987) Principles of surface water quality modeling and control.
Harper & Row, New York
45. Clark MM (1996) Transport modeling for environmental engineers and scientists. Wiley,
New York
46. Weber W Jr, DiGiano FA (1995) Process dynamics in environmental systems. Wiley, New
York
47. Weilenmann U, O’Melia CR, Stumm W (1989) Particle transport in lakes: models and
measurements. Limnol Oceanogr 34:1–18
48. Mackenthun AA, Stefan HG (1998) Effect of flow velocity on sediment oxygen demand:
experiments. J Environ Eng 124:222–230
49. Qian Q, Clark JJ, Voller VR, Stefan HG (2009) Modeling of vertical dispersion of a solute in
a stream or lake bed enhanced by wave induced interstitial flow. J Am Water Resour Assoc
45:343–354
50. Qian Q, Clark JJ, Voller VR, Stefan HG (2009) Depth-dependent dispersion coefficient for
modeling of vertical solute exchange in a lake bed under surface waves. J Hydraul Eng ASCE
135:187–197
51. Elliott AH, Brooks NH (1997) Transfer of non-sorbing solutes to a streambed with bed forms:
theory. Water Resour Res 33:123–136
52. Packman AI, Salehin M, Zaramella M (2004) Hyporheic exchange with gravel beds: basic
hydrodynamic interactions and bedform-induced advective flows. J Hydraul Eng
130:647–656
53. Roy H, Huttel M, Jorgensen BB (2002) The role of small-scale sediment topography for
oxygen flux across the diffusive boundary layer. Limnol Oceangr 47:837–847
54. Steinberger N, Hondzo M (1999) Diffusional mass transfer at sediment-water interface.
55. O’Connor BL, Hondzo M (2008) Dissolved oxygen transfer to sediments by sweep and eject
motions in aquatic environments. Limnol Oceanogr 53:566–578
56. Nakamura Y, Stefan HG (1994) Effect of flow velocity on sediment oxygen demand: theory.
57. Higashino M, Clark JJ, Stefan HG (2009) Pore water flow due to near-bed turbulence and
associated solute transfer in a stream or lake sediment bed. Water Resour Res 45
58. Thamdrup B, Fossing H, Jorgensen BB (1994) Manganese, iron and sulfur cycling in
a coastal marine sediment, Aarhus Bay, Denmark. Geochimica et Cosmochimica Acta
58:5115–5129
59. O’Connor BL, Harvey JW (2008) Scaling hyporheic exchange and its influence on biogeo-
chemical reactions in aquatic ecosystems. Water Resour Res 44:W12423
60. Mortimer CH (1971) Chemical exchanges between sediments and water in the Great
Lakes-speculations on probable regulatory mechanisms. Limnol Oceanogr 16:387–404
61. Lorenzen CJ, Welschmeyer NA, Copping AE, Vernet M (1983) Sinking rates of organic
particles. Limnol Oceanogr 28:766–769
62. Stefan HG, Hanson MJ (1981) Phosphorus recycling in five shallow lakes. J Environ Eng Div
107:713–730
63. Boudreau BP, Jørgensen BB (2001) The benthic boundary layer: transport processes and
biogeochemistry. Oxford University Press, Oxford
64. DiToro DM (2001) Sediment flux modeling. Wiley, New York
65. Manous JD, Gantzer CJ, Stefan HG (2007) Spatial variation of sediment sulfate reduction
rates in a saline lake. J Environ Eng-ASCE 133:1106–1116
66. Ambrose RB, Wool TA, Connolly JP, Schanz RW (1988) WASP (EUTRO4),
A hydrodynamic and water quality model – model theory, user’s manual, and programmer’s
guide. US Environmental Protection Agency, Athens, Georgia
67. Cole TM, Wells SA (2004) CE-QUAL-W2: a two-dimensional, laterally averaged, hydrody-
namic and water quality model, version 3.2, user’s manual. Environmental Laboratory, US
Army Engineer Research and Development Center, Vicksburg
68. Stumm W, Morgan JJ (1981) Aquatic chemistry. Wiley-Intersciences, New York

69. Stumm W, Morgan JJ (1996) Aquatic chemistry: chemical equilibria and rates in natural
waters, 3rd edn. Wiley, New York
70. Gobas FAPC, McCorquodale JA (1992) Chemical dynamics in fresh water ecosystems.
Lewis, Chelsea
71. Brezonik PL (1993) Chemical kinetics and process dynamics in aquatic systems. Lewis, Boca
Raton
72. Zison SW, Mills WB, Diemer D, Chen C (1978) Rates, constants, and kinetics
formulations in surface water quality modeling. Tetra Tech Incorporation, SEPA, ORD,
Athens, Georgia
73. Bowie GL, Mills WB, Porcella DB et al (1985) Rates, constants, and kinetics formulations in
surface water quality modeling (2nd edn), EPA/600/3-85/040. Environmental Research
Laboratory, Office of Research and Development, US Environmental Protection Agency
74. Novak P, Arnold W, Blazer V et al (2011) On the need for a national (U.S) research
program to Elucidate the potential risks to human health and the environment posed by
contaminants of emerging concern. Environ Sci Technol 45:3829–3830
75. Calabrese EJ (1985) Age and susceptibility to toxic substances. Wiley, New York
76. Scheffer M (1998) Ecology of shallow lakes. Chapman and Hall, London
77. Cooke GD (1993) Restoration and management of lakes and reservoirs. Lewis, Boca Raton
78. Reckhow KH, Chapra SC (1983) Engineering approaches for lake management, vol 1, Data
analysis and empirical modeling. Ann Arbor Science, Boston
79. Henderson-Sellers B (1991) Water quality modeling, vol IV, Decision support techniques for
lakes and reservoirs. CRC Press, Boca Raton
80. Burns F, Powling I (1981) Destratification of lakes and reservoirs to improve water quality.
Australian Government, Canberra
81. Fast AW (1971) The effects of artificial aeration on lake ecology. US Environmental
Protection Agency, Washington, DC
82. Goossens L (1979) Reservoir destratification with bubble columns. Delft University Press,
Delft
83. Zic K, Stefan H (1988) Lake Aerator effect on temperature stratification analyzed by
“MINLAKE” model. Lake Reserv Manag 4:85–90
84. Zic K, Stefan H, Ellis C (1992) Laboratory study of water destratification by a bubble plume.
J Hydraul Res 30:7–27
85. Gulliver JS, Stefan HG (1982) Lake phytoplankton model with destratification. J Environ
Eng Div-ASCE 108:864–882
86. Ellis CR, Stefan HG (1990) Hydraulic design of a winter lake aeration system. J Environ
Engin-ASCE 116:376–393
87. Stefan HG, Bender MD, Shapiro J, Wright DI (1987) Hydrodynamic-design of
a metalimnetic lake aerator. J Environ Eng-ASCE 113:1249–1264
88. Stefan HG, Gu R (1992) Efficiency of jet-mixing of temperature-stratified water. J Environ
Eng ASCE 118:363–379
89. Stefan HG, Gu R (1990) Jet mixing in lake or reservoir stratification simulations. Lake Reserv
Manag 6:165–174
90. Stefan HG, Gu RC (1991) Conceptual design procedure for hydraulic destratification systems
in small ponds, lakes, or reservoirs for water-quality improvement. Water Resour Bull
27:967–978
91. Hanson MJ, Stefan HG (1984) Side effects of 58 years of copper sulfate treatment of the
fairmont lakes, Minnesota. Water Resour Bull 20:889–900
92. Vollenweider R (1968) Scientific fundamentals of the eutrophication of lakes and flowing
waters, with particular reference to phosphorus and nitrogen as factors in eutrophication.
Organisation for Economic Cooperation and Development Technical Report OAS/CSI/68.27,
Paris
93. Gorham E, Boyce FM (1989) Influence of lake surface area and depth upon thermal stratifi-
cation and the depth of the summer thermocline. J Great Lakes Res 15:233–245
94. Alavian V, Jirka GH, Denton RA, Johnson MC, Stefan HG (1992) Density currents entering
lakes and reservoirs. J Hydraul Eng 118:1464–1489
95. Akiyama J, Stefan HG (1984) Plunging flow into a reservoir – theory. J Hydraulic Eng ASCE
110:484–499
96. Akiyama J, Stefan H (1985) Turbidity current with erosion and deposition. J Hydraul Eng
111:1473–1496
97. Akiyama J, Stefan HG (1987) Onset of underflow in slightly diverging channels. J Hydraul
Eng ASCE 113:825–844
98. Akiyama J, Stefan HG (1988) Turbidity-current simulation in a diverging channel. Water
Resour Res 24:579–587
99. Johnson TR, Ellis CR, Farrell GJ, Stefan HG (1987) Negatively buoyant flow in a diverging
channel. Part 2: flow field description. J Hydraul Eng ASCE 113:731–742
100. Johnson TR, Farrell GJ, Ellis CR, Stefan HG (1987) Negatively buoyant flow in a diverging
channel. Part1: flow Regimes. J Hydraul Eng ASCE 113:716–730
101. Farrell GJ, Stefan HG (1988) Mathematical-modeling of plunging reservoir flows. J Hydraul
Res 26:525–537
102. Farrell GJ, Stefan HG (1989) Two-layer analysis of a plunging density-current in a diverging
horizontal channel. J Hydraul Res 27:35–47
103. Fang X, Stefan HG (2000) Dependence of dilution of a plunging discharge over a sloping
bottom on inflow conditions and bottom friction. J Hydraul Res 38:15–25
104. Ellis CR, Champlin J, Stefan HG (1997) Density current intrusions in an ice-covered urban
lake. J Am Water Resour Assoc 33:1363–1374
105. Stefan HG (1970) Modeling spread of heated water over lake. J Power Div 96:469–482
106. Stefan H, Schiebe FR (1970) Heated discharge from flume into tank. J Sanit Eng Div
96:1415–1433
107. Stefan HG (1970) Stratification of flow from channel into deep lake. J Hydraul Div
96:1417–1434
108. Stefan HG, Chu CS, Wing H (1972) Impact of cooling water on lake temperatures. J Power
Div 98:253–272
109. Stefan H, Vaidyaraman P (1972) Jet type model for the three-dimensional thermal plume in
a crosscurrent and under wind. Water Resour Res 8:998–1014
110. Demetracopoulos AC, Stefan HG (1983) Model of Mississippi river pool: dissolved
oxygen. J Environ Eng ASCE 109:1020–1034
111. Stefan HG, Anderson KJ (1980) Wind-driven flow in Mississippi river impoundment.
J Hydraul Div 106:1503–1520
112. Demetracopoulos AC, Stefan HG (1983) Model of Mississippi river pool: mass transport.
113. Manous JD, Stefan HG (2004) Sulfate distribution in a multi-basin, saline lake.
Hydrobiologia 529:169–185
114. Stefan H, Ford DE (1975) Temperature dynamics in dimictic lakes. J Hydraul Div ASCE
101:97–114
115. Ford DE, Stefan HG (1980) Stratification variability in three morphometrically different
lakes under identical meteorological forcing. J Am Water ResourAssoc 16:243–247
116. Stefan HG, Megard RO, Skoglund T (1976) Wind control of algae growth in eutrophic lakes.
J Environ Eng Div 102:1201–1213
117. Herb WR, Stefan HG (2005) Model for wind-driven vertical mixing in a shallow lake with
submersed macrophytes. J Hydraul Eng 131:488–496
118. Markfort CD, Perez ALS, Thill JW et al (2010) Wind sheltering of a lake by a tree canopy or
bluff topography. Water Resour Res 46:W03530
119. Stefan HG, Ambrose RB, Dortch MS (1990) Surface water quality models:modeler’s per-
spective. In: Proceedings international symposium on water quality modeling of agricultural
non-point sources, US Department of Agriculture, American Society of Agricultural

Engineers. St. Joseph
120. Wurbs RA (1998) Dissemination of generalized water resources models in the United
States. Water Int 23:190–198
121. Anon (1986) A numerical one-dimensional model of reservoir water qulity. user’s manual.
Instruction Report E-82-1, US Army Corps of Engineers,Vickburg
122. Imberger J, Patterson JC (1981) A dynamic reservoir simulation model – DYRESM. In:
Fischer HB (ed) Transport models for inland and coastal waters. Academic, New York,
Chapter 9
123. Riley MJ, Stefan HG (1988) Minlake: a dynamic lake water quality simulation model. Ecol
Model 43:155–182
124. Middlebrooks EJ, Falkenbory DH, Maloney TE (1976) Modeling the eutrophication
process. Ann Arbor Science, Boston
125. Bruce LC, Hamilton D, Imberger J et al (2006) A numerical simulation of the role of
zooplankton in C, N and P cycling in Lake Kinneret, Israel. Ecol Model 193:412–436
126. Gal G, Hipsey M, Parparov A et al (2009) Implementation of ecological modeling as an
effective management and investigation tool: Lake Kinneret as a case study. Ecol Model
220:1697–1718
127. Leon LF, Lam D, Schertzer W, Swayne D (2005) Lake and climate models linkage: a 3D
hydrodynamic contribution. Adv Geosci 4:57–62
128. Robson B, Hamilton D (2004) Three-dimenstional modelling of a microcystis bloom event in
the Swan River estuary, Western Australia. Ecol Model 174:203–222
129. Edinger JE (2001) Waterbody hydrodynamic and water quality modeling, workbook and CD-
ROM. ASCE Press, Reston
130. West D, Stefan HG (1998) Simulation of lake water quality using a one-dimensional model
with watershed input. Model description and application to lake riley and lake elmo, Project
Report No.430. St. Anthony Falls Laboratory, University of Minnesota, Minnesota
131. West D, Stefan HG (1999) Modeling of watershed input and potential climate change effects
on water quality in agricultural reservoirs in the little Washita River watershed, Project
Report No.437. St. Anthony Falls Laboratory, University of Minnesota, Minnesota
132. West D, Stefan HG (2000) Simulation of water quality and primary productivity control
strategies for lake McCarrons, Project Report No.426. St. Anthony Falls Laboratory, Univer-
sity of Minnesota, Minnesota
133. Fang X, Stefan HG (1996) Development and validation of the water quality model
MINLAKE96 with winter data. Anthony Falls Laboratory, University of Minnesota,
Minneapolis
134. Fang X, Alam SR, Jacobson P, Pereira D, Stefan HG (2010) Simulations of water quality in
Cisco Lakes in Minnesota. Anthony Falls laboratory, University of Minnesota, Minneapolis
135. Edinger JE, Duttweiler DW, Geyer JC (1968) The response of water temperatures to
meteorological conditions. Water Resour Res 4:1137–1143
136. Ford DE, Stefan HG (1980) Thermal prediction using internal energy model. J Hydraul Div
ASCE 106(HY1):39–55
137. Harleman DRF (1982) Hydrothermal analysis of lake and reservoir. J Hydraul Div ASCE
108(HY3):39–49
138. Hondzo M, Stefan HG (1993) Lake water temperature simulation model. J Hydraul Eng
119:1251–1273
139. Hondzo M, Stefan HG (1992) Water temperature characteristics of lakes subjected to climate
change. St Anthony Falls Hydraulic Laboratory, University of Minnesota, Minneapolis
140. Fang X, Stefan HG (2009) Simulations of climate effects on water temperature, dissolved
oxygen, and ice and snow covers in lakes of the contiguous United States under past and
future climate scenarios. Limnol Oceangr 54:2359–2370
141. Maxwell EL (1998) METSTAT – the solar radiation model used in the production of the
national solar radiation database (NSRDB). Solar Energy 62:263–279
142. Fang X, Stefan HG (1994) Temperature and dissolved oxygen simulations for a lake with ice
cover. Anthony Falls Hydraulic Laboratory, University of Minnesota, Minneapolis
143. Fang X, Stefan HG (1996) Dynamics of heat exchange between sediment and water in a lake.
Water Resour Res 32:1719–1727
144. Gu R, Stefan HG (1990) Year-round temperature simulation of cold climate lakes. Cold Reg
Sci Technol 18:147–160
145. Ashton GD (1986) River and lake ice engineering. Water Resources, Highlands Ranch
146. Fang X, Ellis CR, Stefan HG (1996) Simulation and observation of ice formation (freeze-
over) in a lake. Cold Reg Sci Technol 24:129–145
147. Stefan HG, Fang X (1994) Dissolved oxygen model for regional lake analysis. Ecol Model
71:37–68
148. Mathias JA, Barica J (1980) Factors controlling oxygen depletion in ice-covered lakes. Can
J Fish Aquat Sci 37:185–194
149. Dake JMK, Harleman DRF (1969) Thermal stratification in lakes: analytical and laboratory
studies. Water Resour Res 5:484–496
150. Carslaw HS, Jaeger JC (1959) Conduction of heat in solids. Oxford University Press, New
York
151. Bannister TT (1974) Prediction equations in terms of chlorophyll concentration, quantum
yield, and upper limit of production. Limnol Oceangr 19:1–12
152. Megard RO, Combs WS Jr, Smith PD, Knoll AS (1979) Attenuation of light and daily integral
rates of photosynthesis attained by planktonic algae. Limnol Oceanogr 24:1038–1050
153. Bolsenga SJ (1977) Preliminary observations of the daily variation of Ice Albedo. J Glaciol
18:517–521
154. Lock GSH (1990) The growth and decay of ice. Cambridge University Press, Cambridge
155. Wake A, Rumer RR Jr (1979) Modeling ice regime of Lake Erie. J Hydraul Div 105:827–844
156. Scott JT (1964) A comparison of the heat balance of lakes in winter. Department of
Meteorology, University of Wisconsin, Madison
157. Pivovarov AA (1972) Thermal Condition in Freezing Lakes and Rivers. Wiley, New York
158. Greene GM (1981) Simulation of ice-cover growth and decay in one dimension on the upper
St. Lawrence River. NoAA Technical Memorandum Erl GLERL-36
159. Carlson RE (1977) A trophic state index for lakes. Limnol Oceanogr 22:361–369
160. Brown LC, Barnwell TO (1987) The enhanced stream water quality models QUAL2E and
QUAL2W-UNCAS: documentation and user manual. U.S. Environmental Protection
Agency, Athens
161. Zison SW, Mills WB, Diemer D, Chen CW (1978) Rates, constants and kinetic formulations in
surface water quality modeling. Tetra Tech Incorporation, USEPA, ORD, Athens, Georgia
162. Fang X, Stefan HG (1998) Temperature variability in the lake sediments. Water Resour Res
34:717–729
163. Marshall CT, Peters RH (1989) General patterns in the seasonal development of chlorophyll
a for temperate lakes. Limnol Oceanogr 34:856–867
164. Todd DK (1980) Groundwater Hydrology, 2nd edn. John Wiley, New York, NY
165. Magnuson JJ, Webster KE, Assel RA et al (1997) Potential effects of climate changes on
aquatic systems: Laurentian great lakes and precambrian shield region. Hydrol Process
11:825–871
166. Stefan HG, Fang X, Eaton JG (2001) Simulated potential thermal/DO fish habitat changes in
lakes in response to projected climate warming. Trans Am Fish Soc 130:459–477
167. Fang X, Stefan HG, Eaton JG, McCormick JH, Alam SR (2004) Simulation of thermal/
dissolved oxygen habitat for fishes in lakes under different climate scenarios: Part 1.
Cool-water fish in the contiguous US. Ecol Model 172:13–37
Cold-water fish in the contiguous US. Ecol Model 172:39–54
Warm-water fish in the contiguous US. Ecol Model 172:55–68
170. Eaton JG, McCormick JH, Goodno BE et al (1995) A field information based system for
estimating fish temperature requirements. Fisheries 20:10–18
171. Schindler DW, Bayley SE, Parker BR et al (1996) The effects of climate warming on the
properties of boreal lakes and streams at the experimental lakes area, Northwestern Ontario.
Limnol Oceanogr 41:1004–1017
172. Stefan HG, Hondzo M, Fang X, Eaton JG, McCormick JH (1996) Simulated long-term
temperature and dissolved oxygen characteristics of lakes in the north-central United States
and associated fish habitat limits. Limnol Oceanogr 41:1124–1135
173. Stefan HG, Fang X (1995) A methodology to estimate year-round effects of climate change
on water temperature, ice and dissolved oxygen characteristics of temperate zone lakes with
application to Minnesota. Anthony Falls Laboratory, University of Minnesota, Minneapolis,
MN, St, 55414
174. Stefan HG, Hondzo M, Eaton JG, McCormick JH (1995) Predicted effects of global climate
change on fishes in Minnesota lakes. Fish Aquat Sci 121:57–72 (Special Publication)
175. Stefan HG, Hondzo M, Eaton JG, McCormick JH (1995) Validation of a fish habitat model
for lakes. Ecol Model 82:211–224
176. Fang X, Stefan HG, Alam SR (1999) Simulation and validation of fish thermal DO habitat in
north-central US lakes under different climate scenarios. Ecol Model 118:167–191
177. Jacobson PC, Stefan HG, Pereira DL (2010) Coldwater fish oxythermal habitat in Minnesota
lakes: influence of total phosphorus, July air temperature, and relative depth. Can J Fish
Aquat Sci 67:2003–2013
178. Evans D (2007) Effects of hypoxia on scope-for-activity and power capacity of lake trout
(Salvelinus namaycush). Can J Fish Aquat Sci 64:345–361
179. US EPA (1976) Quality criteria for water. United States Environmental Protection Agency
(US EPA), Washington, DC
180. Chapman G (1986) Ambient aquatic life criteria for dissolved oxygen. US Environmental
Protection Agency, Washington, DC
181. Barica J, Mathias JA (1979) Oxygen depletion and winterkill risk in small Prairie lakes under
extended ice cover. J Fish Res Board Can 36:980–986
182. Fang X, Stefan HG (2000) Projected climate change effects on winterkill in shallow lakes in
the northern U.S. Environ Manag 25:291–304
183. Fang X, Alam SR, Jiang LP et al (2010) Simulations of Cisco Fish Habitat in Minnesota
Lakes under Future Climate Scenarios. Anthony Falls laboratory, University of Minnesota,
Minneapolis, St
184. Boer GJ, McFarlane NA, Lazare M (1992) Greenhouse gas-induced climate change
simulated with the CCC second-generation general circulation model. J Clim 5:1045–1077
185. McFarlane G, Boer J, Blanchet JP, Lazare M (1992) The Canadian climate centre
second-generation general circulation model and its equilibrium climate. J Clim 5:1013–1044
186. Kim S-J, Flato GM, Boer GJ, McFarlane NA (2002) A coupled climate model simulation of
the Last Glacial maximum, Part 1: transient multi-decadal response. Clim Dynam
19:515–537
187. Kim S-J, Flato GM, Boer GJ (2003) A coupled climate model simulation of the Last Glacial
maximum, Part 2: approach to equilibrium. Clim Dynam 20:635–661
188. Hasumi H, Emori S (2004) K-1 coupled model (MIROC) description. University of Tokyo,
Center for Climate System Research
Books and Reviews
Eaton JG, McCormick JH, Stefan HG, Hondzo M (1995) Extreme-value analysis of a fish
temperature-field database. Ecol Eng 4(4):289–305
Ellis CR, Stefan HG (1989) Oxygen demand in ice covered lakes as it pertains to winter aeration.
J Am Water Resour Assoc 25(6):1169–1176
Ellis CR, Stefan HG (1991) Field testing of an ice-preserving winter lake aeration system. J Am
Water Resour Assoc 27(6):903–914
Ellis CR, Stefan HG, Gu R (1991) Water temperature dynamics and heat-transfer beneath the ice
cover of a lake. Limnol Oceanogr 36(2):324–335
Fang X, Stefan HG (1995) Interaction between oxygen transfer mechanisms in lake models.
J Environ Eng 121(6):447–454
Fang X, Stefan HG (1996) Long-term lake water temperature and ice cover simulations/
measurements. Cold Reg Sci Technol 24(3):289–304
Fang X, Stefan HG (1997) Simulated climate change effects on dissolved oxygen characteristics in
ice-covered lakes. Ecol Model 103(2–3):209–229
Fang X, Stefan HG (1998) Potential climate warming effects on ice covers of small lakes in the
contiguous U.S. Cold Reg Sci Technol 27(2):119–140
Fang X, Stefan HG (1999) Projections of climate change effects on water temperature
characteristics of small lakes in the contiguous US. Clim Change 42(2):377–412
Gao S, Stefan HG (1999) Multiple linear regression for lake ice and lake temperature
characteristics. J Cold Reg Eng 13(2):59–77
Gao S, Stefan HG (2004) Potential climate change effects on ice covers of five freshwater lakes.
J Hydrol Eng 9(3):226–234
Gu R, Stefan HG (1993) Validation of cold climate lake temperature simulation. Cold Reg Sci
Technol 22(1):99–104
Gu R, Luck FN, Stefan HG (1996) Water quality stratification in shallow wastewater stabilization
ponds. J Am Water Resour Assoc 32(4):831–844
Henneman HE, Stefan HG (1998) Snow and ice albedo measured with two types of pyranometers.
J Am Water Resour Assoc 34(6):1487–1494
Henneman HE, Stefan HG (1999) Albedo models for snow and ice on a freshwater lake. Cold Reg
Sci Technol 29(1):31–48
Herb WR, Stefan HG (2004) Temperature stratification and mixing dynamics in a shallow lake
with submersed macrophytes. Lake Reserv Manag 20(4):296–308
Herb WR, Stefan HG (2005) Dynamics of vertical mixing in a shallow lake with submersed
macrophytes. Water Resour Res 41(2):W02023, doi:10.1029/2003wr002613
Herb WR, Stefan HG (2006) Seasonal growth of submersed macrophytes in lakes: the effects of
biomass density and light competition. Ecol Model 193(3–4):560–574
Hondzo M, Stefan H (1991) Three case studies of lake temperature and stratification response to
warmer climate. Water Resour Res 27(8):1837–1846
Hondzo M, Stefan HG (1993) Regional water temperature characteristics of lakes subjected to
climate change. Clim Change 24:187–211
Hondzo M, Stefan HG (1996) Long-term lake water quality predictors. Water Res
30(12):2835–2852
Hondzo M, Stefan HG (1996) Dependence of water quality and fish habitat on lake morphometry
and meteorology. J Water Resour Plan Manag 122(5):364–373
Hondzo M, Ellis CR, Stefan HG (1991) Vertical diffusion in small stratified lake: data and error
analysis. J Hydraul Eng 117(10):1352–1369
Horsch GM, Stefan HG (1988) Convective circulation in littoral water due to surface cooling.
Limnol Oceanogr 33(5):1068–1083
Johnson SL, Stefan HG (2006) Indicators of climate warming in Minnesota: Lake ICE covers and
snowmelt runoff. Clim Change 75(4):421–453. doi:10.1007/s10584-006-0356-0
Manous JJD, Stefan HG (2003) Projected sulfate redistribution as impacted by lake level stabili-
zation scenarios: devils lake, North Dakota. J Water Resour Plan Manag 129(5):399–408
Rasmussen AH, Hondzo M, Stefan HG (1995) A test of several evaporation equations for water
temperature simulations in lakes. J Am Water Resour Assoc 31(6):1023–1028
Stefan HG, Fang X (1993) Model simulations of dissolved oxygen characteristics of Minnesota
lakes: past and future. Environ Manag 18(1):73–92
Stefan HG, Fang X (1997) Simulated climate change effects on ice and snow covers on lakes in
a temperate region. Cold Reg Sci Technol 25(2):137–152
Stefan HG, Fang X, Hondzo M (1998) Simulated climate changes effects on year-round water
temperatures in temperate zone lakes. Clim Change 40:547–576
Stefan HG, Fang X, David W, Eaton JG, McCormick JH (1995) Simulation of dissolved oxygen
profiles in a transparent, dimictic lake. Limnol Oceanogr 40(1):105–118
Stefan HG, Hondzo M, Fang X (1993) Lake water quality modeling for projected future climate
scenarios. J Environ Qual 22(3):417–431
Stefan HG, Horsch GM, Barko JW (1989) A model for the estimation of convective exchange in
the littoral region of a shallow lake during cooling. Hydrobiologia 174(3):225–234
Stefan H, Cardoni J, Schiebe F, Cooper C (1983) Model of light penetration in a turbid lake. Water
Resour Res 19(1):109–120. doi:10.1029/WR019i001p00109
Stefanovic DL, Stefan HG (2002) Two-dimensional temperature and dissolved oxygen dynamics
in the littoral region of an ice-covered lake. Cold Reg Sci Technol 34(3):159–178
Williams G, Layman KL, Stefan HG (2004) Dependence of lake ice covers on climatic, geo-
graphic and bathymetric variables. Cold Reg Sci Technol 40(3):145–164
Williams SG, Stefan HG (2006) Modeling of lake ice characteristics in North America using
climate, geography, and lake bathymetry. J Cold Reg Eng 20(4):140–167
Chapter 12
Oceanic Fate and Transport of Chemicals
Robert P. Mason
Glossary
Anthropogenic Produced by or derived from human-related activities.

Biogeochemical cycling The overall transport of chemicals through the ocean
waters as modified by chemical, physical, and
biological processes.
Chemical inputs The external sources of chemicals to the ocean from the
atmosphere or from deep ocean environments or from
rivers and other terrestrial sources.
Chemical sinks All elements and compounds can be removed from the
ocean by various processes. The relative ratio of their
input to their removal provides an indication of how
they are distributed in the ocean, and whether human
activity has increased their ocean concentration.
Major ions Those chemicals present in the ocean at high
and nutrients concentrations and the major nutrients (nitrogen, phos-
phorous, and silica).
Metalloid An element in the periodic table that acts both as a metal
and a nonmetal, depending on the chemical
environment.
Micronutrients Those chemicals present in the ocean at low
and trace elements concentrations but which still have an important impact
of ocean biological productivity, either because they are
R.P. Mason (*)
Department of Marine Sciences & Chemistry, University of Connecticut,
Groton, CT 06340, USA
288 R.P. Mason
essential nutrients (e.g., iron and zinc) or toxic

compounds (e.g., lead and mercury).
Organic chemicals In the context of this chapter these are chemicals
manufactured purposely (e.g., PCBs, pesticides) or
inadvertently (e.g., dioxins) by humans, and also
released to the environment by human activity (petro-
leum compounds).
Persistent organic Organic compounds that are stable in the environmental
pollutants (POPs) and not rapidly degraded and which are also
bioaccumulative through the food chain and potentially
toxic to organisms.
or their consumers, are nonessential elements or compounds, or are major components

with essentially uniform concentration across the ocean system. Given that there are
over 100 elements on Earth, and a finite but very large number of organic chemicals,
derived either from an original biochemical source or manufactured by humans, it
would be impossible to deal with all these compounds and elements in any detail.
Therefore, while this chapter contains a description of the cycling of compounds
through the marine environment and identifies the main pathways of movement and
the major sources and sinks for compounds, in both the solid and dissolved phases, the
focus of the discussion is a smaller list of/or their consumers, are nonessential elements
or compounds, or are major components with essentially uniform concentration across
the ocean system. Given that there are over 100 elements on Earth, and a finite but very
large number of organic chemicals, derived either from an original biochemical source
or manufactured by humans, it would be impossible to deal with all these compounds
and elements in any detail. Therefore, while this chapter contains a description of the
cycling of compounds through the marine environment and identifies the main
pathways of movement and the major sources and sinks for compounds, in both the
solid and dissolved phases, the focus of the discussion is a smaller list of compounds
whose inputs to the ocean and whose fate and transport have important consequences
for oceanic life and the environmental health of the ocean and the biosphere in general.
The four main groups of elements/compounds and their fate and transport, and how
these have changed in the recent and geological past, and the consequences of these
changes on the ocean are described.
The four main classes of compounds are: (1) nutrients, including both the major
nutrients (nitrogen (N), phosphorous (P), silica (Si)), and the minor nutrients, which
are constituents of important enzymes and biochemicals (e.g., iron (Fe), zinc (Zn)
and other transition metals); (2) toxic metals and metalloids, such as mercury (Hg),
lead (Pb), arsenic (As) and silver (Ag); (3) organic compounds, primarily those
12 Oceanic Fate and Transport of Chemicals 289
manufactured or extracted from the Earth by humans (petroleum-derived

compounds, pesticides and fungicides, pharmaceutical products and related
chemicals); and (4) radioactive compounds and other environmental pollutants
not covered in the other classes. While the addition of carbon dioxide (CO2),
released by fossil fuel burning and other human activities, to the ocean has and
will have a large and currently uncertain impact on the basic ocean chemistry
through its exacerbation of ocean acidification, and will impact, for example, the
ability of calcifying organisms to precipitate their external shells, this will not be
covered in this chapter.
Much of the discussion focuses on chemicals whose concentration in the ocean
has been exacerbated by human activities. It concentrates on their dominant sources
to the ocean, the factors that control their overall distribution and fate, with a brief
discussion of the mechanisms for their accumulation into the marine food chain.
Human health and environmental health impacts of chemicals assimilated into
marine biota will not be directly discussed but this is clearly the most important
concern of the increased ocean concentrations of many compounds. Historically, it
was thought that the “solution to pollution was dilution” and that the ocean was
a large and limitless reservoir that could not be severely impacted by the input of
human wastes and other chemicals into the environment. This notion was prevalent
in the early part of the twentieth Century, and before, but it has become increasing
clear that this idea is a fallacy and that human activity has markedly impacted the
ocean, as it has the other reservoirs of the biosphere.
Therefore, there has been a resultant change in attitude that has resulted in
numerous conventions and international activities to mitigate past anthropogenic
impacts and to ensure a sustainable future for marine resources. The debate
however continues about the degree to which the ocean can be used as a “waste
disposal site.” Some scientists and ecosystem managers focus on the “precautionary
principle” while many others think that the inputs can be effectively managed and
assessed through the quantification of loading and an understanding of the level of
the chemical in a particular environment that can be tolerated. These two
approaches require different levels of understanding and knowledge. Therefore,
there is the need to further examine and ascertain the degree to which the ocean can
buffer ongoing inputs of chemicals. However, many such debates lack an apprecia-
tion for the complexity of the ocean system and how the inputs of chemicals in one
location can impact the ocean more generally, and how chemicals are transported
from one location to the other, and how long they actually reside in the ocean water
column before being removed. This chapter attempts to provide the background and
insights necessary to make such informed decisions.
Introduction
All the elements are found in the ocean, but their concentrations differ by many
orders of magnitude in total concentration. The concentration of many elements and
compounds has not changed substantially for thousands of years while those
290 R.P. Mason
chemicals that have a strong anthropogenic component to their emissions to the

atmosphere and to coastal waters have changed substantially in the last 100 years.
The overall processes important in the cycling of chemicals in the ocean and across
the important interfaces between the ocean and the rest of the biosphere are
shown in Fig. 12.1. Exchange of chemicals occurs across the air–sea interface
due to deposition of gases and particulate, as both wet and dry deposition. For
gases, the direction of exchange – from the water to air or vice versa – depends on
the relative concentrations of the compounds in each phase, and the relative
solubility or volatility of the compound. Inputs of chemicals from the terrestrial
environment are also important via riverine discharge as well as subsurface
inputs from groundwater. Removal of chemicals from the surface ocean is due
to mixing of ocean currents and internal cycling as well as the transport of
chemicals through the sinking of particulate material to the deep ocean. For
most chemicals that reach the deep ocean, the ultimate sink is removal to the
sediment, and the relative rate of sedimentation to input from the surface ocean
determines the net difference in surface and deep ocean concentrations. This is
discussed further in the remainder of the chapter.
The major ions are sodium (Na) and chloride (Cl), both which are present at
around 0.5 M concentration, with the other major ions (Calcium (Ca), magnesium
(Mg), potassium (K), sulfate, and bicarbonate) being present at concentrations of
5–20 times less (Table 12.1) [77, 100]. In contrast, the major nutrients are present
around 10 6 M or lower (often more than a thousand times less), the major
transition metals are present at 10 8–10 10 M and some of the more rare metal
elements are present at <10 12 M (a trillion times less) (Table 12.1). Total organic
matter ranges from 10 3 to 10 5 M (on a carbon basis). Concentrations of many
trace species are highest in coastal waters compared to the deep ocean and for
elements and chemicals that interact and bind strongly with suspended particulate
matter in the water, their total concentration is determined to a large degree by the
suspended solids (TSS) levels, which vary from >100 mg/L for some rivers and
estuaries to <1 mg/L in open ocean waters [100].
In terms of organic compounds, the focus of the discussion is contaminant and
toxic compounds rather than the large suite of biochemicals and their degradation
products found in seawater. The most important from an environmental health point
of view are the so-called “persistent organic pollutants” (POPs), as defined by the
Stockholm Convention. The POP designation includes a series of pesticides and
other compounds initially termed the “dirty dozen” in earlier classifications of toxic
and bioaccumulative organic compounds (aldrin, chlordane, dieldrin, endrin,
heptachlor, DDT, and its breakdown products (designated here as ∑DDT), hexachlor-
obenzene, mirex, polychlorinated biphenyls (PCBs), polychlorinated dibenzo-p-
dioxins (referred to as dioxins in this chapter), polychlorinated dibenzofurans, and
toxaphene). Since then, polycyclic aromatic hydrocarbons (PAHs) and certain
brominated flame-retardants (e.g., polychlorinated diphenyl ethers (PBDEs)), as well
as some organometallic compounds such as tributyltin (TBT) and related products
(referred to here as alkylated tin compounds) have been added [57, 64].
Dry gaseous and

Wet Deposition
Particle Deposition
Water-Air
Exchange
NOM Complexes
Inorganic Species &
& Interactions
Neutral Compounds
Photochemical
Dissolved Colloids
Degradation
ions
Transformations
Redox Chem
Microbial
Alkylation
Degradation
Bioaccumulation
& Trophic Transfer
Solute Resuspension Settling

Exchange Bioturbation
Diagenesis Bioturbation
Net Burial Diffusion Partitioning
Transformations
Fig. 12.1 Diagram showing the major sources and sinks for chemicals in the ocean and the major
processes that are involved in determining their horizontal and vertical distributions and
concentration
While not specifically on the POP list, other organic chemicals that are of
concern will be mentioned throughout the chapter. Concentrations of the classes
of organic compounds are given in Table 12.1 and are typically low (pg/L range)
[58, 64]. Given the range in compounds in each class of organic chemicals in the
Table, and discussed in the chapter, concentrations are given on a mass basis.
As noted above, this chapter cannot cover all aspects of chemicals in the ocean.
Important topics of marine environmental pollution that will not be dealt with in
detail include the transport of sediments themselves to coastal waters due to
enhanced erosion and human activities within the watershed, such as deforestation,
agriculture, or due to activities in coastal waters, such as dredging and bottom
trawling, and the impact of the resultant excessive sediment on the ecosystem
health of coastal waters [12, 45]. Excessive turbidity and sedimentation in such
waters will impact the growth and survival of bottom-dwelling (benthic) plants (sea
grasses and other plants) and animals (e.g., bivalves), either as a result of light
limitation for plants, or due to smothering of benthic organisms and habitat degra-
dation. Excessive sediment input is an important environmental concern. Addition-
ally, sediment transport is an important vehicle for contaminants and other
chemicals [100], and this aspect will be considered in detail.
292 R.P. Mason
Table 12.1 Estimated range and average concentrations for open ocean waters for major ions,
trace elements and organic chemicals, and for the elements, the estimated residence time of each
metal in the global ocean, and the major dissolved inorganic species in solution. The data for the
organic chemicals are for the early 1990s. Taken from the literature ([17, 18, 21, 64, 111])
Average Conc. (nM, Residence Major dissolved
Ocean Conc. Range except where time inorganic species
Element (nM) indicated) (103 year) for elements
Major ions Na+, Cl 0.49, 0,55 M >105 Free ions
K+, Ca2+, Mg2+ 9.7, 8.3, 4.4 mM
SO42 , HCO3 13, 1.8 mM
Nitrate <100–100,000 – – Free ions, acid–base
Phosphate <100–5,000 – pairs
Silicate 1,000–150,000 –
Al 0.3–40 20 0.6 Al(OH)x3 x, x = 3,
4
Cr 3–5 4 8.2 CrO42
Mn 0.08–5 0.3 1.3 Mn2+
Fe 0.01–2 0.5 0.05 Fe(OH)x3 x, x = 2,
3
Co <0.01–0.3 0.02 0.34 Co2+
Ni 2–12 8 8.2 Ni2+
Cu 0.5–4.5 4 0.97 CuCO30
Zn 0.05–9 5 0.51 Zn2+
As 20–25 23 39 HAsO42
Mo – 105 820 MoO42
Ag <0.01–0.04 0.02 0.35 AgClx1 x, x = 1–4
Cd <0.01–1 0.5 – CdClx2 x, x = 1–4
Hg <0.01 0.002 0.56 HgClx2 x, x = 1–4
Pb <0.01–0.15 0.1 0.81 PbCO30
Se 0.5–2.3 1.7 26 SeO42
Total PCBs 1–50 pg/L 15 pg/L NA –
∑DDT 0.2–10 pg/L 2 pg/L NA –
∑Chlordanes 2–20 pg/L 5 pg/L NA –
∑HCH 0.05–2.1 ng/L 0.2 ng/L NA –
The chapter will also not directly consider the complex mixture termed “natural”
organic matter (OM), either particulate or dissolved (POM or DOM) that is
primarily derived from natural sources and biochemical processes, either those
occurring in situ in the ocean or being derived from terrestrial inputs. Some of
the OM comes from human activities such as sewage releases and dumping, and
enhanced inputs to coastal waters have helped to drive water column oxygen
depletion, although the major culprits are excessive amounts of nutrients. So, for
this reason, OM will not be discussed directly although it has an impact on the
cycling of the other chemicals to be discussed, and this will be included in the
discussion.
As the chapter is focused on chemicals, there will be no discussion of the fate
and transport of bacteria and related organisms, and their potential impact on
human health [12, 45]. The most common and well-known contamination of this
kind is from Escherichia coli bacteria, a mammalian intestinal bacteria, and the
number of such bacteria in a water sample is often used as a pollution indicator as
they do not survive for long periods in seawater. Shellfish can also accumulate these
and other toxic bacteria and their chemical products, and therefore cause toxicity
onto their consumers. Additionally, there is the potential for the harmful algal
blooms (HAB) in marine waters as a result of the rapid growth of dinoflagellates
and other organisms, which again can result in human health impacts and disease.
Overall, infections as a result of contaminated seafood can be viral, parasitic, or
bacterial.
Finally, marine litter and debris is an ever-increasing ocean problem, especially
in terms of small plastic materials and other smaller products that are more easily
ingested by marine organisms. Many ships dump garbage and wastes while at sea,
which is not biodegradable and accumulates on the ocean surface. Additionally,
gear is lost from ships while at sea, and beach litter and other material are washed to
sea during storms. Sewage inputs and storm-water overflow into coastal waters can
be a source of household, medical, and other wastes. Plastic particles can be large or
small (<1 mm in size) [110]. Given ocean circulation patterns, there are now regions
of the ocean where there is an ever-growing accumulation of marine debris that is
extensive in size, with densities of >10,000 particles per square kilometer. The main
impact of large plastics is on marine animals, such as turtles, and birds. Recent action
has lead to regulations banning dumping plastics and other non-degradable products
at sea. Additionally, it is now being recognized that the smaller plastic particles can
accumulate hydrophobic organic compounds [44] and therefore could have a role in
the bioaccumulation of these compounds through their ingestion by benthic
organisms and filter feeders. Further discussion on this topic can be found in
a number of publications ([110]; etc.) and the important problem of ocean debris
will not be further considered in this chapter.
The Transport of Chemicals to the Open Ocean
The three important sources of chemicals to the ocean are: (1) inputs from the
atmosphere of chemicals in the gas phase and attached to particles (termed dry
deposition) and inputs of chemicals with precipitation (wet deposition); (2) inputs
from the terrestrial environment (watershed inputs, surface runoff, and point and
nonpoint discharges and groundwater inputs); and (3) inputs from the deep ocean
via release from sediments and/or from hydrothermal vents and other natural point
source inputs (Fig. 12.1). Additionally, some chemicals present in seawater as
a dissolved gas can build up in ocean surface waters to supersaturated
concentrations and thereafter can be lost to the atmosphere via gas exchange
processes; this gas exchange can be thought of as negative dry deposition. The
relative importance of each source for the elements and organic chemicals is
collated in Table 12.2.
294 R.P. Mason
Table 12.2 Estimated inputs of some elements and chemicals to the ocean from the atmosphere,
coastal/riverine inputs and from hydrothermal vents. The atmospheric inputs include both wet and
dry (gaseous and particulate) net deposition. The relative importance of each source is indicated
for each metal. Data taken from the literature. Fluxes are in Gmol year 1 for the elements and in
Mg year 1 for the organic compounds. Taken from the literature as noted in the text
Element Fluvial flux Pluvial flux Hydro-therm. %Pluviala
N 1.5–5 3.5–10 – 70
Al 5,445 2,156 0.4 28
Mn 35 11 25 16
Fe 1345 469 75 25
Co 0.65 0.23 0.070 24
Ni 3.0 1.5 – 34
Cu 3.4 0.65 0.75 14
Zn 6.7 1.3 2.0 13
As 5.0 0.08 – 2
Ag 0.60 0.050 0.001 8
Cd 0.042 0.04 0.007 45
Hg 0.003 0.017 0.001 81
Pb 0.78 1.3 0.001 62
∑PCB 60 230 NA 80
∑DDT 4 170 NA 98
∑HCH 60 4700 NA 99
HCB 4 80 NA 95
Dieldrin 4 40 NA 91
Chlordane 4 22 NA 85
NA not applicable
a
Estimates of percentage are the fraction of the total flux
Atmospheric Inputs
The information collected on the input of elements and chemicals to the ocean
surface from the atmosphere (Table 12.2) is difficult to summarize for two reasons.
One, sampling has occurred over a span of nearly 30 years and during these times
there have been large changes in their deposition rates due to changes in anthropo-
genic inputs and loadings, and changes in human activity globally. For example,
data from Bermuda suggest that the concentration of Pb in deposition has decreased
by about 90% between 1982/93 and 1996/97 (Fig. 12.2c) [15, 106] and a similar
decrease in ocean surface water concentrations has been found over the same period
(Fig. 12.2a, b; 60% decrease). Similarly, cadmium (Cd) has decreased by 80%,
Zn by about 55%, copper (Cu) and nickel (Ni) by about 60%, and the decreases for
the coastal east coast USA is of the same order [60].
This is also true for the POPs. The data for organic chemicals given below are
from papers published in the early 1990s [33, 47, 111] and as many of these
chemicals are now banned, these values have changed substantially as well. For
example, emissions of a-HCH (lindane) has decreased by more than 90% since the
early 1980s; emissions of PCBs more than 90% since their peak emissions in the
Pb (pmol kg−1)
0 50 100 150 200
0
1997 1993
100
150
200
1989
300
1984
Pb (pmol kg−1)
400
Depth (m)
100
500
600
1979
700 50
800
900
a 1000 b 0
1980 1985 1990 1995 2000
Year (year marked at Jan. 1)
Pb concentration in the atmosphere, 10−10 g/m3
180
Pb concentration in seawater, pmol/kg and
Marine dissolved Pb
160
Atmospheric Pb
140
120
100
80
60
40
20
0
1970 1975 1980 1985 1990 1995 2000
c Year
Fig. 12.2 (a) Vertical concentration profiles for lead in the North Atlantic at a location close to
Bermuda over time; (b) A plot of the data from surface samples collections at the same location
over time; and (c) A compilation of data for both the surface ocean and the atmosphere (atmo-
spheric particle concentrations) from the same location. The first two figures are reprinted from
“Anthropogenic Trace metals in the Ocean,” a chapter by E.A. Boyle, pp. 273–280, with permis-
sion from Elsevier, and the third figure from “Atmospheric Input of Pollutants,” a chapter by R.A.
Duce, pp. 281–290, with permission from Elsevie, both chapters were published in Marine
Chemistry and Geochemistry, Steele, J.H., Thorpe, S.A. and Turekian, K.K. (Eds), a derivative
of the Encyclopedia of Ocean Sciences, 2nd Ed., Elsevier, Amsterdam, copyright (2010). Data are
taken from various publications as indicated in the original figures
1970s; and DDT emissions have decreased by 50% in the last decade [57]. Overall,
the data also suggest that the rate and timing of the changes are not all similar,
reflecting the different anthropogenic sources to the atmosphere, changes in usage
at different times and the differences in locations of emissions over time (e.g., for
many chemicals, currently decreasing for North America and Europe; increasing
for Asia, or in Africa).
296 R.P. Mason
Table 12.3 Range in the flux of metals from the atmosphere to the ocean surface. All values are
given in mmol m 2 year 1. Data taken from the literature and converted as necessary ([3–6, 21, 34,
60 74, 87, 95, 121, 123])
Open ocean Coastal/seas
Element Wet Dry Total Totala
Al 40–555 7–110 100–2,000 590–104
Cr 1.2 1.7–3.3 1.7–40
Mn 0.55–1.6 0.11 0.73–10.4 10–70
Fe 5.0–100 3.0 132–600 220–4570
Co 0.003–0.017 0.001 0.042–0.46 0.60–6.6
Ni 1.4–5.1 6.3–66
Cu 0.32–1.1 0.55 0.63–16 5.0–205
Zn 0.31–10 0.30–0.15 0.37–21 50–1,370
Cd 0.036 0.21–0.80 0.20–38
Pb 0.97–4.1 0.005 0.068–7.0 1.0–130
Hg 0.001–0.048 <0.001 0.001–0.048 0.05–0.15
Ag 0.28 0.009 0.18–0.32 0.28–0.83
As 0.13–0.4 0.09 0.36–0.4 1.3–38
Se 0.051–0.51 0.13 0.13–1.8 2.8–6.1
Secondly, compiling data is difficult as it is often reported in different units and

not all the same metals have been the focus of the various investigations. Thus, for
some metals, there is a relatively large number of datasets to work from, while for
other metals, data limitation is a problem [33, 34]. Also, for the crustal elements, the
concentrations vary depending on the relative importance of dust input and there-
fore there are large ranges in the reported concentrations. For example, studies on
Bermuda have shown that there are large seasonal differences in the concentration
of metals in rain and in aerosols [60] and this reflects the dominance of inputs from
the east, with a relatively high dust component in summer, and inputs from the west,
off the North American continent, being more dominant in fall-winter. Addition-
ally, seasonal differences in rainfall amounts are obviously also important.
In addition to the total inputs, differences in the extent of input are also present
given the dominance of anthropogenic sources near cities, and the locations of
many large urban environments in coastal locations. This is illustrated by the
compilation of data for the metals and the metalloids (Table 12.3) [21, 34]. An
examination of the data indicates firstly that for the open ocean, wet deposition is
the more important source, especially for the “anthropogenic metals” (Pb, Zn, As,
Hg) and this reflects the typically low aerosol concentrations found over the open
ocean. Thus, the dry deposition component is relatively small, and scavenging of
the particles from the atmosphere by precipitation is the main mechanism of their
removal. For the crustal metals (Al, Fe, Mn), the wet deposition is of somewhat
lesser importance but still the dominant atmospheric flux, while for the other
metals, the wet is typically greater than 70% of the total deposition. Total deposi-
tion fluxes vary most for the main crustal elements, while the variability appears to
be lower for metals such as Cr, V, Ni, and Co, which have both a crustal source and
an anthropogenic source. Metals such as Pb, Cu, and Cd show a large range in their
values and this reflects the changes in their inputs over time, as noted above. For the
crustal metals, changes on a decadal basis have been relatively small in the last
30 years.
There is a strong contrast between the fluxes for the open ocean and those for the
coastal areas, and marginal seas (Table 12.3) (see references in Table legend). This
is true both for the crustal metals and for the anthropogenically derived metals, and
is likely also true for the organic chemicals. The increased flux for the crustal metals
is primarily due to the higher aerosol concentrations in the coastal environments,
while the higher fluxes for the other metals are mostly a consequence of their
proximity to anthropogenic sources. Dust fluxes decrease exponentially offshore
and most of the open ocean has inputs that are two orders of magnitude lower that
the surrounding coastal environments, except for regions of high dust input (to the
North Pacific from Asia; to the tropical Atlantic from Saharan Africa). Addition-
ally, for those chemicals with a dominant anthropogenic signal (Pb, Hg, organic
chemicals), inputs are highest in the Northern Hemisphere and have historically
been higher into the North Atlantic, although this is changing depending on the
specific chemical or element. Overall, the remote Atlantic, being a smaller ocean,
has higher inputs, and there is clearly a strong gradient between the Northern and
Southern Hemispheres.
Additionally, many cities are located in coastal areas or close to the coast, and
therefore contribute substantially to the coastal fluxes, which are highest within
a local radius of 50–100 km. For the anthropogenically dominated elements and
organic chemicals, it has been shown that the fluxes measured in the urban
environment are substantially greater than those representing the regional average
value. In addition, continents with high urbanization and which typically have
offshore air mass flow for all or part of the year (Asia and North America in the
Northern Hemisphere, in particular) result in a substantial input of contaminants to
the coastal waters. The much higher fluxes for the coastal regions of Asia is a result
of the current rapid urbanization in the region, as well as the substantial input of
dust, particularly during the spring “dust period” (March–May) when a large
fraction of the annual dust input occurs. In the case of Europe, it has been shown
that due to meteorological conditions, there are periods where there is substantial
transport from the continent to the Arctic region, and also clear evidence of
contamination of the North Atlantic coastal waters and the Mediterranean, due to
anthropogenic inputs [1, 33, 34, 48, 90]. Overall, there is much uncertainty in the
flux estimates as a result of the sporadic nature of the measurements and the lack of
consistency in the metals that are quantified in specific studies, and due to the
relatively short duration of many studies.
For organic chemicals, their interaction between the atmosphere and the surface
ocean is complex and this is particularly so for the chemicals where the fluxes are
dominated by gas exchange as these chemicals are volatile enough to be reemitted
to the atmosphere after deposition. Scientists therefore discuss the environmental
fate of organic chemicals in terms of their overall mode of long-range transport.
Based on their chemical properties, organic chemicals are classified as “fliers,”
298 R.P. Mason
“multi- hoppers,” “single hoppers,” and “swimmers” ([57, 64, 105]) and this
designation was developed mostly to distinguish the ability of a chemical to be
transported and deposited in the Arctic region. Most lower molecular weight
organics that are volatile are “multi-hoppers” as after they are deposited to the
ocean and land surfaces they can be efficiently reemitted on short timescales and
this enhances their overall lifetime in the biosphere and the extent of their long-
range transport. Such multiple exchanges are mostly driven by seasonal and spatial
temperature differences. Less volatile chemicals are “single hoppers” as they tend
to associate with particles and are not reemitted for the most part after deposition,
except if the particles to which they are attached are resuspended. “Swimmers” are
those chemicals that are relatively soluble and therefore are transported spatially
mostly through movement of ocean currents while “fliers” are those compounds
that do not have substantial air-sea exchange and their transport and fate is mostly
within the atmosphere.
Deposition of N to the ocean is related to the distribution of its sources (urban
areas and agriculture) as the primary sources of N are from release of nitrogen
compounds (collectively termed NOx) from coal and other fuel combustion, and
from the use of N-containing fertilizers, as well as other sources related to human
activity [34, 48]. It is estimated that 30–50% of the N input to the ocean is from
anthropogenic sources [33]. Nitrogen fixation in the ocean provides an input that is
about 50% of the external sources. Currently, inputs of N are changing from being
dominated by inputs from North America and Europe to having a larger Asian
signal, as is found for many other compounds which have an anthropogenic
component. Given the importance of agriculture as a source, inputs from Africa
are also expected to rise in the future [33].
Inputs from the Terrestrial Environment
The export of material from rivers to the coastal zone is not evenly distributed
globally because of the dominance of export from large rivers such as the Amazon,
and the large Asian rivers, and also as a result of spatial differences in terrain [42,
75]. In addition, material export is relatively higher for rivers draining steep
topography compared to those rivers draining regions with a large coastal plain,
especially for chemicals associated with particulate material. The export of partic-
ulate material from rivers to the ocean has been estimated by various groups. The
three regions of high particulate discharge are: (1) the Amazon River region (14%
of the total input); (2) southern Asia (34%), and (3) eastern Asia (36%). These
regions all have large rivers and periods of high rainfall that drive the high
particulate discharge. Much of the particulate material is however removed in
estuaries and at the coastal boundaries due to the mixing of fresh and saline waters,
changes in water flow rate and coagulation processes [21, 75]. The difference in
ionic strength, pH, and the concentrations of dissolved solids, causes a general
coagulation and removal of organic material in the mixing zone. The continents
with the widest coastal plain therefore tend to be locations where particulate
removal in the estuary is enhanced. The estimates in Table 12.2 are for the net
transport of elements and chemicals from the coastal zone to the open ocean.
While it is most likely that many particle-reactive chemicals will be removed
from the water column during estuarine mixing, it is possible for them to pass
through without their concentration changing dramatically (so-called conservative
elements) or they may even be added to the water column as a result of estuarine
processing. If there is a linear change in concentration of a particular element or
chemical relative to salinity through the estuary, then it is assumed that the
concentration is the result of dilution of the riverine signal with seawater. In
estuaries, salinity changes are not spatially linear with most of the salinity change
occurring in a small mixing zone where the freshwater and saline waters meet. This
region is also a region of high turbidity and particle settling.
For some elements and chemicals, the estuarine mixing of freshwater and seawater
and the resultant chemical interactions that occur result in the elements being added or
removed from the water column by precipitation, adsorption to suspended matter and
resultant sinking to the sediment, or from release from particles with increasing
salinity. Many metals and hydrophobic organic chemicals are highly particle reactive
and are strongly retained in estuaries as most (95%) of the suspended matter
entering from rivers is not exported out of the coastal zone [21]. The role of other
factors, such as biological productivity in surface waters and microbial degradation at
depth also needs to be considered. While many of the redox processes that occur in
estuaries can occur abiotically, it is now known that many of these processes are
biotically mediated. Phytoplankton productivity plays an important role in estuarine
processing of elements and chemicals. Phytoplankton productivity is driven by the
supply of nutrients, which are typically in higher concentration in river waters than in
coastal waters, and thus the input of freshwater and the estuarine mixing regime
determines to a large degree the location where maximum productivity occurs, but it is
often below the region of maximum turbidity, where light levels are low. Many
elements and chemicals are actively accumulated by algae in this region, or bind
strongly to the plankton, detrital material, and other particles. All these processes lead
to the removal of chemicals in the high turbidity zone and in the maximum region of
primary production, therefore potentially decreasing the dissolved concentration in
surface waters. Their ultimate fate is sinking into the deeper waters and accumulation
at the sediment–water interface. Thus, there are multiple mechanisms for the removal
of dissolved constituents in the low salinity mixing regime of estuaries.
Chemical Inputs from Hydrothermal Vents, Ocean Seeps,

and Sediments
The discovery of the presence of hydrothermal vents on the deep ocean floor, near
the mid-ocean ridges and spreading centers in the late 1970s, and demonstration of
300 R.P. Mason
the input of high temperature fluids from these systems changed our understanding
of the sources and sinks of major and minor elements to the ocean [43, 104]. The
percolation of seawater through the ocean crust leads to its heating followed by the
addition or subtraction of chemicals due to reactions of the heated water with the
crustal material. Temperatures can be as high as 350–400 C. The water becomes less
dense and rises back to the surface where it is emitted and mixed with the surrounding
cold seawater. Such processes are very important in the overall oceanic chemistry and
fate of the major ions. Detailed chapters on hydrothermal vents and their importance to
ocean geochemistry have been recently published and these provide much more
information than can be detailed here [43, 54].
Magnesium and, to a lesser degree, Na and K, are removed during hydrothermal
activity, while Ca, Fe, Mn (both in reduced forms), and other metals are added to
the fluids. Sulfate is converted to sulfide in the low oxygen environment and
bicarbonate (HCO3 ) is converted into methane. The low pH (<4) leads to the
conversion of dissolved carbonate species (primarily HCO3 ) into CO2 [43, 100].
The enriched elements in hot hydrothermal fluids are subsequently precipitated as the
fluids cool or as these sulfidic, hot solutions are emitted into the oxygenated bottom
ocean waters. Sulfides of Fe and other metals are precipitated, forming precipitate
chimneys through which the fluids vent. Any remaining Fe and Mn will subsequently
be oxidized and precipitate as their (hydr)oxides. Given the low solubility of many
metal sulfides and/or their propensity to bind to and be scavenged by Fe-sulfide
minerals and metal oxides precipitates, it is likely that most trace metals in hydrother-
mal solutions are removed to the solid phase close to the emission source.
It has been difficult to accurately estimate the importance of these sources for
trace metals to ocean waters (Table 12.2). The situation is further complicated by the
fact that there is still limited measurement of the concentrations of trace metals in such
plumes, and the measurements that have been made have found a large variation in
concentration, reflecting the heterogeneous nature of this overall process. Average
values for the concentration of trace metals in hydrothermal fluids are given in
Table 12.2. Overall, it is apparent that the concentrations of some metals, such as
Fe, Mn, Al can be enriched by many orders of magnitude over their concentration in
seawater [43]. Most transition metals, and heavy metals such as Ag, Tl, Pb, and Hg,
and the metalloids (As, Se, and Sb), are also enriched in vent fluids. Overall, however,
it is concluded that few of the first row transitions metals and the heavy metals (Ag,
Cd, Sn) have a strong net hydrothermal source (Table 12.2).
Inputs to the deep ocean from the sediment through changes in chemical form for
elements due to diagenesis within the sediment are thought to be small compared to
the other sources discussed. As discussed above, the mobility of some elements is
a function of their oxidation state and many can be released into solution in low
oxygen environments. However, as discussed above for hydrothermal systems, these
elements are rapidly removed from solution to the particulate phase on encountering
oxygenated waters above the sediment and therefore, even if released from the
sediment, are not more than a local source. However, in contrast to metals and
most elements, there are important natural sources of organic petroleum chemicals
in some ocean regions, specifically natural petroleum hydrocarbon seeps. These are
found in the Gulf of Mexico and off the coast of California, for example. The amount
of oil that is released from these seeps (15% of total inputs; [46]) is difficult to
estimate accurately but is not trivial compared to other sources releasing petroleum
products into the environment.
Comparison of Sources
Generally, for the ocean, based on the estimates in Table 12.2, an overall estimation of
the importance of atmospheric versus riverine inputs can be made. For the
human-derived chemicals, hydrothermal inputs are not considered. In making these
estimates, different values in the literature needed to be reconciled (e.g., [7, 9, 10, 13,
20, 21, 33, 34, 36, 37, 41, 43, 55, 57, 58, 70, 74, 82, 86, 95, 100]). Additionally, in
making these calculations, it was assumed that 90% of the riverine suspended material
was trapped in estuaries and did not get transported to the open ocean, and it was
assumed that if there was no data available for hydrothermal inputs for a particular
element, it was not significant. Clearly, this assumption may not be valid. Also, there is
little information on the potential importance of groundwater inputs or the release of
metals from coastal sediments and their transport offshore, which has recently been
suggested to be an important source of open ocean Fe. Overall, the estimated values
and relative contributions give an idea of the important sources to the ocean and allow
a comparison of the relative sources and the biogeochemical reasons for these
differences.
It is evident from Table 12.2 that riverine inputs are the dominant source for
most trace metals to the ocean. Based on these values, Hg has the highest relative
atmospheric input from the atmosphere at 85% of the total flux. The input from
riverine sources is higher than the hydrothermal input, but this flux is not well
known. For Pb, the atmosphere is also the dominant source, with the estimate being
63% of the total (Table 12.2). Again, the global cycle of Pb has been substantially
perturbed by human emissions of Pb to the atmosphere and this is likely the reason
for the importance of the atmosphere as a source [36]. In addition, Pb is highly
particle reactive and is therefore removed to a high degree during transport through
the estuary. Cadmium is another metal with a relatively high atmospheric signal,
which is also a relatively volatile metal with important anthropogenic sources.
Hydrothermal sources are also relatively important for Cd. Additionally, hydrother-
mal sources are somewhat important sources for Co, Cu, and Zn, but they are the
dominant input for Mn, accounting for about 40% of the total inputs, almost as
much as is entering the ocean from riverine sources. The atmospheric source is not
important for Mn, or any of the other crustal metals, especially Al, Fe, and Co.
For the organic chemicals, the data in Table 12.2, even though they are from an
early compilation of data by Duce et al. [34], confirm that the atmosphere is the
dominant source for these chemicals to the ocean. Therefore, these compounds are
determined to a substantial degree by long-range transport from their source regions
302 R.P. Mason
to the locations of their deposition. The extent of their global versus regional
deposition is a function of their residence time in the atmosphere. For petroleum-
derived products, the inputs to the ocean come both from long-range transport
(atmospheric and terrestrial inputs) and from local and regional inputs that are
significant, and often highly variable [46]. Total inputs per year are in the region of
2–3 million tons. These include the inputs from marine accidents, discharges
associated with offshore drilling, marine storage facilities and operations, and
discharges from tankers and other ships.
For most metals and organic chemicals, removal to deep ocean sediments is the
major and dominant sink for any of the compound transported into the deep ocean
from the surface [21, 57, 64, 105]. However, this observation is only really true for
those chemicals that are present in the dissolved phase as ionic species, or those that
do not have a significant vapor pressure. This is true for most elements except those
that have compounds that are volatile (e.g., Hg, Se) and the organic compounds. For
Hg, which can exist in surface waters as a dissolved gas (elemental Hg), and many
organic compounds loss to the atmosphere via gas exchange is the dominant loss
term for the ocean [74]. For the majority of the organic chemicals, this is also true
and the surface ocean concentration of these chemicals is mostly driven by the
partitioning between the atmosphere and the surface waters. As discussed further
below, the ocean can therefore be a source or sink of these compounds via gas
exchange depending on the relative concentrations in each phase. As many of these
chemicals have changed in atmospheric concentration rapidly compared to the
mixing time of the upper ocean (years to decades, depending on location), many
oceans are now a source of organic chemicals to the atmosphere where previously
they may have been a net sink.
Ocean Cycling of Chemicals
The concentration and distribution of an element or chemical in the ocean is not

directly related to its crustal abundance or its concentration in the surface soils, or
its rate of release into the biosphere, but is rather determined by the complex
interaction between its solubility, and its volatility, its ability to bind to or be
incorporated into other phases, such particulate and dissolved organic matter
(POM & DOM) or TSS (POM is a fraction of the TSS), and its potential to be
present as colloidal material [17]. Scientists mostly measure the “dissolved” frac-
tion of a chemical in seawater by filtration through membranes with openings of
0.2–0.8 mm, and therefore the filtered fraction can contain some colloidal material.
Much of the DOM is likely in colloidal suspension and many insoluble metals such
as Fe are potentially present as colloidal material as well in ocean waters.
A chemical’s rate of biotic and abiotic transformation is also important and there
are many microbial processes that can degrade organic chemicals or transform them
into a different, and perhaps more toxic or mobile state. Microbial processes can
also transform the toxicity of metals (e.g., methylation of Hg into methylmercury

(CH3Hg) produces a more toxic and bioaccumulative compound while methylation
of As appears to reduce its toxicity). Intracellular transformations of organic
compounds by Cytochrome P450 and other cellular degradation pathways can
lead to more or less toxic products. The fate of some elements is also determined
by the system oxidation state as their solubility is a strong function of this (Fe is
more soluble when present as reduced FeII, chromium (Cr) is more soluble when
oxidized as hexavalent Cr).
Besides the impact of biogeochemistry, particles, and organisms on the chemical
distributions, there is also the impact of ocean circulation [17]. Ocean waters mix in
a complex manner and this has a large impact of the fate and transport of chemicals.
Therefore, a brief description of ocean circulation is required to understand the
ocean distribution of chemicals. Vertically, the ocean is divided into: (1) the mixed
layer (a few hundred meters deep), which is efficiently mixed by winds and
currents, and where most of ocean photosynthesis occurs, and where most
organisms reside; (2) the thermocline region, where there is a rapid change in
temperature from the warmer upper waters to the colder deeper waters, and
where there is strong density stratification and limited vertical mixing; and (3) the
deep ocean. Below the thermocline, around 1,000 m in many oceans, is a region of
oxygen depleted waters where much of the remineralization of organic particles
transported vertically from the surface ocean occurs. In the deep ocean, waters are
cold and relatively uniform in characteristics. The complexities of the ocean
circulation can be dived up into two classes – “wind-driven” and “thermohaline”
circulation [85]. Surface currents are mostly driven by heating and wind-related
forcing, as the passage of the wind over the ocean surface will move the upper water
with it. For many ocean regions, winds are consistently in one direction (e.g., the
trade winds) and this has set up an overall surface ocean large-scale circulation,
which also results from the fact that the water is being moved on a the Earth
spinning on its axis, and therefore the water movement is also influenced by the
“Coriolis effect.” This results in the water moving to the right of the wind direction
in the Northern Hemisphere, and vice versa. In addition to this direct effect, when
there is movement of water away from a coast or other barrier, conservation of mass
will require the transport of water from depth to the surface to replace this water.
Such “upwelling” of deeper water is an important part of the upper ocean circula-
tion. Additionally, upwelling can occur at the equator as water masses are moving
away from the equator due to the Coriolis force (there is a water mass divergence at
the equator). Similarly, at higher latitudes, convergence and sinking upper air
masses results in divergent air flow over the ocean surface, which forces conver-
gence of ocean waters, and sinking of water masses which feed the intermediate
waters of the main ocean basins. These sinking intermediate waters travel laterally
toward the equatorial regions.
Such wind-driven circulation is enhanced and complemented by the thermoha-
line circulation that is driven by density differences in water masses, which is
primarily a function of temperature but also related to salinity. As surface waters
move away from the Equator they are heated and net evaporation increases salinity.
304 R.P. Mason
When these water masses reach the mid-latitudes in winter, their cooling increases
their density so that they become more dense that the waters below and sink. These
sinking waters then travel in the main thermocline region (500–100 m) back toward
the equator forming a cyclonic circulation that feeds upwelling at the equator. In the
high North Atlantic and around the Antarctic Peninsula, surface waters become
dense enough to sink to the deep ocean as formation of polar ice additionally
increases the salinity of the surface waters, and cooling is greater. These sinking
waters reach the deep ocean (>2,000 m) and supply and drive the deep water
circulation of the ocean, and transport chemicals on a journey that takes 1,000 years
on average. Overall, sinking water in the North Atlantic is transported south into the
South Atlantic, skirting the Southern Ocean and passing through the Indian Ocean,
and finally moving northward from the South to the North Pacific. To complete the
water mass cycle, waters are generally transported upward (net upwelling) in the
North Pacific Ocean to the surface, completing the water cycle.
The distribution of the major important characteristics of ocean waters that
results from this circulation is illustrated by the data in Fig. 12.3 [93]. The salinity
distribution along this Atlantic Ocean transect shows the presence of the interme-
diate water that has sunk from the surface in the South Atlantic and is moving north
toward the equator. A similar process occurs in the North Atlantic and is also shown
by the difference in salinity although the distribution is more diffuse. Sinking water
in the Southern Ocean near Antarctica can be seen as deep water in the Southern
Hemisphere. Similarly, water sinks in the North Atlantic around Iceland but this
signal is not as evident in the salinity, but is shown in the phosphate distribution.
The phosphate distribution also shows very clearly that its concentration has been
depleted in the surface waters and is higher at depth. It also shows the two sinking
water masses from the southern ocean regions which have higher phosphate
concentrations than their surrounding waters. The oxygen distribution clearly
shows the very low concentrations in the intermediate waters around the equator
– due to oxygen consumption during organic matter remineralization. The south
flowing deeper waters, that forms through water sinking in the North Atlantic
Ocean, show a decreasing oxygen concentration as they progress south, further
illustrating the consumption of oxygen in these deep waters due to organic matter
degradation.
The movement of the water masses in the deep ocean can be tracked using the
major chemical and nutrient distributions, as discussed in reference to Fig. 12.3, but
there are also many other ways to track and age these water masses. The so-called
“tracers” are generally of two types: (1) radioactive chemicals whose concentration
will change over time due to radioactive decay [61, 78, 112] and (2) chemicals
whose input to the ocean is well known and transient [38]. The radioactive elements
can be used as “clocks” as their rate of decay is first order and well known and
therefore the ratio of the source (“parent”) and product (“daughter”) can give
information on the time since the element was added to the ocean. Also, as many
of the parent-daughter components have different reactivity or solubility, their
relative concentrations can tell something about ocean processes. One example is
the decay of uranium (U) that is very soluble to thorium (Th), which is very particle
(a) Salinity (pss-78)

0
37
36.5
2000
Depth (m)
36
Ocean Data View

35.5
4000
35
34.5
6000 34
40°S 20°S Eq. 20°N 40°N
(A) 0
(b) Phosphate (µmol kg−1)
12
0 2.5
Pressure (dB)
2 2000 20
2000 30
1.5
Depth (m)
Ocean Data View

4000
1
4000
0.5
6000
–5 5 15 25 35 45 55 65
6000 0
40°S 20°S Eq. 20°N 40°N 60°N (B) 0
(c) Oxygen (µmol kg−1) 30 30
0
300
Pressure (dB)
2000
250 >30 years
2000
Depth (m)
Ocean Data View
200
4000
150
4000
100
6000
6000 –30 –25 –20 –15 –10 –5 0 5 10 15 20 25 30 35 40 45 50
a 40°S 20°S Eq. 20°N 40°N 60°N b Latitude
Fig. 12.3 (a) Meridional sections along the WOCE section A16 in the Atlantic Ocean for salinity,
phosphate, and oxygen, with the cruise taken shown in the insert, indicating the vertical and
horizontal distributions in the ocean, which result from ocean circulation, biological activity, and
particulate remineralization; (b) Sections in the eastern Atlantic Ocean of the ratio of two
chlorofluorocarbons (CFC-11 to CFC-12) along 20 W from 65 N to 5 S in summer 1998 and
along 135 W in the eastern Pacific ocean from 54 N to 32 S in summer 1991. The first figure set is
reprinted from “Inverse Modeling of Tracers and Nutrients,” a chapter by R. Schlitzer,
pp. 188–199, and the second figure set from “CFCs in the Ocean,” a chapter by R.A. fine,
pp.155–163, both published in Marine Chemistry and Geochemistry, Steele, J.H., Thorpe, S.A.
and Turekian, K.K. (Eds), a derivative of the Encyclopedia of Ocean Sciences, 2nd Ed., Elsevier,
Amsterdam, copyright (2010). Data are taken from various publications as indicated in the original
figures
reactive and removed from the ocean by sinking particles [2, 19]. The resultant
deficit in the ratio compared to that expected based on their decay rates can be used
as a measure of the rate of particle removal from the ocean [22].
Some human-produced chemicals are also useful tracers. An example of this is
the chlorofluorocarbons (CFSs), which were widely used for a number of industrial
applications but which are now banned. Their extent of use and release to the
environment is well known and so these compounds can be used to trace ocean
water mass movement. An example is shown in Fig. 12.3, which shows the
distribution of the ratio of CFC-11 and CFC-12, and the estimated age of the
water in years based on this ratio [38]. This shows that the waters of the deep
North Atlantic are relatively “young” as they are formed by sinking of surface
ocean waters within the last 50 years. In contrast, in the Pacific, only the upper
water masses show the presence of these chemicals and the deep waters show little
306 R.P. Mason
Constituent/Salinity Concentration Concentration

0 0 0
1 1 1
Depth (km)
2 2 2
3 3 3
4 4 4
5 5 5
Fig. 12.4 Diagrammatic depiction of the distribution of a chemical that has (a) a conservative;
(b) a nutrient-like; and (c) a scavenged profile in the ocean
evidence. Again, these distributions concur with the understanding of deep ocean
circulation.
Other important tracers of ocean circulation are the radioactive elements
released due to above ground atomic weapon testing (137Cs, 14C, 3H) and the
radioisotopes derived from the decay of U and Th (their specific isotopes and
those of radium (Ra) (multiple radioisotopes), radon (222Rn), 210Pb and others)
[112], and cosmogenic radioisotopes (14C, 3H, beryllium (Be) isotopes and many
others) [113, 114].
Factors Controlling Ocean Chemical Distributions
The distribution of an element or chemical in the ocean has a characteristic

distribution that relates to its solubility, degree of reaction, its requirement as
a nutrient, and other factors (Fig. 12.4). Additionally, as discussed further below,
many elements and chemicals are converted between different forms due to photo-
chemical processes in the surface ocean, due to redox changes in the low oxygen
intermediate waters, or due to their association with DOM. For many metals,
complexation to dissolved ligands plays an important role in defining their concen-
tration and distributions. The importance of mixing and other physical transport
processes must also be considered, as detailed above. The overall range and average
concentrations for a number of important elements or compounds are given in
Table 12.1. For the elements, their major form and state in seawater is also given.
A compilation of distributions for all the elements in the periodic table for the North
Pacific has been compiled by Nozaki [81] (Fig. 12.5).
The major ions are found at a constant concentration across all the major ocean
basins, and their distribution in the ocean has been termed a “conservative” distri-
bution (Fig. 12.4) [17, 21]. Except for rare and specific circumstances in specific
locations, all the major ions have this type of distribution. To exhibit such
Fig. 12.5 A depiction of the vertical distributions of the elements in the North Pacific Ocean.
Originally published in EOS supplementary material in 1997, for an article by Y. Nozaki “A fresh
look at element distributions in the North Pacific”, copyright 1997, American Geophysical Union.
Reprinted with permission from the American Geophysical Union
a distribution, their residence time of the element or compounds must be much

longer than the mixing time of the ocean (which is around 1,000 years) so that they
can become well mixed throughout the ocean basins, and most conservative
elements have residence times of >106 years [17]. Most conservative elements
are found at relatively high concentrations compared to their crustal abundance and
this is because of their slow rate of removal from the ocean.
Elements or chemicals with a “nutrient”-type distribution are those whose
concentration is depleted in the surface ocean due to their uptake into biomass,
especially into photosynthetic microorganisms, and their removal from the upper
waters by the sinking of particulate material (dead cells and organisms, fecal
pellets). Decomposition of this material in deeper waters leads to a rapid increase
in concentrations at mid-depth in the ocean and higher and relative uniform
concentrations in the deep ocean, and this decomposition depletes the oxygen in
these mid-depth regions. Furthermore, if these compounds and elements persist in
the deeper waters and are not effectively removed by deep ocean particle scaveng-
ing, their concentration in the deep Pacific Ocean will be greater than that of the
deep North Atlantic given the increased input of these chemicals over time from
above as the water masses move through the ocean basins. Both nitrate and
phosphate show such a distribution, as do other micronutrients [21].
A correlation between the concentrations of Zn and Si, and cadmium (Cd) and P,
in the ocean was noted in the early 1980s (Fig. 12.6), suggesting that similar
308 R.P. Mason
a Nitrate (µmol kg−1) b Phosphate (µmol kg−1) c Silicic acid (µmol kg−1) d Zinc (nmol kg−1)
0 10 20 30 40 50 0 1 2 3 4 0 50 100 150 200 0 2 4 6 8 10
0 0 0 0
Pacific
1000 1000 1000 1000

Depth (m)
2000 2000 2000 2000
3000 3000 3000 3000

Atlantic
4000 4000 4000 4000
e Cadmium (nmol kg−1) f Nickel (nmol kg−1) g Copper (nmol kg−1) h Manganese (nmol kg−1)
0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 2 4 6 8 10 12 0 1 2 3 4 5 0.0 0.5 1.0 1.5 2.0 2.5 3.0
0 0 0 0
1000 1000 1000 1000

Depth (m)
2000 2000 2000 2000
3000 3000 3000 3000
a 4000 4000 4000 4000
−1
Dissolved Fe (nmol kg )
0.0 0.2 0.4 0.6 0.8 1 1.2 1.4
1,000
Depth (m)
2,000
3,000
b 4,000
Fig. 12.6 (a) The vertical distributions of nutrients (N. P, Si) and metals (Zn, Cd, Cu, Ni, Mn) in
the North Pacific and North Atlantic Oceans; (b) A similar plot for Fe for both oceans. The first
figure set is reprinted from “Trace Metal Nutrients,” a chapter by W.G. Sunda, pp. 17–28,
published in Marine Chemistry and Geochemistry, Steele, J.H., Thorpe, S.A. and Turekian, K.K.
(Eds), a derivative of the Encyclopedia of Ocean Sciences, 2nd Ed., Elsevier, Amsterdam,
copyright (2010), with permission from Elsevier. The iron figure is from K. Bruland and M.C.
Lohan “Controls of trace metals in seawater”, pp. 23–47, a chapter in The Oceans and Marine
Geochemistry, H. Elderfield (Ed), volume 6 in Holland, HD and Turekian, KK (Exec Eds),
Treatise on Geochemistry, Amsterdam, Elsevier. Copyright (2004). Data are taken from various
publications as indicated in the original figures
mechanisms control their ocean distribution and this has since been attributed to
their “nutrient-like” behavior [21, 32]. Zinc has multiple roles in cellular metabo-
lism and, most importantly, it is part of an important enzyme, carbonic anhydrase in
marine phytoplankton, which is required to interconvert CO2 and HCO3 within
cells [76]. Thus, Zn is actively accumulated into plankton in surface waters and is
released back into the water at depths when the organic matter decomposes, and has
a similar profile with depth to the nutrient, Si. Deep water concentrations of both Si
and Zn are higher in the Pacific Ocean compared to the Atlantic Ocean (Fig. 12.6)
and this reflects the continued buildup of these elements during the “deep conveyor
belt” circulation [17]. The correlation of Cd and P profiles was initially puzzling as
Cd was thought of as a toxic metal but it has been subsequently demonstrated that
Cd can actually substitute for Zn and other transition metals in carbonic anhydrase
and therefore has a biological role in the ocean environment [76]. Given this fact,
the correlation with P is explained.
Iron and Copper (Cu) are other micronutrient elements required for enzymes and
other biochemicals but they have a more complex vertical profile as their relatively
high particle reactivity and the insolubility of Fe results in their scavenging from
deep ocean waters by particles. This results in a lack of increase in concentration in
the North Pacific relative to the North Atlantic [18, 76]. Additionally, for Fe, its
redox chemistry can result in changes in redox state in the low oxygen waters that
can further complicate its ocean distribution. Finally, Fe, Cu, and other metals are
released into the deep ocean by hydrothermal activity and this can also influence the
ocean distribution of these metals. Finally, in a similar manner to Cd, Cu is both an
essential element and a potential toxic element and therefore there is a fine balance
between its concentration needed to support growth and that which can hinder
organisms. Other elements in this category are Se, and to a lesser degree As.
Another oceanic distribution is characterized by elements or chemicals that are
particle reactive and/or relatively insoluble, and especially those metals or organic
compounds that also have the majority of their input from the atmosphere
(Fig. 12.4). These are referred to as “scavenged elements” and they have deep
water concentrations that tend to decrease along deep water flow paths between the
Atlantic and Pacific. Hence, their distributional features are often dominated by
their point of entry and removal in the ocean. These elements and compounds have
short residence times (<103 years). For example, surface water concentrations of
Hg and Pb are typically elevated compared to deep waters [14, 40, 92], reflecting
their dominant atmospheric input pathway to the ocean. Such metals are scavenged
and removed from the surface ocean via particle transport, released back into the
water column during the dissolution of the biological particles, but are often re-
scavenged from deep ocean waters and therefore do not have a continuing increas-
ing concentration with depth, and their concentration is depleted in deep ocean
waters, or does not increase with depth, or with time.
Organic chemicals are added to the ocean via a variety of processes. Human-made
chemicals such as pesticides, polychlorinated biphenyls (PCBs) and other industrial
chemicals are typically added to the ocean from both direct runoff and atmospheric
deposition. Some compounds, such as petroleum-derived compounds, are naturally
310 R.P. Mason
formed in the deep Earth but are pollutants when added to the ocean in high
concentrations. Most of these compounds are highly insoluble in water and have
a high tendency to associate with particles (POM) and are taken up into
microorganisms, and therefore are effectively removed from the surface ocean by
particle sinking. Such chemicals are also likely strongly removed from solution in
estuaries and coastal waters and therefore their input to the ocean is dominated by
atmospheric inputs. These compounds likely have a scavenged distribution in the
ocean water column. Other pharmaceuticals and organic chemicals are manufactured
to be highly soluble in water but typically have a relatively short half-life before they
are degraded by photochemical processes and microbes, and as many of these
compounds have only been recently (on a ocean mixing timescale) added to the
ocean, their distribution and fate are poorly known. Many of these compounds are
accidentally added to the ocean or are released from coastal sources, such as treatment
plant effluents and stormwater outflows.
In terms of the nutrients, it is often their excessive addition to the ocean as a result
of anthropogenic inputs and human activity that have been the focus of study and
environmental management. For example, excessive amounts of nitrogen, in partic-
ular, in coastal waters has lead to an increase in extent, duration, and severity of
oxygen depletion in deep waters, and the persistence of hypoxic (oxygen content
<4 mg/L) and anoxic waters (undetectable oxygen/presence of sulfide), which have
been termed “dead zones” in many instances in the media and elsewhere. Oxygen
depletion in coastal waters has important consequences and while the formation of
such zones did occur in the historic past, human activity has definitely increased the
problem [115]. However, it is also important to understand the role of nutrient
limitation in ocean chemistry and environmental health. Nutrient limitation is an
important concern for open ocean ecosystems, and this is mainly due to the limita-
tion of Fe input to the ocean. Such limitation has been little impacted by human
activities in the last century and has its origins in the geological evolution of the
Earth in the very distant past. In contrast to Fe, Si limitation is also important in some
ocean locations and this can be linked to human activity as one important source of
Si to the ocean is riverine inputs. While not all microorganisms require silica, it is an
important nutrient for one class of phytoplankton (diatoms) and for other organisms
that construct silicate shells, and therefore its limitation can have an important
impact on coastal and ocean primary productivity. Construction of reservoirs on
many rivers around the world has decreased the Si input to the ocean and therefore
could be an important factor in determining the extent of this limitation [11, 23].
Detailed Description of the Cycling of Important

Elements and Chemicals
Iron, Manganese, and Aluminum Cycling in Open Ocean Waters
Atmospheric and riverine inputs are the main sources of Fe, Mn, and Al (all
crustally abundant) to the oceans (Table 12.1) and much of this input is in the
particulate or colloidal fraction. While Fe is primarily present as FeIII complexes
and solids in oxic seawater, Mn can be found in its more reduced +2 oxidation state,
and the reason for the persistency of the reduced Mn forms is its relatively slow
oxidation rate compared to Fe [63, 65, 66]. Aluminum (Al) ranges widely in
concentration and is elevated in regions of enhanced atmospheric input, such as
the equatorial zones where rainfall is high, and is very low in polar surface waters. It
is present as AlIII (Al(OH)4 , Al(OH)3) [84]. Concentrations are higher in the North
Atlantic (8–30 nM) than the North Pacific (<2 nM) due to its depletion via particle
scavenging. It has high concentrations in the surface Mediterranean (>100 nM) due
to the enhanced atmospheric (dust) inputs in that region. Aluminum is the element
with the largest ocean variation.
While riverine inputs are important, most of it is removed in estuarine and
coastal waters as all these metals are highly particle reactive. Given their increased
insolubility with pH, these elements have low ocean concentrations even though
they are some of the most abundant elements on the surface of earth. The relative
insolubility of both FeIII and MnIV leads to the rapid removal of these metals from
the surface ocean. While both Mn and Fe have a role in the biochemistry of marine
organisms [76], the concentration of Mn is relatively high in ocean waters com-
pared to its nutrient requirement, and depletion of its surface concentration due to
biological activity is not often found. In contrast, as noted above, dissolved Fe can
show surface ocean depletion in regions of high primary productivity.
Representative profile for Fe in the major ocean basins are shown in Fig. 12.6c. In
seawater, the concentration of dissolved Fe is higher than would be predicted based on
its primary solubility in terms of Fe (hyd)oxides due to the presence of Fe-binding
ligands in solution [18, 76]. In the absence of these ligands, the predicted concentration
in equilibrium with Fe(OH)3 (s) is around 0.1 pM, even considering that it is strongly
hydrolyzed in seawater with the dominant inorganic complexes being Fe(OH)42 , Fe
(OH)3 and Fe(OH)2. The average measured concentration is 0.5 nM, three orders of
magnitude higher (Fig. 12.6c). This is consistent with the electrochemical
measurements which estimate that 99% or more of the Fe in surface ocean waters is
complexed to organic ligands. Additionally, given that the measured values are for
filtered waters, it is likely that this value reflects the presence of both organic and
inorganic Fe-containing colloidal phases. In contrast, Mn forms only weak associations
with both inorganic and organic ligands, and its dissolved concentration is much lower
than the concentration expected from precipitation of MnIV hydroxide or carbonate
phases.
312 R.P. Mason
The distributions of Fe and Mn in the water column are modified in the regions
of low oxygen or anoxic waters. In some ocean locations, oxygen concentrations
are sufficiently low that both Fe and Mn can be reduced, either through biologi-
cally mediated pathways or abiotically, and this has been observed, for example,
in the oxygen minimum zone of the equatorial Pacific Ocean [21]. In addition to
these sub-oxic environments, the distributions of Fe and Mn are highly modified
in anoxic environments with elevated levels of the reduced forms of the metals
being present. However, the distributions are complex for Fe because, while in
low oxygen environments Fe(II) is formed with an increase in solubility as sulfide
levels increase, Fe(II) is precipitated as Fe-sulfide phases (FeS and FeS2) and
this leads to a maximum in dissolved Fe in low oxygen, but low sulfide
environments.
In permanently or seasonally stratified systems, the redox cycling of metals can
lead to a large gradient in their concentration, as well as for metals and other
constituents that are strongly associated with the oxic particulate phases of Fe and
Mn. The concentration of particulate Fe and Mn, and associated metals, are
enriched above the interface due to the diffusion of dissolved, reduced Fe and Mn
from below, and their subsequent oxidation and precipitation in the higher oxygen
waters. Sinking of these particulate materials and their dissolution below the
interface leads to a peak in dissolved species below the interface. This cycling
across the interface has been termed the “ferrous wheel” in analogy to the amuse-
ment park ride.
Another location where increased concentrations of dissolved Mn and Fe can be
found is in association with hydrothermal vents, especially for Mn given the slower
rate of MnII oxidation. Elevated concentrations have been found to persist in the
vicinity of hydrothermal vents, in plumes that are enriched in other tracers of such
activity (3He) [21, 43]. While hydrothermal vents are sources of many metals as
discussed above, the removal of metals with the precipitation of Fe and Mn oxides
after emission and oxidation of fluids from hydrothermal systems results in the local
scavenging and removal, and decreases the importance of hydrothermal sources as
input to the global ocean. These processes are important in the formation of oxide
deposits and also in the formation of metal-rich nodules.
The Cycling of the Other First Row Transition Metals,

Zinc, and Cadmium
The distribution of Zn in open ocean waters appears to be strongly controlled by

biogeochemical processes, and its distribution is that of a “classic” nutrient metal.
The release of Zn from decomposing organic material in the sub-thermocline
waters results in higher concentrations in deeper waters and the relationship
between the concentration of Zn and that of Si [116]. As the main source of Zn to
the ocean is riverine and terrestrial input (Table 12.1), and the atmospheric input is
a relatively small component, depletion of surface waters is more likely. Zn is

relatively soluble in seawater in the presence of inorganic ligands and has an
average concentration of 5 nM (Table 12.1), greater than that of Fe and Mn,
which are much more abundant elements in the terrestrial environment and in
riverine inputs. Additionally, there is no important redox chemistry for Zn in
ocean waters, and Zn forms weak complexes with inorganic ligands. Complexation
of Zn by dissolved organic ligands likely also stabilizes its dissolved concentration
and it is suspected that these ligands are directly or indirectly produced by microbial
organisms, even if they may not be produced specifically for Zn complexation.
Even given its direct biological role, Zn has a relatively long residence time
compared to the other first row transition metals that exist in solution predominantly
as cations.
As Zn is present in the absence of organic ligands mostly as the free ion in
seawater (Table 12.1), it is relatively easily acquired by organisms and therefore
complexation with specific organic ligands to enhance uptake is likely not needed.
Therefore, complexation of Zn by organic matter does not appear to provide
a unique advantage to its uptake in contrast to Fe, which is a demonstrated limiting
nutrient and therefore is acquired actively through organic complexation by some
microbes, and Cu, which can be toxic to some organisms at higher ocean
concentrations, and complexation can result in a reduction in toxicity.
In a similar fashion to Zn, the distribution of Cd, which is from the same group in
the Periodic Table, in open ocean waters is strikingly similar to that of phosphate
(Fig. 12.6b). As with some other elements, Cd can be both a toxic metal at high
concentration and have a beneficial role at low concentration because of its ability
to be incorporated into and act as an active metal center in enzymes. Therefore, Cd
is considered a nutrient metal and there is evidence of its depletion from surface
waters in many ocean regions. Similarl to Zn, Cd inputs to the ocean are not
dominated by the atmosphere, but in contrast to Zn, the inorganic chemistry of
Cd in the ocean is dominated by chloride complexation and the relative fraction of
Cd as the free metal ion is a few percent. It is also apparent that Cd is complexed to
organic ligands in the ocean but the relative degree of complexation is lower than
that of Zn [21] Overall, the major difference in speciation in seawater is likely not
due to differences in the relative binding capacities to organic matter, but the
differences in the binding to Cl. The concentration of Cd in the ocean is relatively
low compared to other transition metals. Its residence time in the ocean is not well
characterized as there is little information on its distribution through all the ocean
basins. Given its sources and biogeochemical cycling, it is probable that it has
a similar residence time to Zn and other relatively soluble transition metals.
Cobalt has received recent attention due to the potential importance of Co as
a cofactor in cobalamin and other enzymes. The uptake of Co by microbes,
especially in surface waters, could lead to differences in the distributions of Co
and Ni, another transition metal with similar chemistry and sources. Copper is also
a required nutrient metal and shows a distribution that is a mixture of that expected
for a nutrient element, modified because of the tendency of Cu to complex with
organic matter which enhances its solubility and residence time in deep waters.
314 R.P. Mason
Many studies have demonstrated the importance of organic complexation for Cu

[21, 31]. It has been speculated that these ligands are biologically derived by certain
microorganisms, such as cyanobacteria which are susceptible to Cu toxicity. This
has been demonstrated in a few instances (e.g., [35]). Inorganically complexed Cu
is a very small fraction of the total in deeper waters. As noted in Table 12.1,
inorganic Cu is primarily present as the neutral CuCO3 complex in seawater.
In contrast, both Ni and Co form much weaker complexes with inorganic ligands
and are present predominantly as free metal ions in seawater. These differences in
speciation likely impact the mechanisms of accumulation of these metals into
microorganisms. Most transition metals have been shown to have a biochemical
role, with Cu being incorporated into many enzymes, while the involvement of Ni is
relatively small. The concentration of Ni (8 nM) is higher than that of Cu, while
the concentration of Co is sub-nM. While both Ni and Co are found organically
bound, the relative fraction is smaller than for the other transition metals discussed,
being 30–50%, or greater, in various open ocean locations.
The presence of cobalamin in surface ocean waters has been recently
demonstrated although it is present at low pM concentrations [83]. It is possible
that most of the organically complexed Co in seawater has a biological origin. The
identification of specific biochemicals containing metals raises the possibility that
some of the metal, identified as complexed to organic ligands, has been released
into the environment as a result of cell leakage, purposeful cellular export (e.g.,
metals bound to metallothioniens and phytochelatins), cell death, from the release
due to grazing or even from fecal material. Alternatively, these ligands could be
purposely released into solution to aid in metal assimilation, as occurs for Fe (e.g.,
release of siderophores), or to reduce metal toxicity, as discussed for Cu above.
Another example is the presence and assimilation of heme (Fe-containing)
compounds as it has been shown that microbes can assimilate heme in laboratory
cultures [56].
Anthropogenic Metals: Lead, Silver, and Mercury
The three most important heavy metals in the marine environment are Pb, Hg, and
Ag. Studies of Pb reflect its known input to the atmosphere from Pb compounds in
leaded gasoline, as well as its presence in other anthropogenic sources. The
examination of Hg is driven by human and wildlife health concerns from the
toxicity and bioaccumulation of CH3Hg in marine food chains. Silver is also
known to be toxic to organisms but its levels in the open ocean are generally low
(<1 nM) [88]. Historically, elevated Ag in the coastal environment was considered
an indicator of sewage and anthropogenic-related inputs [91], because of its exten-
sive use in the photographic industry. This is now less the case due to changes in
photographic technology (i.e., the emergence of digital cameras).
All these metals bind strongly to particles and are emitted to the atmosphere
from anthropogenic sources although both Pb and Ag will be removed to a degree
from stacks by particulate emission control devices. Of these elements, Hg has the
highest relative natural source as it can be emitted from various terrestrial
environments in its gaseous elemental form. For the open ocean, atmospheric inputs
are important but additionally coastal inputs also contribute both Hg and Pb from
terrestrial runoff and point source inputs. Mercury is different as is can be present as
a dissolved gas (Hg0) in seawater and therefore air-sea exchange involves both
deposition and gas evasion.
Concentrations of Ag in the ocean vary from low pM values typically found in
the surface ocean to 100 pM or more in deep waters, and concentrations vary
substantially between ocean basins [79, 89, 108, 109] (Fig. 12.7). Silver typically
has a nutrient-type profile and it has been shown that its vertical distribution mimics
that of Si, suggesting its incorporation into the more recalcitrant tissues of microbes
and other organisms in the surface ocean and its relatively slow release from
sinking particulate material [89, 109]. In the deep ocean, there is continual net
increase as Ag is being continuously released from settling particulate matter but is
not being scavenged in deep waters in a similar fashion to other metals, making the
distribution of Ag similar to that of Al. Additionally, both elements are supplied to
the ocean mainly from river and coastal sources, with atmospheric deposition being
a secondary source [49, 88]. Concentrations are increasing over time in the oceans.
A comparison of intermediate waters in the North Pacific Ocean suggests that Ag
concentrations have increased in the last 20 years, reflecting increased anthropo-
genic inputs [89]. In contrast, a similar comparison for Hg concluded that there had
not been a substantial increase in concentration over that time period [63], although
more recent evidence suggests that concentrations have increased more recently
[101].
A number of ongoing and detailed studies have documented the overall contam-
ination of the ocean by anthropogenic Pb and have demonstrated the resultant
decrease in ocean Pb as a result of the phasing out of Pb additives for gasoline
[107] (Figs. 12.2 and 12.8). However, there is still Pb input to the atmosphere from
other anthropogenic sources and current studies are evaluating the extent of this
input. One unique aspect of examining Pb geochemistry is the fact that it has
numerous isotopes, and these isotopes are daughters of the uranium-thorium
(U–Th) decay series: 206Pb is the stable product of the 238U decay series, 207Pb
the stable product of the 235U decay series, and 208Pb is the stable product of the
232
Th decay series. Thus, the isotopic ratio of Pb in the environment is altered due to
the presence of either U or Th in the medium. It has been demonstrated that Pb from
different locations has ratios which are different enough to track the sources of the
Pb in the ocean and other environments as the different isotope ratios allow for
more resolution of the source signals (Fig. 12.8b). Data for the South and equatorial
Atlantic Ocean collected in 1996 is plotted against the various potential source
signals and demonstrates a range of potential sources, including inputs from North
America, large rivers, and coastal sources. In contrast, the data for the deep waters
of the North Pacific, which represent the “oldest” marine water masses, shows a Pb
Ag (pM)
0 10 20 30 40 50 60 70 80 90 100
0
North Atlantic
South Atlantic
North Pacific
1000
2000
Depth (m)
3000
4000
a 5000
Ag (pM)
0 10 20 30 40 50 60 70 80 90 100
0
1000
2000
Depth (m)
3000
4000
Murozumi, 1981 (20°S, 160°W)

5000 Martin et al., 1983 (36°N,123°W)
Martin et al., 1983 (18°N,108°W)
Zhang et al., 2001 (40°N,145°E)
Station 9 (22°N, 158°W)
b 6000 Station 2 (44°N, 155°E)
Fig. 12.7 (a) Distributions of silver in the various ocean basins and (b) Changes in the
concentrations within the water column in the North Pacific over time. Taken from M.A. Ranville
and A. R. Flegal “Silver in the North Pacific Ocean”, published in Geochemistry Geophysics
Geosystems volume 6, Art. No. Q03M01. Copyright (2005), American Geophysical Union.
Reprinted with permission as stated by American Geophysical Union
a 60°N
40°N
20°N
EQ
Ocean Data View

20°S
180°E 170°W 160°W 150°W 140°W 130°W 120°W
Pb(pM)
b
70
1000 60
50
2000
Depth
40
3000
Ocean Data View

30
20
4000
10
a 5000
EQ 10°N 20°N 30°N
1.25
1.20
Pb
206/207
US aerosols
Canada aerosols
Japanese aerosols
1.15 Chinese aerosols
This Study
Pacific Fe-Mn nodule
Seawater (Flegal, 1984)
Central American aerosols
Vietnamese aerosols
1.10
1.95 2.00 2.05 2.10 2.15 2.20
b 208/206
Pb
Fig. 12.8 (a) Cruise track and horizontal and vertical distributions of Pb collected during the
cruise; (b) Relationship between the various isotope ratios of Pb for the samples collected during
the cruise with comparison to other materials for which isotopes have been measured, and which
could be sources for Pb in this region. Taken from J.F. Wu, R. Rember et al. “Isotopic evidence for
the source of lead in the North Pacific abyssal water” published in Geochimica et Cosmochimica
Acta volume 74(16), pages 4629–4638. Copyright (2010), and published with permission from
Elsevier
318 R.P. Mason
Seawater Inorganic Hg Equatorial Pacific

90°N 0
400
60°N
200
30°N 300
0° 400
30°S 200
600
60°S
Ocean Data view

800 100
90°S
180° 120°W 60°W 0°W 60°E 120°E 180°
pM 1000
170°W 160°W 150°W 140°W 130°W 120°W 110°W 0
0.00 0.50 1.00 1.50 2.00 2.50
North Pacific
Atmospheric Wet + Dry HgII and HgP Deposition
90°N
200 400
60°N
30°N
400 300
0°
30°S 600
200
60°S
800
Ocean Data view

90°S 100
nmol/m2/yr
0.00 37.5 75.0 113 150 1000
a b 30°N 40°N 50°N
Fig. 12.9 (a) Modeled concentrations of inorganic Hg and of wet and dry deposition to the ocean
surface for the global oceans, using the GEOS-Chem model framework. Circles in the figures
represent the location and concentrations obtained during field campaigns; (b) Data for the
concentrations of methylated species (both monomethylmercury and dimethylmercury) in the
equatorial and North Pacific Ocean upper waters. The first figure was derived from A.L.
Soerensen, E. M. Sunderland, et al. “An improved global model for air-sea exchange of mercury:
High concentrations over the North Atlantic” published in Environmental Science & Technology
volume 44(22), pp. 8574–8580. Copyright (2010) and reprinted with permission from the Ameri-
can Chemical Society. The second figure was compiled by Elsie Sunderland of Harvard University
using information in Mason and Fitzgerald [68] and Sunderland et al. [101]
signal reflective of inputs via particle settling and from the anthropogenic enrich-
ment of atmospheric deposition.
Most of the input of Pb to the ocean is atmospheric (Table 12.1) and the
distribution of Pb in the ocean typically has a maximum concentration in the surface
ocean and lower concentrations at depth (Fig. 12.8) [107], as well as higher
concentrations in the Atlantic Ocean compared to the Pacific. The concentrations
in the upper waters of the North Atlantic in the vicinity of Bermuda have been
decreasing in the last 30 years (see Fig. 12.2) [106] from values above 150 pM in
the late 1970s to concentrations around 50 pM today. It can be concluded that while
the surface waters near Bermuda have decreased in concentration they are still
substantially elevated above background and reflect the continual input of Pb from
combustion and other industrial sources.
For Hg, both box and numerical models of the global cycle support the notion
that anthropogenic releases of mercury (Hg) to the environment have impacted the
biosphere substantially, enhancing deposition by a factor of 3–5, and influencing
ocean concentrations ([39, 74, 70, 96, 102]). Besides atmospheric inputs, local and
regional-scale contamination of the coastal zone by Hg has occurred due to runoff
from the terrestrial environment and from point source inputs [103, 117]. Mercury
distributions in surface waters reflect the magnitude of the atmospheric deposition
source and the strength of local removal process (scavenging and gas evasion)
superimposed on water circulation [63, 101, 102]. In some regions, concentrations

change seasonally depending on the variability of atmospheric deposition, evasion,
and removal of Hg by particulate sinking. Recent model output shows the potential
variability that could exist for various Hg species in the ocean (Fig. 12.9a shows
modeled inorganic Hg distributions). Concentrations vary between ocean basins
due to differences in the relative impact of anthropogenic Hg in deposition
(Fig. 12.9a), and these are changing over time – recently, more anthropogenic
inputs from Asia and less from North America and Europe. Surface water
concentrations are higher in the Atlantic Ocean and Mediterranean Sea than in
the Pacific Ocean [25–27, 68, 69, 71, 73, 101, 118], and there is evidence for recent
decreases in concentration in the North Atlantic and the Mediterranean Sea
and these are consistent with model predictions [97, 98, 102].
The upper ocean Hg concentration can be transient, changing on short timescales
due to changes in atmospheric inputs and seasonal water mixing [63]. For example,
the input of Hg from the atmosphere was demonstrated through an increase in the
concentration of Hg in the waters of the North Pacific Ocean seasonal mixed layer
during the summer. However, this signal is eliminated as a result of deep water
mixing in the fall. Seasonal mixing and latitudinal transport of sinking surface
water within the permanent thermocline is a mechanism for the transport of Hg
deposited at higher latitudes, which may have been deposited to the ocean a decade
or more previously [85] to tropical and other regions ([72, 102]). The higher
concentrations in the more recent data suggest the input of Hg to these mid-depth
waters, and this coincides with an increasing anthropogenic signal from Asia, and
such changes are not evident in the deep ocean waters [63, 101]. Such upper ocean
cycling confounds the understanding of how the Hg concentration in ocean surface
waters has changed as a result of increased anthropogenic inputs, and model
predictions suggest that the response time of the ocean to changes in atmospheric
Hg concentrations is decadal or longer [97].
The distribution of Hg in ocean waters reflects the sources and cycling as well as the
internal cycling of mercury, methylation, demethylation, oxidation, and reduction. The
biological production and destruction of methylated Hg species, primarily CH3Hg and
(CH3)2Hg in the ocean, is important as the methylated forms are more toxic. In
numerous profiles there appears to be an enhancement in the concentration of
methylated Hg (CH3Hg and (CH3)2Hg) at mid-depth [52, 101, 119, 122]
(Fig. 12.9b). The analysis of the data and the relationships to environmental parameters
suggest that the profiles can be best explained in terms of production of methylated Hg
during the decomposition and remineralization of organic matter [52, 101]. The more
recent studies examining the production of methylated Hg suggests that the presence of
methylated Hg in low oxygen waters is more due to the same factors that cause the
oxygen depletion – heightened bacterial activity, and slow vertical mixing, lack of
water ventilation, and particulate scavenging – than to the activity of particular
microorganisms [67, 101]. In freshwater and coastal environments and sediments,
sulfate and iron-reducing bacteria have been demonstrated to be the most important
methylating organisms [8, 120]. While little is known about the microorganisms or
processes whereby Hg is methylated in the ocean, the fact that methylation has been
320 R.P. Mason
demonstrated in relatively oxygenated waters, and the dominance in many instances of

(CH3)2Hg, suggests that the pathways may be very different from those for freshwater
and coastal environments, and in sediments. The net concentration of methylated Hg
reflects the complex interactions that occur throughout the water column: production
and destruction of methylated Hg by microbial processes and abiotic mechanisms;
scavenging and release from particles; and uptake into the food chain.
Overall, there are strong similarities between the three major heavy metals in
terms of their strong association with organic matter and the particulate phase.
Mercury and Ag have similarities in that they are both Class B metals and form
strong associations with reduced sulfide, and exist in seawater as chloride
complexes in the absence of NOM. Lead can also be associated with sulfides and
forms strong complexes, but to a lesser degree. Also, Pb does not form strong
chloride complexes and is found as the neutral PbCO3 complex in the absence of
NOM (Table 12.1). Atmospheric sources are all important for these metals to the
open ocean, and this is primarily related to the fact that their global cycles have
been substantially impacted by anthropogenic sources. Besides Hg, the accumula-
tion and fate of these elements in ocean microorganisms and in the oceanic food
chain has been little examined and this could be an important area for future
research.
Metalloids and Other Important Elements
The main metalloids of interest in marine systems (e.g., As, Se) exist as oxyanions, and
in a number of oxidation states (Table 12.1), although they are also found as
methylated compounds, or even as larger metalloid-containing species, such as
arsenobetaine and selenoproteins. These organic species and their formation
mechanisms are not detailed here but their biochemical formation is relatively well
known. Arsenic can be found as either AsIII or AsV and as mono-, di-, and tri-methyl
arsenic in marine waters, and similar forms and speciation are found for Se and the
other metalloids (Sb, Ge). It is thought that the methylation of As is a detoxification
and elimination mechanism for As from phytoplankton as AsV can be taken up
inadvertently by microorganisms in low phosphate waters (both exist as polyprotic
acids with similar pKa’s). Methylation involves pre-reduction of the AsV to AsIII
before being methylated. This is a different mechanism to Hg methylation by sulfate-
reducing bacteria. There is also often AsIII present in conjunction with phytoplankton
in surface waters due to this reduction pathway, which is contrary to what is expected
based on thermodynamic equilibrium calculations. However, AsIII is a small fraction
of the total, as are the methylated species [28, 29]. In estuarine environments and
freshwaters, AsIII and the methylated forms can be a larger fraction of the total
dissolved As [53, 80].
The two main inorganic redox states of Se appear to cover a similar range in
concentration, with deep waters having a ratio of SeIV/SeVI of >0.5, but <1. Both
inorganic species appear to be depleted in the surface waters, likely due to their
uptake and incorporation into biota. Selenium is an essential element although it is

only required at low concentrations. Most of the bioorganic Se compounds are
proteins and are therefore being continually produced through microbial processes.
Overall, there has been little recent study of the inorganic and organic speciation
of the metalloids in the ocean water column. It is probable that new insights and
understanding could be gained from the examination of the various fractions in
more detail, especially the “organic fraction.” It is not clear whether these
compounds are derived directly from microorganisms and other biota, or are
primarily produced during organic matter remineralization. It has however been
shown that microorganisms contain small and large molecular weight metalloid-
containing molecules.
Cycling of Organic Chemicals in the Ocean
There have been few detailed studies of the ocean distribution of organic chemicals
and this relates both to their difficulty of measurement and the lack of research
focus. However, the cycling and fate of other organic chemicals can be estimated
based on the knowledge of these chemicals if their general chemical characteristics
are known. The physiochemical parameters needed to predict their fate and trans-
port include: (1) their solubility and volatility; (2) their Henry’s Law Constant (KH),
which defines their relative solubility in water versus air; (3) their dissolved-
particulate partition coefficient, typically normalized to particulate organic content
(KOM); and (4) their octanol-water partition coefficient (KOW) which is a measure
of the hydrophobicity of the compounds.
The solubility of organic compounds ranges widely but most compounds of
environmental interest are relatively insoluble in seawater. Additionally, their
environmental concentrations do not approach their saturated solubilities (from
10 2 M for small chain compounds to 10 11 M for some PCB congeners; [94]).
Solubility is a strong function of the composition of the organic chemical
as compounds that can be ionized in water due to the presence of acidic or basic
groups (e.g., carboxylic acids, amine, and other N-containing groups, phenols and
alcohols and thiols) will be more soluble that a similar compound without an
ionizable group. The pH of seawater is around 8.2, it has high ionic strength
and water molecules are polar and therefore ionizable compounds will tend to
dissociate or be protonated to some degree as a result. Nonpolar compounds are
the most hydrophobic as there is little interaction between the compounds
and water.
Related to solubility is the air-water partition coefficient (KH = partial pressure
(Pi)/dissolved water concentration (CW)), which for environmental science can be
considered as a measure of the potential for a compound to be lost from solution via
gas exchange. Compounds with high values are volatile, but the degree to which
a compound would be lost from water to the air is a function of KH and the relative
concentration of the compound in each phase. Aliphatic hydrocarbons have
322 R.P. Mason
(mol · kg−1 om)

(mol · L−1)
4
log Kom = 0.82 log Kow + 0.14
3 r2 = 0.93 N = 34
log Kom
2
0
1 2 3 4 5 6 7
a (mol · L−1 octanol)

log Kow
(mol · L−1 water)
PCB usage
90°N
60°N
january
30°N
0°
30°S
july
60°S
90°S
180°W 135°W 90°W 45°W 0° 45°E 90°E 135°E 180°E
Soil Conc (pg)
0 20 40 60 80 100 120 140 160 180 200
Inventory in soil or ocean mixed layer / Inventory in atm boundary layer

b PCB 101
Fig. 12.10 (a) Relationship between the octanol-water partition coefficient and the dissolved-
solid partition coefficient (normalized to organic content) for a variety of organic chemicals.
Taken from R.P. Schwarzenbach, P. M. Gschwend, et al. “Environmental Organic Chemistry”
published by John Wiley and Sons. Copyright (1993) and reproduced with permission; (b) the
distribution of PCB in various reservoirs of the biosphere: the historical usage and soil
concentrations are shown in the left panel; the model predictions of the maximum reservoir
capacity in the center figure, with the right figure showing the particular distribution for the
Atlantic Ocean. Taken from R. Lohmann, K. Breivik, J. Dachs and D. Muir “Global fate of
POPs: Current and future research directions” published in Environmental Pollution, volume 150,
pp. 150–165. Copyright (2007) and published with permission from Elsevier; (c) the global spatial
variability for PCB 52 in the atmosphere (gas phase) (top), dissolved surface ocean (middle) and
predicted air-sea and sinking fluxes (bottom). Taken from J. Dachs, R. Lohmann, W.A. Ockenden,
L. Mejanelle, S.J. Eisenreich and K.C. Jones “Oceanic biogeochemical controls on
global dynamics of persistent organic pollutants” published in Environmental Science and Tech-
nology, volume 36, pp. 4229–4237. Copyright (2002) and reprinted with permission from the
American Chemical Society
relatively high KH (102–104 atm/M), PCBs intermediate (0.01–1 atm/M), and

PAHs lower values (0.001–0.1 atm/M) [94].
The degree to which organic chemicals will remain dissolved in an environmen-
tal solution or partition into particulate phases, or become associated with high
a 90N 90
75
60N 60
45
30N 30
15
Latitude
0 0
−15
30S −30
−45
60S −60
−75
90S −90
180W 90W 0 90E 180E 0 30 60 90 120 150
CA (pg m−3)
0 20 40 60 80 100 120
pg m−3
b 90N 90
75
60N 60
45
30N 30
Latitude
15
0 0
−15
30S −30
−45
60S −60
−75
90S −90
0 3 6 9 12 15
180W 90W 0 90E 180E
CW (pg L−1)
0 2 4 6 8 10 12
pg L−1
c 90N 90
−90 75
60N 60
45
30N 30
15
Latitude
0 0
−15
30S −30
−45
60S −60
−75
90S 0 0.4 0.8 1.2 1.6 2
180W 90W 0 90E 180E
Flux (ng m−2 d−1)
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

c ng m−2 d−1
Fig. 12.10 (continued)

324 R.P. Mason
molecular weight dissolved organic material, such as humic substances, is

a function of their hydrophobicity which is related to their solubility and is related
to KOW, which is the ratio of the concentration at equilibrium in water compared to
octanol. The dissolved-POM partition coefficient is related to these parameters
(Fig. 12.10a) [94]. Octanol is used as the comparative phase as it is thought to be
a reasonable representative medium for the cell membrane and also likely provides
a reasonable proxy for POM in seawater. Again, the values for organic chemicals
range over a wide scale. The more soluble the compound, the smaller is the KOW
and KOM.
As POPs are hydrophobic, they also attach to particles in the water, and the
degree of association is a function of the particulate organic content, and therefore
can be removed from the surface to deep ocean via sinking particles, especially in
regions of high primary productivity. Thus, the factors controlling the overall
distribution and fate of the more volatile POPs are similar to that of Hg. The degree
to which the deep ocean is being enriched in these compounds is related to the rate
of additional from above compared to their rate of removal to the deep sediment. As
most of the organic matter sinking into the deeper ocean is remineralized, it is likely
that the degree of incorporation of POPs into deep ocean sediments is less than that
for metals and other compounds that form relatively strong associations with
mineral phases.
The oceans play an important role in controlling the environmental transport,
fate, and sinks of many POPs at regional and global scales [50, 58]. While the
atmosphere is the most important global transport mechanism for many POPs,
transport via ocean currents is important for some compounds. Deep ocean waters
are usually considered the final sink for POPs, although an evaluation of the
importance of their role in their environmental fate is uncertain, and modeling of
their fate in the deep ocean has not been considered in a detailed fashion. POPs
bound to sinking POM are exported from the mixed layer and will either accumu-
late in the deep ocean waters, as do the nutrients and some elements (Ag, Al), or
will be finally removed to deep sediments. Marine sediments, and potentially the
continental shelf could represent an important reservoir for POPs [59]. Gustafsson
et al. [51] examined the settling fluxes of PCBs in the coastal and offshore North
Atlantic Ocean by measuring their concentrations on particles and using 234Th to
estimate particle settling velocity. As may be suspected, the ratio of the flux to the
sediment surface compared to that of the export flux from the mixed layer decreased
with increasing water depth, and with the aqueous solubility of the PCB congeners,
suggesting overall release from the dissolution of particles with increasing depth.
Depending on the individual congener, the flux rates ranged from <1 to 10 pmol m 2
day 1 at the offshore stations to almost 100 pmol m 2 day 1 at the nearshore
locations. Such fluxes are comparable to estimates by others. These authors inferred
from the PCB distributions that atmospheric input was dominant source to the open
ocean locations, and that particle scavenging and removal from the mixed layer was
a more important sink for these compounds than photochemical decomposition in
the marine boundary layer. Given the larger surface area of the open ocean
environment, the flux in this region was a bigger fraction of the total input to deeper
waters than the more nearshore waters, even though the fluxes in the shallow
regions were higher [51].
Deep ocean sediments are likely important for many POPs. The relative impor-
tance of removal to the deep ocean and sediments versus reemission to the atmo-
sphere will be different for different chemicals. Records of POP concentrations in
deep ocean sediments are sparse and therefore it is difficult to evaluate for many
compounds their degree of removal to the deep sediment. Estimates for ∑DDT are
that 4–6% of these compounds have been removed to the deep ocean sediments and
that most of the amount added to the ocean during their use is either still resident in
the ocean or returned to the atmosphere.
Water column distributions and knowledge of the inter-basin variability in
concentration of organic contaminants is much less well known that that of
nutrients and metals but there are some similarities in the chemistry and partitioning
that allows inferences to be made concerning their fate and transport in the ocean
and the main processes controlling their internal cycling. Many of the POPs are
sufficiently volatile and insoluble that air-sea gas exchange is an important part of
their global cycle and determines to a large degree the extent of their accumulation
and persistence in the ocean. For compounds that are relatively volatile, their upper
ocean concentration is determined by the rate of input relative to the rate of re-
emission to the atmosphere. For example, for PCBs, their upper ocean concentra-
tion is determined primarily by the gas exchange component with wet deposition
and dry deposition of PCBs being relatively minor components [30]. The impact of
the historical extent of emissions and atmospheric concentration on the ocean
distribution of PCBs is illustrated with the model data shown in Fig. 12.10b. This
further illustrates the importance of gas exchange on controlling concentrations as
for the open ocean atmospheric particulate concentrations are low and these
aerosols are dominated by large sea salt particles which are not effective is trapping
POPs given their low organic content. It has been estimated that approximately
10% of the global releases of PCBs resides in the upper ocean [16], and similarly
the upper ocean is a major reservoir for other compounds such as ∑DDT [99] and
PFOS and their precursors.
Modeling studies suggest that the ocean, particularly in the Northern Hemi-
sphere, was a net sink for ∑DDT during the period of its heightened use but that it is
now a source as the banning and limited use of DDT has lead to a decrease in the
atmospheric burden and as a result a degassing of ∑DDT from the ocean surface to
the atmosphere [99]. The change from sink to source was slowest for the Northern
Hemisphere as the air concentration in the region was buffered to a larger degree by
re-emission of these chemicals from the terrestrial environment. This is because of
both the relatively higher inputs of these chemicals in the Northern Hemisphere as
well as the higher relative land mass in the north. The model results suggest that the
ocean surface in the Southern Hemisphere became a source to the atmosphere in
the late 1980s/early 1990s while the situation was delayed for about a decade in the
Northern Hemisphere. The ocean gyre regions became a source more rapidly that
regions of higher atmospheric inputs due to enhanced wet deposition, or regions
with higher productivity as the enhanced removal of these compounds from the
326 R.P. Mason
surface ocean to deep in association with particulate sinking delayed the buildup in
surface concentrations relative to those in the atmosphere and resulted in a longer
persistence of ocean uptake. This example again illustrates the importance of ocean
circulation and primary productivity as factors that have a large influence over the
overall fate and cycling of POPs.
Given the lack of data for many chemicals in the ocean, there have been several
efforts over the years to model the systems to understand the fate and transport of
these chemicals. Wania and Mackay [105] developed a zonal climatically
differentiated nonequilibrium fugacity mass balance model for POPs including
air, fresh and ocean water, and sediments/soils with advective connections to
examine the main processes involved in the portioning and distribution of these
chemicals across environmental media. Not surprisingly, temperature and chemical
reactivity were highly important variables, as were the transport rates and the
chemical composition, such as the media organic content and other physicochemi-
cal properties (e.g., solubility, KOW, vapor pressure). Dachs et al. [30] further
examined the biogeochemical controls over the cycling of POPs in the ocean and
demonstrated the importance of air-sea exchange, and as a result, temperature in the
cycling of the volatile components across this interface. Additionally, given the
importance of particulate removal in the cycling of POPs in the upper ocean as
a sink, and air-sea exchange as a potential source or sink, they determined that
influx of gaseous POPs to the ocean was actually higher in productive regions as the
particulate removal depleted the surface layers and resulted in net deposition of
gaseous species from the atmosphere. This is shown in Fig. 12.10c, which shows
the model estimated parameters for gaseous concentrations over the ocean,
predicted dissolved surface ocean concentrations, and the extent of air-sea
exchange for a particular PCB congener. These results show that fluxes are into
the ocean overall and are driven by the atmospheric concentration.
Overall, the results of the limited studies of POPs in the ocean indicate that their
cycling is controlled by their degree to which gas exchange at the sea surface is
important in their geochemical cycling. The more soluble compounds will be less
impacted by gas exchange and the more hydrophobic compounds will attach to
particles and be removed from the surface ocean more efficiently, driving further
gas exchange input. Overall, while many of these chemicals have been banned in
many parts of the world, or have now more restricted usage, there is a strong legacy
remaining in the biosphere because of their buildup in the terrestrial soils, and
because of their overall volatility, they can be reemitted to the atmosphere and
continually contribute to the contamination of the ocean. Finally, it is clear that
there are many similarities in the cycling and ocean biogeochemistry of elements
and organic chemicals and that productivity and carbon export from the surface
ocean are main drivers on upper ocean concentrations and the extent to which the
ocean is a net sink. For many chemicals, buildup is occurring in the deep ocean as
a result of human activity and this legacy will remain for an extended period given
the slow deep cycling of the ocean waters.
Future Directions
The above discussions on the cycling of pollutant elements and chemicals in the
ocean indicates that there has been both an increased awareness of the impact of
human activities on the ocean and on its environmental health, as well as on that of
the organisms that reside in the ocean, and on the consumers of ocean seafood,
including and particularly humans. There has been a notable increase in the amount
of study and regulation globally concerning the ocean and many nations are
involved currently in assessments that will lead to further understanding and
regulation. One such activity is the international GEOTRACERS program which
is designed to measure the concentrations and distributions of a whole suite of
metals and isotopes in all the major ocean basins, and this activity in its initial few
years has produced many interesting and exciting results that are mentioned above.
Additionally, there is a continual improvement in the analytical and sampling
technology, and the development of continuous or in situ samplers for a variety
of chemicals that are increasing enormously the amount of information gathered on
chemicals in the ocean. Previously, ship time and resources have been a serious
limiting factor but the developments of these new technologies are increasing the
databases on chemicals in an exponential fashion, and there is now a need to
develop extensive methods to compare and contrast and validate the measurements
being made by a suite of countries around the globe.
The increasing information and knowledge that is occurring in terms of the
elements is not happening to the same extent for the organic chemicals. This is the
result of the fact that current analytical techniques are still time consuming and their
detection limits are often comparable to the concentrations that are being measured.
Further improvement in analytical methods is needed to further enhance the
databases on these chemicals, which is still very sparse, especially for open ocean
and deep waters. Furthermore, there is a constant development of new chemicals by
industry and therefore while the understanding of past chemicals and their fate and
transport are becoming better understood, there is little information on the newer
chemicals released into the environment. To be able to keep pace with such
developments and changes, it is necessary to develop an understanding of the
underlying fundamental processes that control organic chemical fate and transport
and transformation in the ocean. Only with such understanding can reliable
predictions be made; currently most understanding is developed in an empirical
fashion.
Model development and application can provide insights where there is a lack of
data but the models themselves are only as good as their extent of validation, and so
while it is possible to a degree to use models to fill gaps in data and knowledge,
there is also a need to collect data and to constantly improve the models through
validation and comparison with the data. Only through such interactions can our
understanding of the complex cycling of chemicals in the ocean be further
improved.
328 R.P. Mason
Bibliography
Primary Literature
1. Akeredolu FA et al (1994) The flux of anthropogenic trace-metals into the arctic from the
mid-latitudes in 1979/80. Atmos Environ 28:1557–1572
2. Anderson RF (2004) Chemical tracers of particle transport. In: Enderfield H (ed) The oceans
and marine geochemistry, vol 6. Elsevier, Amsterdam, pp 247–273
3. Arimoto R et al (1995) Trace-elements in the atmosphere over the North Atlantic. J Geophys
Res Atmos 100(D1):1199–1213
4. Arimoto R, Duce RA, Ray BJ, Tomza U (2003) Dry deposition of trace elements to the
western North Atlantic. Global Biogeochem Cycles 17(1)
5. Arimoto R, Gao Y, Zhou M-Y, Soo D, Chen L, Gu D, Wang Z, Zhang X (1997) Atmospheric
deposition of trace elements to the western Pacific Basin. In: Baker J (ed) Atmospheric
deposition of contaminants to the great lakes and to coastal waters. SETAC Press, Pensacola,
pp 195–208
6. Baker J, Poster DL, Clark CA, Church TM, Scudlark JR, Ondov JM, Dickhut RM, Cutter
G (1997) Loadings of atmospheric trace elements and organic contaminants to the
Chesapeake Bay. In: Baker J (ed) Atmospheric deposition of contaminants to the great
lakes and coastal waters. SETAC Press, Pensacola, pp 171–194
7. Barbante C et al (2004) Historical record of European emissions of heavy metals to the
atmosphere since the 1650s from alpine snow/ice cores drilled near Monte Rosa. Environ Sci
Technol 38(15):4085–4090
8. Benoit JM, Gilmour CC et al (2003) Geochemical and biological controls over methylmer-
cury production and degradation in aquatic ecosystems. Biogeochem Environ Important
Trace Elem 835:262–297
9. Berg T, Steinnes E (2005) Atmospheric transport of metals. In: Sigel A, Sigel H, Sigel RKO
(eds) Biogeochemistry, availability and transport of metals in the environment, vol 44, Metal
ions in biological systems. Taylor & Francis, Boca Raton
10. Bergan T, Gallardo L, Rodhe H (1999) Mercury in the global troposphere: a three-
dimensional model study. Atmos Environ 33(10):1575–1585
11. Billen G, Garnier J (2007) River basin nutrient delivery to the coastal sea: assessing its
potential to sustain new production of non-siliceous algae. Mar Chem 106(1–2):148–160
12. Boesch DF, Burroughs RH, Baker JE, Mason RP, Rowe CL, Siefert RL (2001) Marine
pollution in the United States. Pew Oceans Commission Report, Washington, DC
13. Boutron C et al (2004) Anthropogenic lead in polar snow and ice archives. C R Geosci
336(10):847–867
14. Boyle EA, Sherrell RM et al (1994) Lead variability in the western North Atlantic Ocean and
central Greenland ice – implications for the search for decadal trends in anthropogenic
emissions. Geochim Cosmochim Acta 58(15):3227–3238
15. Boyle EA (2010) In: Steele JH, Thorpe SA, Turekian KK (eds) Marine chemistry and
geochemistry. Elsevier, Amsterdam, pp 273–280
16. Breivik K, Wania F (2002) Mass budgets, pathways, and equilibrium states of two hexachlor-
ocyclohexanes in the Baltic Sea environment. Environ Sci Technol 36(5):1024–1032
17. Broecker WS, Peng T-H (1982) Tracers in the sea. Eldigio Press, New York
18. Bruland K, Lohan MC (2004) Controls of trace metals in seawater. In: Elderfield H (ed),
Holland HD, Turekian KK (Exec eds) The oceans and marine geochemistry, vol 6.
Amsterdam, Elsevier, pp 23–47
19. Bruland KW, Coale KH (1986) Surface water 234Th/238U disequilibria: spatial and tempo-
ral variations of scavenging rates within the Pacific Ocean. In: Burton JD, Brewer PG,
Chester R (eds) Dynamic processes in the chemistry of the upper ocean. Plenum Press,
New York, pp 159–172
20. Callender E (2005) Heavy metals in the environment – historical trends. In: Lollar B, Holland
HD, Turekian KK (eds) Environmental geochemistry, vol 5, Treatise on geochemistry.
Elsevier, Amsterdam, pp 67–105
21. Chester R (2003) Marine geochemistry. Blackwell, Malden
22. Clegg S, Whitfield M (1990) A generalized model for the scavenging of trace metals in the
open ocean–I. Particle cycling. Deep Sea Res 37:809–832
23. Conley DJ, Schelske CL et al (1993) Modification of the biogeochemical cycle of silica with
eutrophication. Mar Ecol Prog Ser 81:121–128
24. Cossa D, Averty B et al (2009) The origin of methylmercury in open Mediterranean waters.
Limnol Oceanogr 54(3):837–844
25. Cossa D, Cotte-Krief MH et al (2004) Total mercury in the water column near the shelf edge
of the European continental margin. Mar Chem 90(1–4):21–29
26. Cossa D, Martin J-M, Takayanagi K, Sanjuan J (1997) The distribution and cycling of
mercury species in the western Mediterranean. Deep Sea Res 44:721–740
27. Cossa D, Michel P et al (1992) Vertical mercury profile in relation to arsenic, cadmium and
copper at the eastern North Atlantic ICES reference station. Oceanol Acta 15:603–608
28. Cutter GA (2010) Metalloids and oxyanions. In: Steele JH, Thorpe SA, Turekian KK (eds)
Marine chemistry and geochemistry. Elsevier, Amsterdam, pp 64–71
29. Cutter GA, Cutter LS (2006) The biogeochemistry of arsenic and antimony in the North
Pacific Ocean. Geochem Geophys Geosyst 7, doi:2007JD009750/2005GC001159
30. Dachs J, Lohmann R, Ockenden WA, Mejanelle L, Eisenreich SJ, Jones KC (2002) Oceanic
biogeochemical controls on global dynamics of persistent organic pollutants. Environ Sci
Technol 36:4229–4237
31. Donat J, Dryden C (2010) Transition metals and heavy metal speciation. In: Steele JH,
Thorpe SA, Turekian KK (eds) Marine chemistry and geochemistry. Elsevier, Amsterdam,
pp 72–80
32. Donat JR, Bruland KW (1995) Trace elements in the ocean. In: Salbu B, Steinnes E (eds)
Trace metals in natural waters. CRC Press, Boca Raton, pp 247–281
33. Duce RA (2010) In: Steele JH, Thorpe SA, Turekian KK (eds) Marine chemistry and
geochemistry. Elsevier, Amsterdam, pp 281–290
34. Duce R, Liss PS, Merrill JT et al (1991) The atmospheric input of trace species to the world
ocean. Global Biogeochem Cycles 5:193–259
35. Dupont CL, Nelson RK, Bashir S, Moffett JW, Ahner BA (2004) Novel copper-binding and
nitrogen-rich thiols produced and exuded by Emiliania huxleyi. Limnol Oceanogr
49:1754–1762
36. Erel Y, Patterson CC (1994) Leakage of industrial lead into the hydrocycle. Geochim
Cosmochim Acta 58(15):3289–3296
37. Farrington JW (2010) Chlorinated hydrocarbons. In: Steele JH, Thorpe SA, Turekian KK
(eds) Marine chemistry and geochemistry. Elsevier, Amsterdam, pp 163–174
38. Fine RA (2010) CFCs in the ocean. In: Steele JH, Thorpe SA, Turekian KK (eds) Marine
chemistry and geochemistry. Elsevier, Amsterdam, pp 155–162
39. Fitzgerald WF, Lamborg CH, Hammerschmidt CR (2007) Marine biogeochemical cycling of
mercury. Chem Rev 107:641–662
40. Fitzgerald WF, Mason RP (1996) The global mercury cycle: oceanic and anthropogenic
aspects. In: Baeyens W et al (eds) Global and regional mercury cycles: sources, fluxes and
mass balances. Academic, Netherlands, pp 85–108
41. Gabrielli P et al (2005) Variations in atmospheric trace elements in Dome C (East Antarctica)
ice over the last two climatic cycles. Atmos Environ 39(34):6420–6429
42. Gaillardet J, Viers J, Dupre B (2004) Trace elements in river waters. In: Holland H, Turekian
KK (eds) Treatise on geochemistry, vol 5. Elsevier, Amsterdam, pp 225–272
43. German CA, Von Damm KL (2004) Hydrothermal processes. In: Elderfield H (ed) The
oceans and marine geochemistry, vol 6, Treatise on geochemistry. Elsevier, Amsterdam,
pp 181–222
330 R.P. Mason
44. GESAMP, Bowmer T, Kershaw PT (eds) (2010) Proceedings of the GESAMP international
workshop on microplastic particles as a vector in transporting persistent, bioaccumualtive and
toxic substances in the ocean, Paris. GESAMP Report # 82, 68pp
45. GESAMP (2001) A sea of troubles. GESAMP, The Hague. GESAMP Report # 70, 40pp
46. GESAMP (1993) Impact of oil and related chemicals on the marine environment. GESAMP,
London. GESAMP Report # 50, 180pp
47. GESAMP (1990) Review of potential harmful substances – choosing priority
organochlorines for marine hazard assessment. Food and Agriculture Organization of the
United Nations, Rome. GESAMP Report # 42, 24pp
48. GESAMP (1989) The atmospheric input of trace species to the World’s ocean. World
Meteorological Organization, Geneva. GESAMP Report # 38, 124pp
49. Gioia R et al (2008a) Polychlorinated biphenyls in air and water of the North Atlantic and
Arctic Ocean. J Geophys Res 113(D19302). D19302 doi:10.1029/2007JD009750
50. Gioia R et al (2008) Polychlorinated biphenyls (PCBs) in air and seawater of the Atlantic
Ocean: sources, trends and processes. Environ Sci Technol 42:1416–1422
51. Gustafsson O, Gschwend PM, Buesseler KO (1997) Settling removal rates of PCBs into the
northwestern Atlantic derived from 238U–234Th disequilibria. Environ Sci Technol
31:3544–3550
52. Heimburger LE, Cossa D et al (2010) Methyl mercury distributions in relation to the presence
of nano- and picophytoplankton in an oceanic water column (Ligurian Sea, North-western
Mediterranean). Geochim Cosmochim Acta 74(19):5549–5559
53. Hellweger FL (2005) Dynamics of arsenic speciation in surface waters: As(III) production by
algae. Appl Organomet Chem 19(6):727–735
54. Herzig PM, Hannington MD (2000) Input from the deep: hot vents and cold seeps. In: Schultz
H, Zabel M (eds) Marine geochemistry. Springer, Heidelberg, pp 397–416
55. Hong SM, Candelone JP, Soutif M, Boutron CF (1996) A reconstruction of changes in copper
production and copper emissions to the atmosphere during the past 7000 years. Sci Total
Environ 188(2–3):183–193
56. Hopkinson BM, Roe KL et al (2008) Heme uptake by Microscilla marina and evidence for
heme uptake systems in the genomes of diverse marine bacteria. Appl Environ Microbiol
74(20):6263–6270
57. HTAP, Dutchak S, Zuber A (eds) (2010) Hemispheric transport of air pollutants part c:
persistent organic pollutants. ECE, Geneva. Air Pollution Studies #19
58. Iwata H et al (1994) Geographical distribution of persistent organochlorines in air, water and
sediments from Asia and Oceania, and their implications for global redistribution from lower
latitudes. Environ Pollut 85(1):15–33
59. Jonsson A, Gustafsson O, Axelman J, Sunberg H (2003) Global accounting of PCBs in the
continental shelf sediments. Environ Sci Technol 37:245–255
60. Kim G, Alleman LY, Church TM (1999) Atmospheric depositional fluxes of trace elements,
Pb-210, and Be-7 to the Sargasso Sea. Global Biogeochem Cycles 13(4):1183–1192
61. Krishnaswami S (2005) Uranium-thorium series isotopes in ocean profiles. In: Steele JH,
Thorpe SA, Turekian KK (eds) Marine chemistry and geochemistry. Elsevier, Amsterdam,
pp 214–224
62. Landing W, Bruland KH (1987) The contrasting biogeochemistry of iron and manganese in
the Pacific Ocean. Geochim Cosmochim Acta 51:29–43
63. Laurier FJG, Mason RP et al (2004) Mercury distributions in the North Pacific Ocean –
20 years of observations. Mar Chem 90(1–4):3–19
64. Lohmann R, Breivik K, Dachs J, Muir D (2007) Global fate of POPs: current and future
research directions. Environ Pollut 150:150–165
65. Luther GW (2005) Manganese(II) oxidation and Mn(IV) reduction in the environment – two
one-electron transfer steps versus a single two-electron step. Geomicrobiol J 22(3–4):195–203
66. Luther GW (2010) The role of one- and two-electron transfer reactions in forming thermo-
dynamically unstable intermediates as barriers in multi-electron redox reactions. Aquat
Geochem 16(3):395–420
67. Malcolm EG, Schaefer JK et al (2010) Mercury methylation in oxygen deficient zones of the
oceans: no evidence for the predominance of anaerobes. Mar Chem 122:11–19
68. Mason RP, Fitzgerald WF (1993) The distribution and biogeochemical cycling of mercury in
the equatorial Pacific Ocean. Deep Sea Res 40(9):1897–1924
69. Mason RP, Fitzgerald WF (1996) Sources, sinks and biogeochemical cycling of mercury in
the ocean. In: Baeyens W et al (eds) Global and regional mercury cycles: sources, fluxes and
mass balances. Academic, Netherlands, pp 249–272
70. Mason RP, Fitzgerald WF et al (1994) The biogeochemical cycling of elemental mercury:
anthropogenic influences. Geochim Cosmochim Acta 58(15):3191–3198
71. Mason RP, Lawson NM et al (2001) Mercury in the Atlantic Ocean: factors controlling air-
sea exchange of mercury and its distribution in the upper waters. Deep Sea Res II
48:2829–2853
72. Mason RP, O’Donnell J et al (1994) Elemental mercury cycling within the mixed layer of the
equatorial Pacific Ocean. In: Watras CJ, Huckabee JW (eds) Mercury as a global pollutant:
towards integration and synthesis. Lewis, Boca Raton, pp 83–97
73. Mason RP, Sullivan KA (1999) The distribution and speciation of mercury in the south and
equatorial Atlantic. Deep Sea Res Part II 46(5):937–956
74. Mason RP, Sheu GR (2002) Role of the ocean in the global mercury cycle. Global
Biogeochem Cycles 16(4): Art. No. 1093
75. Maybeck M (2004) Global occurence of major elements in rivers. In: Drever JI (ed) Surface
and ground water, weathering and soils, treatise on geochemistry, vol 5. Elesevier,
Amsterdam, pp 207–223
76. Morel F, Milligan AJ, Saito MA (2004) Marine bioinorganic chemistry: the role of trace
metals in the oceanic cycles of major nutrients. In: Elderfield H (ed) The oceans and marine
geochemistry. In: Holland HD, Turekian KK (Exec eds) Treatise on geochemistry, vol 6.
Elsevier/Pergamon, Amsterdam
77. Morel FMM, Hering JG (1993) Principals and applications of aquatic chemistry. Wiley, New
York
78. Murnane R, Sarmiento J et al (1990) Thorium isotopes, particle cycling models, and inverse
calculations of model rate constants. J Geophys Res 95:16,195–16,206
79. Ndung’u K, Thomas MA et al (2001) Silver in the western equatorial and South Atlantic
Ocean. Deep Sea Res Part II 48(13):2933–2945
80. Nice AJ, Lung WS et al (2008) Modeling arsenic in the Patuxent estuary. Environ Sci
Technol 42(13):4804–4810
81. Nozaki Y (1997) A fresh look at element distributions in the North Pacific. http://www.agu.
org/eos_elec/97025e.html
82. Nriagu J (1978) Properties and the biogeochemical cycle of lead. In: Nriagu J (ed) The
biogeochemistry of lead in the environment. Elsevier, Amsterdam, pp 1–14
83. Okbamichael M, Sanudo-Wilhelmy SA (2005) Direct determination of vitamin B-1 in
seawater by solid-phase extraction and high-performance liquid chromatography quantifica-
tion. Limnol Oceanogr Methods 3:241–246
84. Orians KJ, Merrin CL (2010) Refractory metals. In: Steele JH, Thorpe SA, Turekian KK (eds)
85. Pickard GL, Emery WJ (1982) Descriptive physical oceanography. Pergamon Press, Oxford,
249pp
86. Pirrone N, Costa P, Pacyna JM, Ferrara R (2001) Mercury emissions to the atmosphere from
natural and anthropogenic sources in the Mediterranean region. Atmos Environ
35(17):2997–3006
87. Pohl C, Loffler A, Schmidt M, Seifert T (2006) A trace metal (Pb, Cd, Zn, Cu) balance for
surface waters in the eastern Gotland Basin, Baltic Sea. J Mar Syst 60(3–4):381–395
332 R.P. Mason
88. Ranville MA, Cutter GA et al (2010) Aeloian contamination of Se and Ag in the North Pacific
Ocean from Asian fossil fuel combustion. Environ Sci Technol 44:1587–1593
89. Ranville MA, Flegal AR (2005) Silver in the North Pacific Ocean. Geochem Geophys
Geosyst 6: Art. No. Q03M01
90. Ryaboshapko II, Gusev A, Afinogenova O, Berg T, Hjellbrekke A-G (1999) Monitoring and
modeling of lead, cadmium and mercury transboundary transport in the atmosphere of
Europe. EMEP, Meteorological Synthesizing Centre – East, Moscow
91. Sanudo-Wilhelmy SA, Flegal AR (1992) Anthropogenic silver in the southern California
bight – a new tracer of sewage in coastal waters. Environ Sci Technol 26:2147–2151
92. Schaule B, Patterson CC (1983) Perturbations of the natural lead depth profile in the Sargasso
Sea by industrial lead. In: Wong C, Boyle E, Bruland KW, Burton JD, Goldberg ED (eds)
Trace metals in seawater. Plenum Press, New York, pp 487–503
93. Schlitzer R (2010) Inverse modeling of tracers and nutrients. In: Steele JH, Thorpe SA,
Turekian KK (eds) Marine chemistry and geochemistry. Elsevier, Amsterdam, pp 188–199
94. Schwarzenbach RP, Gschwend PM et al (1993) Environmental organic chemistry. Wiley,
New York
95. Scudlark J, Church TM (1997) Atmospheric deposition of trace metals to the mid-Atlantic
bight. In: Baker J (ed) Atmospheric deposition of contaminants to the great lakes and coastal
waters. SETAC Press, Pensacola, pp 195–208
96. Selin NE, Jacob DJ et al (2008) Global 3-D land-ocean–atmosphere model for mercury:
present-day vs. preindustrial cycles and anthropogenic enhancement factors for deposition.
Global Biogeochem Cycles (Accepted)
97. Selin NE, Sunderland EM et al (2010) Sources of mercury exposure for US seafood
consumers: implications for policy. Environ Health Perspect 118(1):137–143
98. Soerensen AL, Sunderland EM et al (2010) An improved global model for air–sea exchange
of mercury: high concentrations over the north Atlantic. Environ Sci Technol 44(22):
8574–8580
99. Stemmler I, Lammel G (2009) Cycling of DDT in the global oceans 1950–2002: world ocean
returns the pollutant. Geophys Res Lett 36(L24602)
100. Stumm W, Morgan JJ (1996) Aquatic chemistry. Wiley, New York
101. Sunderland EM, Krabbenhoft DP et al (2009) Mercury sources, distribution, and bioavail-
ability in the North Pacific Ocean: Insights from data and models. Global Biogeochem Cycles
23
102. Sunderland EM, Mason RP (2007) Human impacts on open ocean mercury concentrations.
Global Biogeochem Cycles 21:GB4022. doi:10.1029/2006GB002876
103. USEPA (1997) The incidence and severity of sediment contamination in surface water of the
United States. Office of Science and Technology, U.S. Environmental Protection Agency,
Washington, DC, (Three volumes) http://water.epa.gov/polwaster/sediments/cs/upload/
nsqszed-complete.pdf
104. Von Damm KL (2010) Hydrothermal vent fluids, chemistry of. In: Steele JH, Thorpe SA,
Turekian KK (eds) Marine chemistry and geochemistry. Elsevier, Amsterdam, pp 81–88
105. Wania F, Mackay D (1995) A global distribution model for persistent organic chemicals. Sci
Total Environ 160/161:211–232
106. Wu JF, Boyle EA (1997) Lead in the western North Atlantic Ocean: completed response to
leaded gasoline phaseout. Geochim Cosmochim Acta 61(15):3279–3283
107. Wu JF, Rember R et al (2010) Isotopic evidence for the source of lead in the North Pacific
abyssal water. Geochim Cosmochim Acta 74(16):4629–4638
108. Zhang Y, Amakawa H et al (2001) Oceanic profiles of dissolved silver: precise measurements
in the basins of the western North Pacific, Sea of Okhotsk and the Japan Sea. Mar Chem
75:151–163
109. Zhang Y, Obata H et al (2004) Silver in the Pacific Ocean and the Bering Sea. Geochem
J 38:623–633
110. Libes S (1992) Introduction to marine biogeochemistry. Elsevier, San Diego, 909 pp
111. Iwata H, Tanabe S, Sakal N, Tatsukawa R (1993) Distribution of persistent organochlorines

in the oceanic air and surface seawater and the role of ocean on their global transport and fate.
Environ Sci Technol 27:1080–1098
112. van der Loeff MMR (2010) Uranium-thorium decay series in the ocean: overview, Chapter
23. In: Steele JH, Thorpe SA, Turekian KK (eds) Marine chemistry and geochemistry.
Elsevier, Amsterdam, pp 203–213
113. Lal D (2010) Cosmogenic isotopes, Chapter 25. In: Steele JH, Thorpe SA, Turekian KK (eds)
114. Key RM (2010) Radiocarbon, Chapter 26. In: Steele JH, Thorpe SA, Turekian KK (eds)
115. Rabalais NN (2010) Hypoxia, Chapter 34. In: Steele JH, Thorpe SA, Turekian KK (eds)
116. Sunda WG (2010) Trace metal nutrients, Chapter 4. In: Steele JH, Thorpe SA, Turekian KK
(eds) Marine chemistry and geochemistry. Elsevier, Amsterdam, pp 17–28
117. Balcom PH et al (2004) Mercury sources and cycling in the Connecticut River and Long
Island Sound. Marine Chem 90(1–4):53–74
118. Gill GA, Fitzgerald WF (1988) Vertical mercury distributions in the oceans. Geochim
Cosmochim Acta 52:1719–1728
119. Cossa D et al (2011) Mercury in the Southern Ocean. Geochim Cosmochim Acta 75
(14):4037–4052
120. Hollweg T, Gilmour C, Mason R (2009) Factors controlling the methylation of mercury in
sediments of the Chesapeake Bay and mid-Atlantic continental shelf. Marine Chem
121. Mason RP (2005) Air-sea exchange and marine boundary layer atmospheric transformations
of mercury and their importance in the global mercury cycle. In: Pirrone N, Mahaffey KR
(eds) Dynamics of mercury pollution on regional and global scales. Springer, New York, pp
213–239
122. Mason RP, Fitzgerald WF (1990) Mercury speciation in the equatorial Pacific Ocean. EOS 71
(43):1413
123. Cutter GA (1993) Metalloids in wet deposition on Bermuda – concentrations, sources and
fluxes. JGR-Atmos 98(D9):16777–16786
Chapter 13
Subsurface Fate and Transport of Chemicals
Frank T. Barranco Jr., Samantha L. Saalfield,

Frederick J. Tenbus, and Brian P. Shedd
Glossary
Abiotic Not relating to life, as in abiotic chemical reactions

that occur independent of living organisms.
Absorption Retention of a chemical within a solid material.
Adsorption Adhesion of a chemical to the surface of a solid.
Advection Transport of a solute within a fluid in the direction of
the bulk fluid’s flow.
Aerobic Requiring oxygen.
Air sparging Injecting air or oxygen into an aquifer.
Aliphatic Organic compounds not containing an aromatic ring.
Anaerobic Without oxygen.
Best management Techniques generally accepted as effective for
practice (BMP) achieving a particular goal, for example minimizing
the environmental impact of remediation.
Biodegradation Use of living organisms to clean up contaminated
environmental media.
Biotic Relating to life, as in a biotic reaction mediated by
living organisms.
Bioventing The addition of air (or oxygen) under, at times, an
induced lowering of water table to promote aerobic
biodegradation of subsurface contaminants in the
unsaturated zone.
F.T. Barranco Jr. (*) • S.L. Saalfield • F.J. Tenbus • B.P. Shedd
EA Engineering, Science, and Technology, Inc., 15 Loveton Circle, Sparks, MD 21152, USA
e-mail: [email protected]; [email protected]; [email protected]; [email protected]
336 F.T. Barranco Jr. et al.
Catalytic oxidizer Remediation technology equipment that uses

a catalyst to accelerate the chemical oxidation of
hydrocarbons with oxygen in a vapor effluent stream.
Chemisorption Adhesion of a chemical to the surface of a solid,
specifically through a chemical reaction occurring at
the surface.
Confined aquifer A water-bearing geologic strata that is situated
between impermeable layers (clays and silt layers),
leading to higher pressure of the groundwater in this
unit.
Extraction well A well used to remove liquid or gas from the
subsurface.
Fate and transport Encompasses how contaminants move through envi-
ronmental media and how long the contaminants per-
sist or how fast they are degraded.
Feasibility study A document that describes and analyzes potential
cleanup alternatives for a site and recommends selec-
tion of an effective and efficient alternative.
Green and sustainable Environmental cleanup that is designed and
remediation performed with consideration of the environmental
impacts of the technologies used.
Half-life Time required for half of the molecules of a chemical
to decay or be degraded.
Henry’s law Law that is used to describe the volatility of
a chemical, by describing the equilibrium between
the vapor phase and dissolved forms of the chemical.
Hydrocarbons Chemical compounds that consist of carbon and
hydrogen.
Hydrodynamic dispersion Transport within a fluid in directions other than the
primary direction of fluid flow. This process decreases
contaminant concentrations while increasing the total
volume of fluid contaminated.
Hydrodynamics The process of the motion of groundwater.
Hydrogeology Discipline dealing with the properties and
characteristics of groundwater.
Hydrolysis Reaction that splits a chemical into two parts by
adding a water molecule, through addition of
a hydrogen ion to one fragment of the chemical and
addition of a hydroxyl group to the other fragment.
Hydrophobic Having an aversion to water. Typically describes
a contaminant that associates with nonpolar
substances (such as oils and organic matter) rather
than polar substances like water.
Injection well A well used for injection of fluids, gases, and/or
chemicals for remediation.
13 Subsurface Fate and Transport of Chemicals 337
Inorganic Describes chemicals that are not organic, including

metals and common anions (sulfate, nitrate, etc.).
Interfacial tension Tension at the interface between a liquid of one
chemical and a solid, liquid, or gas of another chemi-
cal. One of the primary determinants of NAPL mobil-
ity in the subsurface.
Interim remedial action A remedial action taken to address immediate risks to
human health or the environment before long-term
remedial goals are achieved.
Leaching Dissolution of relatively soluble chemicals and
removal by water transport.
Life cycle analysis (LCA) Evaluation of the environmental impacts of all stages
of a product or process.
Light nonaqueous phase A nonaqueous phase liquid that is less dense than
liquid (LNAPL) water and therefore floats on the water table, includ-
ing petroleum hydrocarbon fuels and lubricating oils.
Liquid density Mass per unit volume of a liquid.
Liquid viscosity Resistance of a liquid to being deformed. Higher
viscosity is associated with more resistance to flow,
or less fluidity.
Lower explosive limit The concentration of a compound in air below which
it will not ignite.
Microaerophilic Requiring only small amounts of oxygen.
Monoaromatic Organic chemicals containing one aromatic ring,
hydrocarbons (MAHs) which are common petroleum derivatives.
Nonaqueous phase A liquid, such as oil, that remains in a separate phase
liquid (NAPL) in the groundwater and can act as a source of organic
contaminants to groundwater and soil.
Organic Describes a category of chemicals that typically con-
tain carbon with hydrogen, oxygen, and/or nitrogen.
Most organic compounds can be degraded to carbon
dioxide, water, and other simple components.
Oxidation Chemical reaction in which a chemical of interest
loses electrons. The chemical that takes the electrons
is known as the oxidant. Includes “rusting” of metals
and processes that degrade organic matter to carbon
dioxide.
Partitioning Distribution of a chemical between the solid, fluid,
and/or gas phases, in proportions reflecting its affinity
for each phase, as described by the partition
coefficient.
Permeability Tendency of a material to allow fluids to flow
through it.
Persistent organic Chemicals that do not readily degrade under environ-

pollutants (POPs) mental conditions and, therefore, persist in environ-
mental media.
Polychlorinated Organic chemicals with chlorine atoms attached to
biphenyls (PCBs) two benzene (aromatic) rings, which were widely
used as dielectric and coolant fluids, for example in
transformers.
Polycyclic aromatic Organic chemicals containing more than one aromatic
hydrocarbons (PAHs) ring, which are common by-products of coal
combustion.
Porosity Fraction of a material that is void space. Can be
primary (original, from when the geological material
was formed) or secondary (formed later, by selective
dissolution or fracturing).
Precipitation Formation of a solid from dissolved chemicals in
a solution.
Preferential flow Faster movement of groundwater through certain,
more porous or permeable, portions of the subsurface,
which can result in localized, rapid contaminant
transport.
Redox Term used to describe the related processes of reduc-
tion and oxidation.
Reduction Chemical reaction in which electrons are gained by
a chemical of interest. The chemical donating the
electrons is known as the reductant. Includes the
reduction of oxygen gas during aerobic respiration
and the reduction of other chemicals (nitrate, iron,
carbon dioxide) during anaerobic respiration.
Remedial action Action taken to remove or contain a hazardous sub-
stance in the environment.
Remediation Cleanup or other methods used to remove or contain
hazardous materials.
Risk assessment Qualitative and quantitative evaluation of the risk
posed to human health and/or the environment by
contaminants.
Saturated zone The portion of the subsurface below the water table,
where the pressure of water within the pores is at
a pressure equal to or greater than atmospheric
pressure.
Solubility Ability of a chemical to dissolve into (i.e., mix with
and become incorporated into) another substance.
Unless otherwise specified, in an environmental con-
text, solubility is typically used to refer to solubility of
a chemical in water.
Sorption (verb: to sorb) Attachment of a chemical to a solid, which removes

the chemical from the dissolved phase. See also
adsorption, absorption, and chemisorptions.
Speciation The chemical form (phase, redox state, molecular
structure) in which an element exists. Important deter-
minant of metal mobility in the environment.
Subsurface The zone beneath the surface of the earth, including
geologic strata and groundwater.
Transport mechanisms Processes by which contaminants move through the
environment.
Unsaturated/vadose zone The subsurface zone between land surface and the
water table where the moisture content is less than
atmospheric pressure (i.e., soil pores are not
completely filled with water).
Vapor pressure Pressure of the vapor of a chemical that exists in
equilibrium with the chemical’s solid or liquid phase.
Volatility Tendency of a chemical to vaporize, or go into the
gaseous phase.
Since the onset of subsurface remediation in the 1970s, there has been a need for
a more appropriate balance between the protectiveness of environmental cleanup
technologies and the concept of environmental sustainability. This entry explores
the implementation of innovative green and sustainable practices deemed appropri-
ate for the remedial technologies that address the most common classes of persistent
and toxic subsurface contaminants.
Introduction
Preceding and during the early timeframe of environmental remediation, there was
a lack of concern and knowledge about the fate and persistence of chemicals
released to the subsurface. There was a prevailing but unfounded assumption that
the subsurface environment would sorb or attenuate almost unlimited amounts of
contaminants. Much to our dismay, this has been shown to be false through major
advancements in analytical chemistry techniques over the last 2 decades. Once
subsurface transport mechanisms were understood, it became obvious that
contaminants released at or near the surface may make their way deep into
subsurface environments, including confined aquifers and bedrock settings. The
fate and transport of environmental contaminants in subsurface environments are
Table 13.1 Summary of properties that affect fate and transport of organic and inorganic
contaminants in the subsurface
Contaminant properties
Organic contaminants Inorganic contaminants
Solubility Solubility
Fluid density Redox (reduction-oxidation)
Viscosity Speciation
Interfacial tension Adsorption
Carbon partition coefficient Reactivity
Henry’s law constant Vapor pressure
Biological degradation
Vapor pressure
Subsurface/hydrogeologic properties
Rock type and characteristics
Hydraulic conductivity
Aquifer redox chemistry
Specific surface area of minerals
significantly affected by two categories of properties (Table 13.1): (1) the geologic
characteristics of the subsurface environment, which defines the intrinsic properties
of the soil (or rock) and imparts the characteristics of groundwater flow through that
media, and (2) the properties of the contaminants, which define the physicochemi-
cal and biological processes that affect their fate and persistence.
The prevailing group of processes affecting subsurface fate and contaminant
transport are hydrodynamic processes, partitioning, biotic reactions, and abiotic
reactions. Hydrodynamic processes impact contaminant transport through
groundwater advection, hydrodynamic dispersion, and potential preferential
flow. Partitioning affects contaminant distribution and dispersal by allowing
interchange of contaminant from one subsurface medium to another (e.g., soil,
groundwater, soil gas) through means of adsorption, absorption, or chemisorp-
tion. Biotic reactions can affect contaminant transport by degradation (or
immobilization) of the contaminant in oxidation or reduction reactions. More
specifically, biotic processes, which occur under aerobic, microaerophilic, or
anaerobic conditions, can lead to degradation or immobilization (by reaction or
precipitation), depending on the type of contaminant. Abiotic reactions affect
contaminant transport in the subsurface by promoting interactions between the
contaminant and groundwater or stationary media (e.g., soil, bedrock), causing
the contaminant to degrade or change in form (e.g., hydrolysis, redox reactions).
These processes take place within the saturated zone and the unsaturated zone of
the subsurface. Of these two subsurface zones, heightened concern is often paid to
the saturated zone because contaminants are in direct contact with groundwater,
which may be used as a potable or irrigational resource or for other purposes. The
unsaturated zone, or vadose zone, overlies the saturated zone (i.e., above the water
table) and is an important contributor to contaminant fate and transport through
processes such as leaching and migration to the saturated groundwater zone.
Sources of chemicals released to the subsurface are varied, but generally include
(1) underground and above ground storage tanks, (2) septic tanks, (3) agricultural
activities, (4) municipal and industrial landfills and dumps, (5) regulated and
abandoned hazardous waste sites, (6) injection wells, and (7) other industrial sites
[1]. The types of chemicals that have been historically released to the subsurface
and have the potential to cause adverse effects to human health or the environment
are generally divided into organic compounds and inorganic compounds (or inor-
ganic elements). Several of the more common classes of organic and inorganic
contaminants found in soil and groundwater are as follows:
• Petroleum hydrocarbons and derivatives used as fuels such as gasoline, diesel
fuel, jet fuel, and heating oil. These fuels consist of many organic chemical
components, including monoaromatic hydrocarbons (MAHs) like benzene, tol-
uene, ethylbenzene, and xylene (BTEX), and low-molecular-weight polycyclic
aromatic hydrocarbons (PAHs), such as naphthalene, and are generally less
dense in hydrocarbon fluid form than water, contain components that are highly
volatile, and are sparingly to moderately soluble in water. When present as an
immiscible phase (in water), they tend to persist over reasonably long time
frames as light non-aqueous phase liquids (LNAPLs). However, LNAPL
constituents will dissolve in water, sorb onto soils, and/or partition into the
vapor phase.
• Chlorinated aliphatic hydrocarbons used as solvents, degreasers, and dry
cleaning fluids, such as tetrachloroethylene (PCE), trichloroethylene (TCE),
and carbon tetrachloride (CT). These compounds generally are distilled for use
in industrial settings in a pure form or as relatively simple mixtures. When
present as an immiscible phase, they generally are denser than water, highly
volatile, and also sparingly soluble in water. Hence, they can exist as dense
non-aqueous phase liquids (DNAPLs), dissolved in water, sorbed onto soils,
and/or present in the vapor phase.
• Polychlorinated biphenyls (PCBs), pesticides, wood-preserving chemicals (such
as creosote), and fossil fuel combustion/gas manufacturing gas by-products,
such as coal tars and high-molecular-weight PAHs. These contaminants tend
to be dense, highly viscous fluids (when present an immiscible phase) with
constituents that sorb strongly onto soils, are nearly insoluble in water, exhibit
low volatility, and are persistent in the subsurface environment because they
chemically and biologically degrade very slowly. When present as an immisci-
ble phase, they generally are found as DNAPL or are sorbed onto soils.
• Explosives and energetic compounds such as solid-rocket fuels and propellants,
including trinitrotoluene (TNT), plastic explosives, perchlorate, and munitions
components. These constituents are often found as solids on or near the land
surface, although some (such as perchlorate) can be found dissolved in
groundwater.
• Metals such as arsenic, lead, chromium (VI), mercury, cadmium, and others.
These generally are found either dissolved in groundwater, present as elements
in the solid phase, or present within the rock matrices as mineral components.
Each type of contaminant has a distinct set of physicochemical characteristics

that define its behavior and migration within subsurface environments. Hydrogeol-
ogy and rock/mineral geochemistry also have a significant influence on defining the
fate and migration of contaminants. As a function of the contaminant’s physico-
chemical properties and geologic characteristics, the above listed classes of organic
chemicals exist in the subsurface as one or more of four phases: (1) mobile or
residually entrapped nonaqueous phase liquid (NAPL), (2) dissolved phase in
groundwater, (3) sorbed phase to solid aquifer media, and (4) vapor phase in soil
gas. Both the properties of the chemical and that of the subsurface control the
dynamic evolution of phase transfer, including the duration of time that these
organic chemical remains within these phases following a spill or release. From
the point of release, organic chemicals generally exist in the NAPL phase, with
eventual partitioning to one or more of the other phases with time. Subsurface
NAPLs can exist as a pure chemical or as a bulk mixture of chemicals.
Physical and chemical properties that have a major effect on the fate, transport,
and persistence of the classes of typical organic contaminants are shown in
Table 13.2. Properties such as solubility (in water) determine the degree to which
a contaminant persists in the subsurface as an immiscible fluid or solid. If an
immiscible fluid phase persists in the subsurface, fluid density governs whether
the fluid acts as an LNAPL or DNAPL (e.g., petroleum hydrocarbons, with a fluid
density less than 1 g/mL, form LNAPLs, whereas chlorinated hydrocarbons form
DNAPLs). Properties like organic carbon partition coefficient (Koc) and Henry’s
law constant effect the tendency of a chemical to partition under equilibrium
conditions from a source release (e.g., contaminant liquid or solid) to subsurface
media (e.g., soil, groundwater, and/or soil gas). The relatively high Henry’s law
constants and solubilities of hydrocarbons, for example, indicate that hydrocarbons
are more likely to partition into air and groundwater than are PAHs, PCBs, and
pesticides.
The half-lives for biological degradation of the contaminants, also shown in
Table 13.2, provide an indication (using first order kinetics) of how quickly
a chemical will biologically degrade in the subsurface. Petroleum hydrocarbons
and chlorinated hydrocarbons have been observed to degrade under aerobic and
anaerobic conditions, respectively, at reaction rates determined to be moderate to
fast for both natural and engineered remedial systems. Conversely, PCBs and
certain PAHs have been shown to undergo very slow to negligible rates of aerobic
and/or anaerobic biodegradation in natural subsurface settings. Explosives and
energetic compounds have been shown on a constituent by constituent basis to
biologically or chemically degrade (in presence of reductants and oxidants, respec-
tively) under subsurface conditions attainable with the aid of remedial technologies.
The most common inorganic compounds identified in the subsurface include
metal contaminants. Although metals are natural constituents of soils, anthropo-
genic metals enter the soil through a variety of means including (1) leaching of
municipal or industrial solid wastes, (2) storm water runoff and infiltration,
(3) industrial by-products, (4) dredged materials, (5) mining and smelting
operations, (6) atmospheric emissions from coal or oil combustion, (7) ash and
13
Table 13.2 Physicochemical properties of organic contaminants that affect contaminant fate, transport, and persistence (Refs. [2–5])
Henry’s law constant Fluid density Biodegradation half-life

Class of contaminant (atm. m3/mol) Solubility (g/L) (g/cm3) log Koc Aerobic Anaerobic
Chlorinated hydrocarbons 0.00674–0.015 0.15–6.3 1.2–1.6 0.4–2.4 4 weeks–1 year 11 weeks–4.5 years
Petroleum hydrocarbons 0.00046–0.015 0.03–200 0.86–0.88 1.5–3.6 2 days–6 months 8 days–2 years
Heavy PAHs Minimal–0.0003 0.0000003–0.004 NA (solid) 1.3–7.5 12 days–5.2 years 50 days–21 years
Subsurface Fate and Transport of Chemicals
PCBs 0.000003–750 0.00001–0.002 1.182–1.566 2.4–6.4 1.5–9 years –

Pesticides 0.0000002–0.002 0.0000004–0.5 NA (solid) 2.1–6.3 2 days–16 years 1–294 days
Explosives/energetics Minimal–0.00000002 0.13–250 NA (solid) NA 4 weeks–6 months (TNT) 4 weeks–6 months (TNT)
NA not applicable
343
slag from coal or oil combustion, (8) and sludge residues from wastewater treat-
ment. Typical metals identified as contaminants in subsurface environments include
arsenic, chromium, cadmium, copper, lead, mercury, nickel, silver, and
zinc. Radionuclides are a separate class of contaminants that are inorganic and
due to their specificity are not discussed in this entry.
The physicochemical properties of inorganic chemicals that govern their fate,
partitioning, and migration within subsurface environments include solubility,
reduction–oxidation (redox) speciation, reactivity, and vapor pressure. Properties
of subsurface media that play a defining role in the fate of inorganic contaminants
within subsurface environments include rock type and characteristics (primary and
secondary porosity, mineral composition, fracture density), hydraulic conductivity,
redox chemistry, and specific surface area of minerals present. Metals are not
degraded by biological or chemical reactions, though they can conveniently be
rendered unavailable through precipitation reactions or transformed via oxidation/
reduction reactions to less toxic species.
Properties of subsurface media that play a defining role in the fate, partitioning,
and transport of organic contaminants within subsurface environments include rock
type (i.e., porous or fracture flow), primary and secondary porosity, hydraulic
conductivity, and total organic carbon.
Background
Beginning in the 1970s with the advent of subsurface remediation, the environmen-
tal industry invested heavily in remediation systems without an adequate under-
standing of the degree of contaminant cleanup (or the duration of the cleanup).
Although these early efforts were well intended, the actions taken with
the technologies utilized were not justified based on the costs of the cleanups, the
overutilization of resources, the intensive amount of energy consumed, and
the insufficient contaminant removal. Past presumptive remedies such as soil
“excavation and off-site disposal” and groundwater “pump and treat” are prime
examples of these generally wasteful remedies considered unsustainable onto the
future.
Beyond this early misunderstanding about remedial technology selection to
achieve efficient and sustained removal of subsurface contaminants, there has
been a long-standing misperception that cleanup shall continue to pristine
conditions, a goal that as turns out is largely unattainable in most circumstances.
At the expense of nearly 2 decades of numerous examples of unsuccessful and
costly cleanups, we have learned considerably from these early mistakes. Today, in
many cases, there is a robust process and a wealth of regulatory guidance to
determine more pragmatic yet adequate remedial action objectives and cleanup
performance criteria. These cleanup goals and objectives are most often defined by
human health risk assessment [6], ecological risk assessment [7], risk-based
Fig. 13.1 Key elements of

green and sustainable
remediation
corrective action [8, 9], or by what is reasonably achievable based on the best
practicable remedial technology options (generally referred to as best available
technology [BAT]).
In addition to these well-considered approaches that have provided a practicable
risk-related basis for cleanup end goals, there is growing interest in the last 5 years
to incorporate green and sustainable remediation (GSR) concepts throughout the
remedial action process, while continuing to provide acceptable long-term protec-
tion of human health and the environment. As suggested by the Sustainable
Remediation Forum (SURF), sustainable remediation is defined as a “remedy or
combination of remedies whose net benefit on human health and the environment is
maximized through the judicious use of limited resources” [10]. To this end,
sustainable remediation employs solutions that minimize the environmental foot-
print while providing maximum net environmental benefit over the remedial
lifecycle. To realize the benefits of sustainable remediation requires the use of
green and/or renewable energy sources, conservation of water and energy, decreas-
ing waste, and formulating integrated sustainability policies. With sustainable
remediation, the goal should be to (1) develop and implement safe remedial
solutions that are minimally disruptive to the environment, (2) realize energy
savings through creative design and value engineering, (3) embrace waste minimi-
zation and recycling concepts, and (4) emit the least amount of pollutants and
greenhouse gases to the atmosphere. To the extent practical, the follow-
ing sustainability elements should be applied to remedial solutions incorporating
GSR practices (Fig. 13.1): (1) short-and long-term energy and water consumption,
(2) greenhouse gas emissions and air pollutants, (3) ecosystem impacts, (4) material
consumption, and (5) waste minimization and/or recycling.
The recent recognition of the balance between adequacy of cleanup and
sustainability concepts has not, in some cases, dispelled the notion of initiating
cleanups with little thought to the protective character (human health and environ-
mental), resource utilization, or safety of the action. As evidenced by the continued
use of pump and treat systems, there are still regulatory mandates for cleanup with
little decrease in contaminant mass (and therefore little decrease in overall risk).
There are, in fact, examples of sites where natural attenuation is actively decreasing
contaminant concentrations in the subsurface, yet the overseeing regulatory
agencies require a more traditional, less sustainably oriented, significantly more
costly approach because of outdated or arbitrary remediation goals set for the site.
Clearly a new paradigm is warranted, whereby cleanup protectiveness of a remedy
is balanced with sustainability elements. The analysis to establish this balance
should be evaluated early in remedial planning, such as during a feasibility study
of remedial alternatives. Incorporation of sustainability into the balancing criteria
evaluated during a feasibility study would help to assure that this process happens.
In the interim, many professionals in the environmental industry have been
documenting sustainability metrics by qualitatively and/or quantitatively scoring
the degree of sustainability core elements to be utilized on cleanup actions. Such
evaluations have ranged from simple qualitative review of the available list of best
management practices (BMPs), utilizing those that are applicable, to the perfor-
mance of quantitative and complex life cycle analysis (LCAs) for optimization of
the cleanup over the remedial lifetime. The following section provides valuable
information on industry lessons learned from implemented LCAs and BMPs for
sustainable remediation practices, as applied to the typical classes of subsurface
contaminants described in this document.
Sustainable Remediation Practices for Classes of Typical

Subsurface Contaminants
Remediation sites comprise a range of sizes, proximity to human and/or ecological

receptors, proximity to man-made infrastructure, site accessibility, environmental
complexity, type of contaminants and their chemical, physical, and toxicity
characteristics, complexity of the circumstances surrounding the release(s), and
so on. All of these factors affect the feasibility of corrective actions, and even the
ability to implement any corrective action. Because of this, no generalized discus-
sion of sustainable remediation practices can be complete, because the topic is
simply too broad. However, the information described below, is comprised of
remedial measures applied recently that can be characterized as sustainable
practices.
With this in mind, this chapter will focus on sustainable practices honed over
time for the general classes of subsurface contaminants discussed above. The
discussions are premised on the following assumptions:
• Releases to the subsurface have stopped; i.e., that the pipelines have been
repaired, the underground storage tanks have been removed, the uncontrolled
landfill is no longer receiving toxic materials and has been capped, etc.
• Any discussion of “sources,” with regard to groundwater, refers to concentrated

areas of contaminants such as NAPL or highly contaminated soils that are
present but relatively stable.
• A risk assessment has been completed (if necessary), and remedial action
objectives have been defined before remedial implementation.
• Emergency or interim measures (e.g., protection or replacement of domestic
water supplies affected by the release) are in place as needed.
• Site characterization has progressed to a point where remediation methods can
be considered within a feasibility study (FS) or a focused feasibility study (FFS).
• Remedial systems are designed to destroy (or remove) contaminants with the
knowledge and benefit of exploiting the physical and/or chemical characteristics
of the contaminants.
• Sustainable measures for the contaminants discussed have been optimized over
time as a result of trial and error evaluation with full-scale remedial systems
and technology innovation or breakthroughs.
These discussions readily address the lessons learned and resulting sustainable
optimizations (or BMPs) of various technologies for typical contaminants.
Petroleum Hydrocarbons
Petroleum hydrocarbons, including MAHs and low-molecular-weight PAHs, often

occur in persistent LNAPLs in the subsurface. Therefore, the primary method of
remediation historically has relied on product recovery within a cone of depression
produced by water table drawdown from groundwater pumping.
As a result of energy inefficiencies and high cost of groundwater extraction (and
treatment), a more sustainable approach evolved involving product recovery with
skimmer pumps. Skimmer pumps are designed to remove LNAPL from the water
table surface. The skimmer floats on the water table and has an interval with
a hydrophobic (water-rejecting) screen that is open to the LNAPL layer within
a monitoring well. LNAPL is drawn into the skimmer and flows through a flexible
tube to a reservoir where it is pumped to the surface. The following practices are
often employed with product recovery systems to reduce energy consumption,
minimize site impacts, and improve the overall sustainability of the treatment [11]:
• Power product recovery components or auxiliary equipment with low energy
demand, such as renewable energy off-grid wind turbines or photovoltaic (PV)
systems.
• Such systems relying on off-grid energy should be equipped with deep-cycle
batteries to provide steady power.
• Eliminate the long-distance transport of incoming materials and equipment or
outgoing remedial-derived wastes (i.e., recovered product). To that end, consol-
idate deliveries/pickups to avoid deploying partially filled vehicles.
• Recycle separated product (LNAPL) through local fuel or waste recyclers.

• Optimize product recovery through proper equipment sizing and frequent reas-
sessment based on treatment performance.
• Establish operating or cleanup performance criteria that could trigger use of less
intensive polishing technologies as cleanup progresses and LNAPL recovery
rates decline.
Skimmer pumps are capable of removing LNAPL down to a sheen, but do not
reduce dissolved contaminant concentrations or the mass of contaminants sorbed
onto soil. In addition, skimmer pumps, unless supplemented with vacuum
enhancements, generally are not capable of providing complete capture of
a mobile accumulation of free-phase LNAPL that is migrating under natural
subsurface conditions. In such cases, groundwater extraction is sparingly used to
capture and contain the mobile, free-phase LNAPL for recovery. The following
sustainable practices are often employed to minimize groundwater extraction while
optimizing LNAPL capture:
• Perform groundwater capture zone analyses using empirical calculations or
numerical groundwater modeling as basis of design for groundwater capture
and, therefore, LNAPL containment.
• Calibrate and refine groundwater extraction network and flow rates with
calculations or modeling after system startup with observed drawdown
conditions from corrected groundwater elevations.
• Monitor and periodically optimize the groundwater extraction network and flow
rates to maximize LNAPL recovery rates as a result of higher producing wells.
• Monitor extraction well change in head at a given rate over time, or specific
capacity, to ensure continued efficient well operation. Rehabilitate wells period-
ically, if decreased specific capacity is observed, to maximize extraction well
longevity.
• Design system to minimize the total amount of piping including length, surface
area, bends, and elbows to maximize transport efficiency.
• Engineer extraction wells to maximize efficiency through decreased head loss
and optimized flow rates by designing the well gravel pack to match the
formation and using the largest feasible slot size to maximize open cross-
sectional area. This decreases velocity and prevents migration of fines by
allowing a less-turbulent flow into the well.
Sustainable features of skimmer pumps are that they require little energy to
operate, and the energy can be obtained from sustainable sources such as solar
panels (Fig. 13.2). Recovered LNAPL can be recycled to minimize waste. The
method is relatively slow and inefficient for LNAPL removal, but it can be effective
at small sites. Skimmer systems can be enhanced by applying a vacuum, which can
speed up LNAPL recovery for a relatively minor additional energy input that could
also be obtained from sustainable power sources. The use of skimmer pumps
Fig. 13.2 Example of solar powered system with photovoltaic panels for powering product
skimming
generally does not exacerbate existing subsurface conditions or make additional

remediation efforts more difficult or less effective.
Soil vapor extraction (SVE) can be effective at removal of volatile organic
compounds (VOCs) of petroleum hydrocarbons within the unsaturated zone of
the subsurface present as LNAPL, sorbed to contaminated soil, or in the vapor
phase. One or more SVE wells (screened in the unsaturated zone above the water
table) are constructed, and a vacuum system is installed and manifolded to the
wells. The vacuum draws the soil air (which is contaminated with VOCs) out of the
subsurface through a treatment system (described below). The contaminated vapors
are replaced by fresh air from vent wells or from other parts of the subsurface. This
allows the volatile fraction of the VOC to evaporate at a rate that is proportional to
the Henry’s law coefficient. These volatiles are then carried by the air into the SVE
system, allowing for further evaporation into the fresher soil air, and so on.
SVE systems are efficient for VOC removal in the unsaturated zone primarily
because volatiles preferentially partition into the vapor phase, and air is relatively
easy to move in the subsurface compared to liquids. SVE systems tend to lose
efficiency with age as the volatile fractions of the contaminants are removed and the
less volatile fractions remain. Because not all of the components of refined petro-
leum hydrocarbons are volatile, SVE by itself cannot achieve complete subsurface
remediation. In addition, SVE does not affect contaminants below the water table,
because they are not in contact with the soil air. However, certain adaptations can
be made to an SVE system to facilitate contaminant removal from the subsurface.
Examples of these adaptations include artificially lowering the water table to

expose more soil contaminants to the air or forcing air into the aquifer below the
water table to strip out the dissolved or sorbed volatile compounds in the saturated
zone through a process known as air sparging.
Nearly all SVE systems require treatment of the vapors removed from the
subsurface soil. The most efficient treatment depends on the contaminant
concentrations within the vapor stream. Essentially, low concentrations are most
cost effectively removed using a sorption medium such as granular activated carbon
(GAC), which is a specific type of charcoal. GAC requires no energy input to
remove organic compounds from soil vapors while deployed on an SVE system.
Because GAC is an absorbent material, however, it does not destroy contaminants,
but transfers them to another medium. Once most of the sorption sites are filled and
“breakthrough” (i.e., contaminants are no longer captured by the absorbent mate-
rial) occurs, GAC must be regenerated off-site (requiring energy) or replaced with
virgin material (also requiring energy). The virgin material commonly used to
manufacture GAC is coconut shells, which are a sustainable resource.
If vapor concentrations are too high, breakthrough will occur quickly and GAC
change-out costs can become prohibitive. Above certain concentration threshold,
however, a treatment method that uses a catalytic oxidizer may become the most
appropriate choice. Catalytic oxidizers are energy intensive because their optimal
performance occurs at a relatively high temperature (300–600 C). However, the
oxidation of fuel components at sufficient concentrations releases heat that is used
to maintain the optimal temperature, greatly increasing efficiency and decreasing or
eliminating the need for external energy inputs. As a result, the sustainability of the
remediation system is enhanced.
For very high concentrations of VOCs that approach or exceed the lower
explosive limit of the vapors, high temperature oxidation of the fuel components
can be used to actually run the SVE system in whole or in part. The technology uses
an internal combustion engine powered by the extracted vapors and an auxiliary
fuel source such as propane if needed. Under the right conditions, such a system
requires no external power supply, and if a generator module is added, the system
can supply power that can be used to operate lights or other electrical equipment
onsite. Off-gas emissions from these units are equivalent to those from operating an
automobile engine.
If properly used, SVE does not generally exacerbate existing subsurface
conditions or make additional remediation efforts more difficult or less effective.
If improperly vented, however, the negative pressure generated by the SVE can
cause mounding of the water table that creates a hydraulic gradient capable of
spreading LNAPL away from the vapor extraction well(s), possibly increasing the
lateral extent of the contaminated area.
Most SVE systems have esthetic impacts that reduce their green characteristics.
One undesirable aspect is that they tend to be noisy. This can be mitigated by
methods such as surrounding the SVE unit with fences lined with commercially
available sound-insulating blankets. Another undesirable aspect is that the systems
have a visual impact. They typically are not large, with a footprint on the order of
10 m2 or less and a height less than 3 m, but they have an “industrial” look about
them. This could be mitigated with appropriate landscaping that could include
components like fencing with small trees, shrubs, or even large potted plants.
Typically, SVE systems are operated almost constantly for a period of approximately
2–3 years, so the esthetic impacts can be a substantial nuisance to nearby workers,
passers-by, or residents. The following BMPs are examples of ways to promote
a more sustainable SVE approach:
• Optimize extraction configuration and rates by manifolding several vapor points
and periodically manipulating valves to minimize “dead zones” and reduce
remediation timeframe, energy consumption, and noise.
• Utilize appropriate vapor treatment with GAC by vapor concentration (low
concentrations – energy efficient; low concentrations result in infrequent change
out; cat-ox for intermediate to high concentrations – appropriate concentrations
improve oxidation efficiency; thermal oxidation for very high concentrations –
can utilize contaminant vapors as a fuel source).
Multiphase extraction combines a pump system to remove LNAPL and
contaminated groundwater along with SVE to remove soil vapors. The pump
system has the effect of lowering the water table, creating a gradient for the
LNAPL to migrate toward the extraction well while it removes contaminated
water and LNAPL. This effect also exposes more of the subsurface to the soil air,
enabling SVE to act on a larger volume of subsurface material.
If the water table is relatively shallow (less than about 8 m below land surface),
removal of all three contaminant phases can be accomplished from the surface by
applying SVE within the unsaturated zone along with suction to a downhole drop
tube with its end located at or slightly below the oil–water interface. At the surface,
the oil and water are separated. The oil can then be recycled (or used to power the
internal combustion engine if appropriate), and the water gets treated and released
back into the environment (Fig. 13.3).
Multiphase extraction can be energy intensive, but as with SVE, green and
sustainable options exist. For example, the power to operate the system can
sometimes be obtained from the petroleum hydrocarbon vapors and/or the
LNAPL, considerably reducing or possibly eliminating the need for external
power sources. Multiphase extraction generally is fast and effective, reducing
overall energy use along with reducing long-term esthetic impacts from noise and
infrastructure. It can be deployed quickly as all or part of an interim remedial action
to mitigate urgent cleanup requirements and reduce the need for additional remedial
actions that may be less sustainable. Even if it is not effective at a site, multiphase
extraction generally does not exacerbate existing conditions or make additional
remediation efforts more difficult or less effective. As with SVE, visual and noise
impacts from the extraction system can represent an esthetic nuisance, but can be
mitigated in a similar fashion with appropriate landscaping and noise dampening
components. Although groundwater pumping (and treatment) is included in multi-
phase extraction for the purpose of exposing subsurface soil to unsaturated
conditions, many of the pitfalls (e.g., energy-intensive systems, low dissolved
Fig. 13.3 Concept-level remedial strategy for green and sustainable multiphase NAPL recovery
phase contaminant removal, and high treatment cost) described earlier in this entry
for pump and treat systems would also apply here. Therefore, caution is warranted
when employing groundwater extraction and treatment for multiphase recovery or
for product recovery. The following BMPs are offered to provide a sustainable
approach to groundwater extraction and treatment with multiphase extraction
systems:
• System design should consider modular treatment components that can be
removed or added as needed.
• Variable frequency drive pumps can be used to optimize performance and
reduce energy usage.
• Use of gravity flow where feasible to reduce the need for transfer pumps.
• System operation should have optimized extraction network to minimize
pumping of clean water.
• Use of energy efficient equipment and green energy from alternative energy
providers.
• Regenerate GAC onsite; recycle process residuals.
Monitored natural attenuation (MNA) is a highly sustainable remediation
method that can be used for releases of petroleum hydrocarbon fuels, distillates,
or contamination from other organic compounds. The method utilizes naturally
occurring processes such as biodegradation to reduce contaminant mass and

concentrations that are dissolved in groundwater, present as LNAPL, or sorbed to
the soil. Biodegradation generally destroys most of the components of refined
petroleum hydrocarbons, producing innocuous by-products such as carbon dioxide
and water. As a result, MNA requires little to no external energy input and generates
minimal waste.
Biodegradation of petroleum hydrocarbons readily occurs in many
environments. Microbes have undergone natural selection for millennia, resulting
in microbial communities that have evolved remarkable capabilities to utilize every
bit of energy that can be extracted from oil constituents [12]. Biodegradation occurs
even in extreme conditions. Fouling of fuel system components in aircraft due to
microbial growth in kerosene-based jet fuel was recognized as early as 1956 [13].
For practical purposes, this means that in many (if not most) cases, degradation of
refined petroleum hydrocarbons will proceed under natural conditions without
external energy inputs.
MNA as the sole remediation method has certain limitations that must be
addressed prior to and during implementation. First, it must be demonstrated that
adverse impacts to human and ecological receptors are not occurring and are
unlikely to occur in the future. Second, sufficient evidence for effective natural
attenuation must be shown to exist at a given site. Such evidence can include among
other things a demonstrably shrinking or stable dissolved plume, a reduction of
dissolved concentrations in groundwater over time, and the use of stable isotopes to
determine degradation rates. Third, long-term monitoring is required to ensure that
conditions do not develop to change the efficacy of natural attenuation as a remedial
action. MNA, in a practical sense, does not diverge from the activities associated
with site investigation and monitoring. For this reason, the following BMPs utilized
for site investigation are introduced with MNA activities [14]:
• Perform fewer field mobilizations through the use of flexible work plans and
real-time field measurements as well as onsite mobile laboratory analyses to
determine the next course of action during a single sampling event.
• Utilize small-scale direct push technology drilling equipment for invasive work
or monitoring well installation to reduce fuel consumption, reduce drilling time,
lower air emissions, lower water consumption, produce less noise, and minimize
site clearing and physical impact.
• Use groundwater low-flow sampling equipment to minimize sampling purge
volumes, reduce energy consumption, and reduced derived waste.
• Onsite treatment and recycling of MNA-derived wastes, including site clearing
by onsite composting or landscaping and treatment/reuse of extracted ground-
water for equipment decontamination.
• Collect the meteorological information (e.g., sun duration, wind direction and
velocity) necessary to support the design and installation of off-grid alternative
energy for auxiliary power for MNA monitoring.
• Use of solar or wind-powered telemetry systems to remotely transmit logging
data directly to project offices.
In a variety of circumstances, MNA is used in conjunction with other methods

that may not be sufficient to provide complete remediation of a site by themselves.
There is a saying within the remediation industry that “90% of the remedial effort
goes into cleaning up the last 10% of contamination.” While this is not always true,
monitored natural attenuation can provide a means in which the “last 10%” of the
contamination does not have to be actively remediated because the final cleanup
can be accomplished naturally and sustainably.
Enhanced bioremediation is designed to stimulate contaminant biodegradation
by indigenous microbial populations [15]. Petroleum hydrocarbons degrade most
rapidly under aerobic (i.e., oxygenated) conditions, although other electron
acceptors may be utilized as reactants in biodegradation. Enhanced bioremediation
of these compounds involves the injection of amendments to the contaminated
subsurface to increase oxygen/electron acceptor concentration in the unsaturated
and/or saturated zones. Enhanced bioremediation can supplement monitored natu-
ral attenuation in many cases to improve performance and reduce the time needed
for complete remediation.
Several sustainable methods for promoting bioremediation by adding oxygen
to the subsurface exist. One method used for soil remediation is known as
bioventing. Bioventing works by injecting atmospheric air at low rates into the
unsaturated subsurface zone to displace oxygen-depleted air, thereby stimulating
the growth of aerobic microorganisms and improving biodegradation rates. Gen-
erally, the air is injected into the ground through well points screened in the
contaminated unsaturated zone using a blower similar to that found on many SVE
systems operated at low injection rates. Low air injection rates are important,
particularly in populated areas, due to the potential for soil vapor migration into
basements or other inhabited spaces. The low injection rates are advantageous for
sustainable remediation, as external energy inputs are relatively low and could be
supplied by renewable sources. A novel approach for air injection using mechanical
windmill power (US Patent No. 6,109,358) represents a potential application of
green bioventing technology.
Oxygen addition below the water table for bioremediation purposes can be
accomplished through chemical or mechanical means. Proprietary slow-release
oxygenating compounds for enhancing aerobic bioremediation are available in
several forms, including powder (designed to be mixed with water), solids, and
filter bags. Compounds in powder form generally are mixed with water and injected
into the subsurface using direct push technology. Compounds in solid form or filter
bags are typically suspended within the screened or open interval of monitoring-
type wells.
Mechanical means of adding oxygen (for bioremediation or physical air strip-
ping) below the water table usually takes the form of air sparging. Air sparging
works by injecting air under pressure through one or more wells completely
screened in the saturated zone for the purpose of providing oxygen to the ground-
water. The air travels upward through the porous medium, along preferred flow
paths that form a dendritic pattern (similar to the branches on a tree), until it reaches
the unsaturated zone and becomes exposed to the soil.
Introduction of air in this manner will have two desirable outcomes if it is

working well. First, the air will strip out volatile organic compounds from the
water and saturated soil, transferring these compounds into the vapor phase where
they can be removed through SVE (as needed). Second, the groundwater will pick
up oxygen and become more aerobic, stimulating biodegradation. Air flow rate and
injection pressure generally define whether physical air stripping (of VOCs) or
in-place bioremediation will be the dominant process. If aerobic bioremediation is
promoted through air introduction, the process is referred to as biosparging. The
following sustainable BMPs can be applied to various types of bioremediation
systems for treatment of petroleum hydrocarbons [15]:
• Maximize use of existing or new wells (to avoid resource overutilization) for
addition of reagents that will act as electron acceptors in the biodegradation of
petroleum hydrocarbons.
• Design and use of bioremediation recirculation cells allowing multiple passes of
oxygenated groundwater through fewer wells.
• If oxygen additive is in liquid form, add to the subsurface via trickling gravity-feed
system if high-pressure injections are unnecessary to assure proper distribution.
• If pressurized injection of air is required, evaluate the feasibility of pulsing
rather than continuous injection to increase the efficiency of delivery.
• Employ modular, portable units that can be modified or incrementally reduced as
needed.
• Employ photovoltaic panels or wind turbines to generate auxiliary power for
trailer or for equipment, such as air blowers.
Physical air sparging (for the purposes of stripping and removing VOCs from the
saturated zone) does not work at every site. It will not succeed if the dendritic
patterns of the upwardly traveling sparge zone form with only a few branches, as
contact between the air and the contaminants is essential. The efficacy of air
sparging can be determined through relatively simple pilot tests, however.
Air sparging will require energy inputs to run the air pumps for a considerable
period of time, depending on the size of the impacted area. The systems generally
are run in conjunction with SVE, which is needed for vapor control, so 2–3 years of
operation is not uncommon. If run with an SVE system, visual and noise esthetic
impacts of air sparging are not substantially increased from an SVE system alone.
Phytoremediation is a remedial technology that uses plants to extract, destroy,
and/or contain contaminants in environmental media. Petroleum hydrocarbons can
be degraded within the plants or by plant by-products that are excreted into the soil,
or volatilized into the atmosphere through transpiration.
To date, one of the most commonly implemented types of phytoremediation of
petroleum hydrocarbons in shallow soils is rhizodegradation, or the destruction of
contaminants by microbes whose activity is promoted by plant roots. Plants used for
rhizodegradation include mulberry, hybrid poplars, grasses, cattails, and rice [16].
Phytodegradation, or destruction of the contaminants within the plant, and
phytovolatilization of the contaminants can also be performed by a variety of
trees, scrubs, and herbaceous plants.
Trees including poplars, cottonwoods, and willows can also be used to achieve
hydraulic control to contain groundwater plumes. Often the same plants are used for
both contaminant remediation and containment through hydraulic control. BMPs
for phytoremediation include techniques for minimizing the impact of generic site
operations, including energy conservation, waste minimization, and use of onsite
resources.
The sustainability benefits of phytoremediation technologies include minimal
site disturbance, leading to operational and esthetic benefits, and minimal energy
inputs. Current technologies are limited in their applications by such factors as
plants’ limited contaminant tolerance, limited depth of influence of plant roots, and
often lengthy remediation timeframes. Ongoing research and pilot studies are
focused on developing new and improved methodologies for phytoremediation.
In situ chemical oxidation (also known as ISCO or Chem-Ox) is a method
designed to oxidize contaminants using reactions that break apart chemical bonds,
completely destroying the petroleum hydrocarbon compounds. The primary delivery
method for the chemical oxidants is injecting them in liquid form into the aquifer
using direct push methods or permanent injection wells. The method can work quite
well at many sites, but like any remediation method, it is not universally applicable.
Typical problems include daylighting (chemical oxidants flowing out onto the land
surface or other inappropriate places), preferential oxidation of ambient organic
carbon not related to contamination, desorbing of petroleum hydrocarbons from the
soil into LNAPL form, and poor contact between the oxidant and the contaminants.
The sustainability advantages of ISCO is that it is fast acting, that it destroys
contaminants rather than transferring them to another medium, and that it is injected
using small equipment with a resultant small carbon footprint and minimal site
disturbance. Direct push methods usually leave a boring with a diameter of 5 cm or
less, which is filled in accordance with regulations and finished at the surface with
a sod plug, asphalt patch, or other material such that little to no visual evidence of
the hole exists. Remedial actions using ISCO are most applicable to small sites and
generally are completed after two to six injection events, leaving no long-term
esthetic impacts. Some or all of the following BMPs are examples of ways to
promote a more sustainable ISCO approach:
• Minimize operational impact by constructing little to no long-term or permanent
infrastructure.
• Use of extracted groundwater for onsite mixing of ISCO reagents.
• Use of direct push injections over injection wells if minimal injections are
required.
• Reuse existing injection wells if multiple ISCO injection events are necessary.
Excavation and treatment or disposal is a remediation method not commonly
associated with a green and sustainable approach. However, in many cases, it can be
a direct and verifiable way of removing grossly contaminated material, which can
then allow for more sustainable remedial methods to be implemented. Excavation
can have many undesirable aspects during and after implementation, including
things such as noise, dust, heavy equipment, large trucks, scarred landscapes, and
so on. While these things may be necessary, they can be mitigated with proper
planning and implementation of the remedial action. Some or all of the following
BMPs are examples of ways that can help improve the green characteristics of
excavation actions [17]:
• Devote time and resources to conducting low-impact site characterization (using
direct push methods, field screening, etc.) both before and during the remedial
action. Excavation footprints can sometimes be reduced if efforts are made to
minimize the removal of clean material.
• Use appropriately sized equipment for excavation and hauling to minimize
noise, dust, erosion, and fuel efficiency. Utilize energy efficient operat-
ing procedures such as minimizing idling and performing routine maintenance
to improve fuel efficiency.
• Consider onsite treatment or using the closest treatment and disposal facility to
minimize hauling. Purchase supplies that are produced locally and use local
contractors when possible. Scout for local or onsite backfill material in the
planning stages and during implementation.
• Salvage uncontaminated vegetation and organic debris for use as mulch or
compost.
• Salvage uncontaminated objects and materials (such as steel, storage containers,
etc.) for recycling, reuse, resale, or donation if possible.
• Revegetate and restore excavated areas as quickly as possible. Use native rather
than imported plants. Judicious use of diverse plants and prior planning of final
landforms can be utilized to create valuable habitat.
Chlorinated Hydrocarbons
Chlorinated aliphatic hydrocarbons, which include chlorinated VOCs, are impor-

tant industrial chemicals frequently used as solvents for dry cleaning, degreasing,
metal cleaning, paint stripping, and electronics. They tend to be highly volatile,
toxic to humans, and denser than water. Common chlorinated solvents include
chlorinated ethenes such as tetrachloroethylene (also known as perchloroethy-
lene or PCE), trichloroethene (TCE), 1,2-dichloroethene (1,2-DCE), and vinyl
chloride (VC); chlorinated ethanes such as 1,1,1-trichloroethane and 1,2-dichlo-
roethane; and chlorinated methanes such as carbon tetrachloride, chloroform, and
methylene chloride.
Like refined petroleum hydrocarbons, chlorinated VOCs can be found in the
vapor phase, dissolved in groundwater, sorbed to soil, or as a NAPL. However,
because chlorinated solvents are denser than water, NAPL will be present as
DNAPL. Because of its higher density and other properties such as high viscosity,
a DNAPL tends to be considerably harder to remove, destroy, or otherwise remedi-
ate than LNAPL.
Remediation methods for chlorinated VOCs include some that are the same as or
similar to those discussed in the section on petroleum hydrocarbon remediation.
These and other remediation methods are discussed below in the context of
chlorinated hydrocarbons cleanup. However, the BMPs offered as sustainable
practices discussed previously by technology for petroleum hydrocarbons have
not been repeated in this section.
Soil vapor extraction (SVE) can be effective at removal of chlorinated VOCs
present in the unsaturated zone of the subsurface that is sorbed to contaminated soil
or present in the vapor phase. As described earlier, one or more SVE wells
(screened in the unsaturated zone above the water table) are constructed and
a vacuum system is installed with a manifold to the wells. The vacuum draws
the air from the soil (which is contaminated with VOCs) out of the subsurface
through a treatment system. The contaminated vapors are replaced by fresh air
from vent wells or from other parts of the subsurface, allowing the VOCs to
evaporate and drawn to the SVE for removal. Representative BMPs for SVE
systems discussed in the previous section on petroleum hydrocarbons would be
applicable for chlorinated VOCs with the exception of those related to vapor
control technology.
Treatment of extracted soil vapor for chlorinated VOCs is more complicated
than it is for fuel hydrocarbons. Chlorinated VOCs are not combustible, and when
oxidized, one of the by-products is hydrochloric acid, which is highly corrosive, and
can severely damage or ruin catalytic or thermal oxidizers. The vapors can be
treated with GAC, but this is not green or sustainable and could result in consider-
able change-out expense if concentrations are high.
For higher concentrations, it is sometimes possible to recover the chlorinated
solvents from the vapor stream for recycling. One method uses refrigerated con-
densation to recover VOCs from the vapor phase as liquid [18]. This type of
technology is energy intensive, but can be considered a green remediation method
based on the high efficiency of product recovery from the vapor stream.
Multiphase extraction is sometimes used to recover DNAPL, contaminated
water, and soil vapor from the subsurface. DNAPL, however, is usually more
viscous than LNAPL and is more prone to adhering to the soil matrix, making it
difficult to draw the liquid into a well for product recovery. In most cases, the
recoverable volume is small and the recovered solvent has been degraded to less
than commercial quality, so recycling or reuse of the DNAPL generally would not
be cost effective. As a result, most of the material would be disposed of off-site.
As with the use of multiphase extraction for the recovery of petroleum
hydrocarbons, caution is warranted when employing groundwater recovery and
treatment. The BMPs offered for groundwater recovery with multiphase extrac-
tion of petroleum hydrocarbons would also apply to the recovery of chlorinated
VOCs.
Monitored natural attenuation is a highly sustainable remediation method that can
be very effective for the destruction of certain chlorinated hydrocarbons, particularly
the chlorinated ethenes. The method exploits naturally occurring processes such as
abiotic and biotic degradation to reduce contaminant mass and concentrations that are
dissolved in groundwater, present as DNAPL, or sorbed to the soil. The process

known as reductive dechlorination sequentially strips chlorine atoms off of each
compound, starting with highly chlorinated compounds such as PCE and TCE
through 1,2-DCE and VC, ultimately producing innocuous by-products such as
carbon dioxide, chloride ions, and water. As with petroleum hydrocarbons, monitored
natural attenuation of chlorinated ethenes requires little to no external energy input
and generates minimal waste.
Biodegradation of chlorinated hydrocarbons through reductive dechlorination
does not occur in all subsurface environments. Reductive dechlorination proceeds
best under anaerobic (i.e., oxygen-depleted) conditions as opposed to the aerobic
conditions that are most favorable for degradation of petroleum hydrocarbons.
Additionally, reductive dechlorination requires both electron acceptors (i.e., the
chlorinated ethenes) and electron donors (generally organic carbon that can come
from anthropogenic sources such as landfill leachate or natural sources such as
decaying vegetation) [19]. Third, microbial populations capable of facilitating
reductive dechlorination are not always present, even if anaerobic conditions
exist and adequate organic carbon is available. Therefore, even though MNA can
be a green and sustainable remediation method, it is not always applicable or
effective. In such cases, other methods need to be implemented.
Enhanced bioremediation, which is a method designed to stimulate contaminant
biodegradation by indigenous microbial populations [15], can be a green and
sustainable alternative similar to MNA. Enhanced bioremediation of chlorinated
compounds involves the injection of amendments to the contaminated subsurface to
promote bacterial processes that consume oxygen in the unsaturated and/or
saturated zones, thereby initiating or enhancing reductive dechlorination. In cases
where natural attenuation is actively occurring, enhanced bioremediation can
supplement monitored natural attenuation to improve performance and reduce the
time needed for complete remediation. If biodegradation is not occurring under
natural conditions, enhanced bioremediation can initiate the process by creating
conditions that are amenable to reductive dechlorination.
Enhanced biostimulation of chlorinated compounds typically uses fast-acting
and/or slow-acting, food-grade carbon sources such as lactate, emulsified oils,
molasses, or proprietary compounds developed specifically for this purpose. In
many cases, injection of these carbon sources can be done with low energy input
(such as controlled gravity feed), and they can be left in the subsurface because they
are nontoxic and will eventually be completely consumed by the microbial
populations. Enhanced bioremediation also destroys contaminants rather than trans-
ferring them to another medium, decreasing undesirable environmental impacts that
can occur with some other remedial actions.
Like MNA, enhanced bioremediation can be used as a green and sustainable
component of a larger scale remedial action involving media other than groundwa-
ter. Specific mixtures of carbon substrate can be used as needed to quickly induce
anaerobic conditions or to react slowly and last longer. Other parameters, such as
viscosity, density, emulsification, and so on, can be modified based on site
conditions and remediation goals or requirements.
In some cases, the natural microbial populations are insufficient to degrade

chlorinated hydrocarbons such as PCE and TCE to nontoxic endpoints, which can
cause an undesirable buildup of 1,2-DCE and/or VC. Because VC is more toxic
than TCE, this result is not acceptable from a regulatory or human health perspec-
tive. In such cases, bioaugmentation can be used, in which microbial consortia
capable of complete degradation of chlorinated ethenes are grown in a laboratory,
acclimated to site conditions, and injected into the subsurface with an appropriate
carbon substrate. At present (2011), the only commercially available microbial
consortia for this purpose are of the Dehalococcoides genus.
In situ chemical oxidation (also known as ISCO or Chem-Ox) is used to directly
oxidize chlorinated hydrocarbons, typically by injection of oxidants in liquid form
into the aquifer using direct push methods or permanent injection wells. The
limitations to using ISCO for chlorinated hydrocarbons are similar to those discussed
above for petroleum hydrocarbons and include daylighting, preferential oxidation of
ambient organic carbon not related to contamination, and poor contact between the
oxidant and the contaminants.
In situ chemical reduction can be used in place of or in conjunction with
enhanced bioremediation. Its purpose is to chemically induce strongly reducing
conditions capable of abiotically destroying chlorinated hydrocarbons. A typical
reducing agent is zero valent iron (ZVI), which is a highly reactive metal that can be
deployed as a powder in a permeable reactive barrier or injected into the subsurface
in the form of a liquid suspension. The method is considered to be green and
sustainable because it is long lasting, not harmful to the environment, and chemi-
cally destroys contaminants. The method also can have a residual effect of creating
sustained reducing conditions in the subsurface that serve to enhance bioremedia-
tion, increasing its utility as a remedial tool.
Phytoremediation can be an effective technology for remediating chlorinated
hydrocarbons in soil or groundwater. Plants used for phytodegradation of PCE,
TCE, and DCE include poplars and cottonwoods. As discussed above, these trees
can also be used to achieve hydraulic control to contain groundwater plumes.
Rhizodegradation and phytovolatilization of chlorinated hydrocarbons have also
been observed [16].
PCBs, Pesticides, and High-Molecular-Weight PAHs
PCBs, pesticides, and high-molecular weight PAHs are three of the most common
classes of persistent organic pollutants (POPs), toxic chemicals that resist chemical
and/or biological degradation and therefore persist in the environment. Although
many of these compounds are now banned or highly regulated in the United States,
their environmental persistence necessitates ongoing remediation efforts of histori-
cally contaminated sites.
Many POPs are polychlorinated or polycyclic aromatic compounds. The

incorporation of multiple halogen atoms (chlorine, bromine, etc.) or multiple
aromatic rings into organic compounds tends to make the compounds more difficult
to degrade both chemically and biologically. Because POPs are difficult to degrade
under environmental conditions, containment or ex situ treatments are typically
required.
POPs are typically characterized by low solubility and, therefore, are most often
found contaminating solid materials including soils and sediments. Incineration is
often the presumed remedy for solid wastes contaminated with POPs, which are
destroyed at high temperatures. Benefits of incineration include near-complete
destruction of a wide variety of POPs and the ability to treat large volumes of
contaminated material on relatively short timeframes. However, incineration can
create toxic by-products (e.g., dioxins and furans) or volatilize heavy metals and,
therefore, must be carefully engineered to prevent adverse health effects. Incinera-
tion also requires large quantities of energy to excavate the contaminated soil and to
reach temperatures that often exceed 1,200 C.
Specialized technologies developed as alternatives to incineration include
a variety of techniques for ex situ chemical degradation of POPs [20]. However,
these technologies typically require high temperatures and caustic solutions or
solvents, due to the innate stability of the chemicals. Onsite reuse of treated soil
for backfilling can make these technologies somewhat greener. Vitrification can
also be used to destroy POPs at high temperature, while containing any residual
contaminants in the resulting glass, though this technology is not considered
sustainable, due to excessive energy consumption. The large environmental
impacts of most technologies for POP treatment highlight both the difficulty of
treating these classes of contaminants and the need for additional innovation in this
area of remediation.
In response to the need for innovation, various specialized technologies have
been developed for bioremediation of PCBs and specific pesticides, typically via
dechlorination. These technologies use specialized organic amendments, nutrients,
or proteins to stimulate microbial activity that drives dechlorination of the POPs.
Bioremediation is useful only for wastes with low POP concentrations, due to the
much slower rate of degradation relative to more energy-intensive processes.
Because of the specific nature of microbial processes and the difficulty of either
dechlorinating or breaking the polyaromatic structure of POPs, technologies are
often specific to the compounds for which they were developed and require testing
for applicability to other POPs. These technologies are significantly more sustain-
able than incineration and chemical degradation processes, due to significantly
lower energy requirements and waste production, and because some of the
technologies can be utilized in situ at sites where conditions allow.
Another emerging technology uses mechanical agitation of POP-contaminated
waste with a chemical reactant in a ball mill or similar machine to drive dechlori-
nation, such as for PCBs. Although this technology requires excavation of the
contaminated material, it has relatively low energy requirements, because mechan-

ical energy is substituted for thermal energy in promoting degradation of the POPs.
The sustainability of variations on this method is further affected by the nature of
the chemical reactant used and waste products created.
Ongoing research into application of phytoremediation to POPs has revealed
plants capable of phytoextraction and rhizodegradation of PCBs, pesticides, and
PAHs. Phytoremediation is only possible where POP concentrations are low
enough so that they are not toxic to the plants, and it is particularly promising for
treatment of low-level pesticide contamination of surface soils. Where feasible, this
technology offers a highly green and sustainable alternative to the high energy
intensity and site disturbance required by most POP remediation technologies.
Explosives and Energetics
Explosive and energetic compounds that are in widespread use in military and
industrial applications are common contaminants of concern, on account of their
relative solubility and persistence in the environment. Common examples of these
contaminants include perchlorate, 2,4,6-trinitrotoluene (TNT), hexahydro-1,3,5-
trinitro-1,3,5-triazine (RDX), and octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine
(HMX).
As with POPs, costly and energy-intensive remedial technologies such as incin-
eration are often used to treat energetic compounds in soil. However, these
compounds are less recalcitrant than POPs and can be degraded biologically
under appropriate conditions.
Although microbial degradation of these compounds is typically slow, enhanced
bioremediation of perchlorate, TNT, RMX, and HMX can be achieved by addition
of a variety of carbon and/or nutrient amendments to groundwater, either in situ or
ex situ. Investigations have shown that bacteria capable of degrading these ener-
getic compounds, under both oxic and anoxic conditions, are widespread in the
environment, but specific bacteria can also be added to promote degradation [21,
22]. Ex situ technologies that have been proven effective include a variety of
bioreactors and composting operations. Organic energetic compounds (TNT,
RMX, HMX) can be degraded to other organic by-products, or mineralized all
the way to carbon dioxide. Perchlorate is reduced to chloride and oxygen gas.
Phytoremediation technologies for energetic compounds include
rhizodegradation and phytodegradation of TNT by a variety of plants including
hybrid poplars. Studies have also provided evidence for phytoremediation of
perchlorate, RDX, and HMX. Constructed wetlands have also been found to be
effective for remediating a variety of energetic compounds, through a combination
of bioremediation and phytoremediation.
The BMPs and sustainability benefits of microbial degradation and
phytoremediation of explosive and energetic compounds are similar to those
discussed above for other classes of organic contaminants.
Metals
The diversity of metals and sources of metal contamination in the environment

necessitates diverse contaminant- and site-specific remedial approaches. Common
metal contaminants identified for remediation include arsenic, chromium, lead, and
mercury. Modern and historical sources of metal contamination include mining and
smelting operations (including acid mine drainage), steel production, landfills,
firing ranges, battery recycling and disposal operations, metal plating facilities,
wood treatment facilities, coal combustion by-products (e.g., coal ash), dyes and
paints, lead arsenical pesticides, and leaded fuels.
Unlike organic contaminants, metals cannot be biologically or chemically
degraded to innocuous by-products. Therefore, remediation of metals typically
relies either on removal or on sequestration in solids, which limits mobility and
potential exposure. Various methods exist for increasing the sustainability of
technologies for both removal and sequestration.
In situ stabilization and geochemical fixation of metals uses chemical or
biological processes to transform metals in soils or aquifers from soluble and/or
toxic to insoluble and/or nontoxic forms. The chemistry of stabilization is depen-
dent on the specific metal(s) of concern. Different metals have distinct properties
that affect their mobility and bioavailability under different environmental
conditions, and remedial technologies must be designed to address the specific
site conditions relative to the properties of the metals of concern.
Redox state can be an important determinant of solubility and toxicity for redox-
active metals, such as chromium, arsenic, iron, and manganese. For example,
hexavalent (oxidized) chromium is much more soluble and toxic than trivalent
(reduced) chromium, whereas iron and manganese are typically less soluble and
toxic under oxidized conditions. Reducing conditions can also lead to sulfide
production, and sequester metals through the formation of metal sulfide minerals.
Therefore, addition of chemicals that stimulate reduction or oxygenation of the
contaminated soils or aquifers can be designed to cause the metal(s) of concern to
precipitate and become less bioavailable.
Various properties of soil and water geochemistry also affect metal behavior. Many
metals are more soluble and bioavailable under acidic (low pH) conditions, and
carbonate, in the form of lime or limestone, can be added to increase pH. The presence
of other geochemical species (e.g., phosphates or organic matter) can cause metals
such as lead, zinc, copper, cadmium, nickel, and uranium to be retained in the solid
phase. Phosphate causes metals to precipitate, whereas addition of organic matter
tends to increase metal sorption to soils. Addition of solids that tend to adsorb metals,
such as iron oxyhydroxides, can also effectively sequester metals.
Implementation of in situ stabilization typically requires that an aqueous or solid
chemical reactant be injected into the contaminated aquifer or mixed with impacted
soils. Application of this technology in aquifers therefore requires the installation of
injection wells, while remediation of soil requires equipment sufficient to mix to the
depth of desired treatment. The BMPs for in situ stabilization include some
discussed above for enhanced biodegradation and ISCO, technologies which

require similar injection of reactants, and for excavation, which requires site
disturbance similar to soil mixing.
As an in situ technology, stabilization offers relatively low energy requirements,
as compared to extractive technologies. The environmental impacts of the stabili-
zation as applied to groundwater are dependent on the frequency of injection events
and level of automation of the system, which can minimize the number trips to the
site for operations and maintenance of the injection system. Metals stabilization in
soils has minimal energy requirements after the initial mixing. Use of recycled
materials, such as composts, wood or coal fly ash, or red mud from the alumina
industry, as treatment additives can further increase the overall sustainability of the
remedial effort.
Limitations of this technology include the difficulty of achieving thorough
mixing with the contaminated medium, possible limitations to future site use, and
the necessity of treatability studies to demonstrate that metal solubilities will meet
criteria after treatment. However, in sites where the contaminant and physical
characteristics are well suited for in situ stabilization, it can be a highly effective
and environmentally sustainable remedial option.
Phytoremediation of metals uses plants to extract and/or contain contaminants in
environmental media. Whereas remediation of organic compounds by plants often
focuses on contaminant degradation, phytoremediation of metals focuses on
accumulating the metals within plant tissues, volatilizing them through the plants,
stabilizing them in soils, or otherwise limiting metal mobility and bioavailability.
A wide variety of metals can be removed from solid materials or water through
phytoextraction, which results in accumulation of metals in plant tissues. Plants
used for phytoextraction are often chosen for their ability to tolerate and/or
“hyperaccumulate” a specific metal of concern. For example, Chinese brake ferns
are used to remove arsenic from shallow soils (Fig. 13.4). Because the metals are
not degraded, plants typically must be harvested and disposed of as waste following
metal uptake. Plants are also used for phytovolatilization of certain metals (e.g.,
selenium, mercury, and arsenic) through transpiration.
Plant-based containment technologies for metals include phytostabilization
within the soil, achieved through chemical or biological processes that decrease
metal solubility, and hydraulic control, which uses water uptake by trees to limit the
mobility of dissolved metals.
Monitored natural attenuation, discussed in detail above in reference to organic
contaminants, is also applicable to dissolved metals in aquifers, in cases where the
natural groundwater chemistry results in immobilization of the contaminants. For
example, hexavalent chromium may be naturally oxidized in aquifers with suffi-
cient dissolved oxygen. Precipitation of metals at naturally high pH or adsorption of
metals to solid materials present in the aquifer (e.g., minerals and organic matter)
can also result in the attainment of remedial goals for dissolved metals.
As with remediation of organic contaminants, monitored natural attenuation of
metals requires collection of evidence to indicate that natural attenuation is occur-
ring and that no adverse impacts to humans or the environment are present. Where
Fig. 13.4 Use of Chinese brake ferns for remediation of arsenic in shallow soils
these conditions are met, this remedial strategy is a highly sustainable method for
achieving remedial goals with minimal energy inputs.
Excavation is not typically considered a green or sustainable technology, due to its
high energy demands and the need to contain excavated waste within a disposal
facility. However, as discussed for petroleum hydrocarbons above, certain methods
can be incorporated into excavation projects to improve their sustainability. For
metals, one additional option is excavation followed by metal recycling, which can
be an environmentally responsible remedial option for waste with high metals
concentrations, such as slag from metal smelting operations and munitions from firing
ranges. The economic feasibility of recycling is often dependent on the concentration
of metals, the quantity of contaminated material, and proximity of recycling facilities.
Future Directions
This entry has provided practical concepts, value engineering principles, and BMPs
for the current state of remedial technologies for the typical classes of environmen-
tal contaminants observed in the subsurface. These concepts and practices represent
the qualitative principles currently utilized to improve or increase the sustainability
of subsurface remediation.
Although these BMPs are helpful, is it not possible based on BMP application to
evaluate their ultimate benefit to improving sustainability. In the near future, the
environmental industry will move away from the general practice of subjectively
reviewing and applying BMPs from a master list to the practice of applying project-
specific, robust, quantitative analyses of sustainability benefit for a project. Green
and sustainable remediation organizations like SURF have already created
a compilation of comprehensive metrics (Metrics Toolbox) that can be used to
evaluate, track, and forecast a remedy’s ability to achieve certain outcomes in
relation to sustainability goals [23]. In the future, metrics like those in SURF’s
Metrics Toolbox will be supplemented with a wider suite of metrics to perform
analyses of sustainability key elements in remedial program decision making.
Although it is envisioned that greener remedies have a distinct place within
regulatory programs, such as the US Environmental Protection Agency (EPA)
Comprehensive Environmental Response, Compensation, and Liability Act
(CERCLA) or Resource Conservation and Recovery Act (RCRA), program-
specific regulatory criteria do not currently address the social and economic
considerations of sustainability. The US EPA is presently attempting to clarify
the role of green remediation within the CERCLA and RCRA programs; however,
its ability to include the social and economic benefits in the remedy selection
process may exceed the authority of these programs. The task of defining what is
meant by the term “sustainable” in terms of remedial measures remains an ongoing
effort [24] and will continue to evolve in coming decades through collaboration
between researchers, regulators, and environmental remediation professionals.
Bibliography
Primary Literature
1. Office of Technology Assessment (OTA) (1984) Protecting the nation’s groundwater from
contamination, report
2. Howard PH, Boethling RS, Jarvis WM, Meylan WM, Michalenko EM (1991) Handbook of
environmental degradation rates. Lewis, Chelsea, p 725
3. Suthersan SS (1997) Remediation engineering: design concepts. Lewis, Boca Raton, p 362
4. Kalderis D, Juhasz AL, Boopathy R, Comfort S (2011) Soils contaminated with explosives:
environmental fate and evaluation of state-of-the-art remediation processes (IUPAC technical
report). Pure Appl Chem 83:1407–1484
5. Liu L, Tindall JA, Friedel MJ (2007) Biodegradation of PAHs and PCBs in soils and sludges.
Water Air Soil Pollut 181(1–4):281–296
6. U.S. Environmental Protection Agency (EPA) (1989) Risk assessment guidance for superfund.
Volume I: Human health evaluation manual, Part A, EPA/540/1-89/002. U.S. Environmental
Protection Agency, Office of Emergency and Remedial Response, Washington, DC
7. U.S. Environmental Protection Agency (EPA) (1989) Risk assessment guidance for superfund.
Volume II: Environmental evaluation manual, EPA/540/1-89/001. U.S. Environmental Pro-
tection Agency, Office of Emergency and Remedial Response, Washington, DC
8. Office of Solid Waste and Emergency Response (1995) Use of Risk-based decision making in
UST corrective action programs. OSWER Directive 9610.17, Mar 1995. Office of Solid Waste
and Emergency Response, Washington, DC
9. Foster Wheeler Environmental Corporation (Foster Wheeler) (1998) RBCA fate and trans-
port models: compendium and selection guidance, prepared for ASTM, Nov 1998, pp 1–26
10. Ellis DE, Hadley PW (2009) Sustainable remediation white paper – integrating sustainable
principles, practices, and metrics into remediation projects, remediation. Wiley Interscience.
doi:10.1002/rem.20210
11. U.S. Environmental Protection Agency (USEPA) (2011) Green remediation best management
practices – sites with leaking underground storage tank systems, EPA 542-F-11-008, June
2011. Office of Solid Waste and Emergency Response, Office of Emergency and Remedial
Response, Washington, DC
12. American Academy of Microbiology (2011) Microbes and oil spills (FAQ), p14
13. Crane CR, Sanders DC (1967) Evaluation of a biocidal turbine fuel additive, Aug 1967.
Federal Aviation Administration AM 67-21, Washington, DC, p12
practices – site investigation, EPA 542-F-09-004, Dec 2009. Office of Solid Waste and
Emergency Response, Office of Emergency and Remedial Response, Washington, DC
practices – bioremediation, EPA 542-F-10-006, Mar 2010. Office of Solid Waste and Emer-
gency Response, Office of Emergency and Remedial Response, Washington, DC
16. U.S. Environmental Protection Agency (USEPA) (2000) Introduction to phytoremediation,
EPA/600/R-99/107, Feb 2000. Office of Research and Development, Cincinnati
17. U.S. Environmental Protection Agency (USEPA) (2008) Green remediation – best man-
agement practices for excavation and surface restoration, EPA 542-F-08-012, Dec 2008.
Office of Solid Waste and Emergency Response, Office of Emergency and Remedial
Response, Washington, DC
18. Kessel L, Squire J, Holland K (2008) Sustainable soil remediation by refrigerated condensa-
tion at sites with “high-concentration” recalcitrant compounds and NAPL – two case studies.
Remediation 19:53–72
19. U.S. Environmental Protection Agency (USEPA) (1998) Technical protocol for evaluating
natural attenuation of chlorinated solvents in ground water, EPA/600/R-98/128, Sept 1998.
Office of Research and Development, Cincinnati
20. U.S. Environmental Protection Agency (USEPA) (2010) Reference guide to non-combustion
technologies for remediation of persistent organic pollutants in soil, 2nd edn. EPA 542-R-09-
007, Sept 2010. Office of Solid Waste and Emergency Response, Office of Emergency and
Remedial Response, Washington, DC
21. Interstate Technology and Regulatory Council (ITRC) (2005) Perchlorate: overview of issues,
status, and remedial options, Sept 2005. ITRC, Washington, DC
22. Esteve-Nunez A, Caballero A, Ramos JL (2001) Biological degradation of 2, 4, 6-trinitrotoluene.
Microbiol Mol Biol Rev 63:335–352
23. Butler PB, Larsen-Hallock L, Lewis R, Glenn C, Armstead R (2011) Metrics for integrating
sustainability evaluations into remediation projects. Remediation 21(3):81–87
24. Interstate Technology and Regulatory Council (ITRC) (2011) Green and sustainable remedia-
tion: state of the science and practice. ITRC, Washington, DC, 43 pp
Books and Reviews
Fetter CW (1999) Contaminant hydrogeology, 2nd edn. Waveland, Long Grove, p500
Holland KS et al (2011) Framework for integrating sustainability into remediation projects.
Remediation 21(3):7–38
Illaszewicz J, Gibson K (2009) Green and sustainable remediation: creating a framework for
environmentally friendly site cleanup. Environ Qual Manag 18(4):1–8
Pankow JF, Cherry JA (1996) Dense chlorinated solvents and other DNAPLs in groundwater.
Waterloo, Portland, 522 pp
Schwarzenback RP, Gschwend PM, Imboden DM (1993) Environmental organic chemistry.
Wiley, New York, 681 pp
Stumm WS, Morgan JJ (1981) Aquatic chemistry: an introduction emphasizing chemical
equilibria in natural waters, 2nd edn. Wiley, New York, 780 pp
U.S. Environmental Protection Agency (USEPA) (2008a) Green remediation: incorporating sustain-
able environmental practices into remediation of contaminated sites, EPA 542-R-08-002, Apr
2008. Office of Solid Waste and Emergency Response, Washington, DC
Index
A turbulent and molecular transport,

Acrolein, 107, 110 177–178
aerobic bioremediation, 354–355 viscous boundary layer, 179
air pollution, 38 atrazine, 8–9, 131
air sparging, 350, 354 pesticide, 131
air-water, 6, 177, 179, 187, 190
gas exchange, 177, 190
gas transfer, 179, 187 B
oxygen transfer, 6 bicarbonate, 290
alkane, 51 bioaccumulation, 48
Alkene, 51 biochemical oxygen demand (BOD), 262
anthropogenic metals, 314–320 bioconcentration factor (BCF), 48
aquasol, 202 biodegradation, 45, 226, 354
aquatic, 5, 215, 242–243 biodiffusion, 205, 211
ecosystems, 5, 243 transport coefficient, 205
environments, 215 biofilm, toxin diffusion, 38
interface compartment (IC) biogeochemical cycles, 42
concept, 215 biomagnification, 49
systems, 242 bioremediation, 354, 359
Arrhenius function, 226 biosparging, 355
Arrhenius temperature, 28 biotransformation, 45
arsenobetaine, 320 bioturbation transport, 202
atmosphere/atmospheric carbon bioventing, 354
dioxide, 177 bluff surface jets, non-buoyant, 153
atmosphere-water exchange, 175, 177–179, bluff wall jet, 152
184, 186–188 Boltzmann constant, 26
description of transport, 179 Boussinesq, 95, 98, 110
eddy correlation flux measurements, 188 approximation, 98
geochemical tracer techniques, 187 assumption, 110
influence of surface films, eddy diffusion coefficient, 95
186–187 brook, 220
influence of waves, 184–186 Brownian, 14, 202
laboratory facilities, 187 diffusion, 202
mass boundary layer, 178 motion, 14
thermography, 188–189 bubble plume, 167
tracer injection, 188 bubbly jet, 165–167
DOI 10.1007/978-1-4614-5731-2, # Springer Science+Business Media New York 2012
370 Index
bulk mean velocity, 115 circular plume crossflow, 161

buoyancy/buoyant, 147–149, 152 Clean Water Act (CWA), 242
flux, 148 cold water, fish oxythermal habitat, 273
jet, 147–149 concentration unit, 31
surface jets, 152–155 contaminant, 204, 342
biological degradation, 342
transport model, 204
C controlled flux technique (CFT), 188
calcium, 290 convection transport, 110
carbon dioxide, 289 convective, 68, 71, 73
carbon sequestration, 11 flux rate, 68, 71
catechol, 57 transport terms, 73
chemical/chemicals, 1–6, 8, 13, 49, 66, 69, Coriolis effects, 303
89, 113, 122, 141 Courant number, 120
diffusion, 2 cultural eutrophication, 242
dispersive transport, 113 cyclic siloxane, 58
fate, 1, 5 cycloalkane, 51–52
hydrophobicity, 49
importance of mixing, 5
in the environment, 1, 13, 89, 113, 141 D
dispersive transport, 113 deep ocean waters, 309
fate, 1 deep ocean, 309, 324
transport with jets and plumes, 141 hydrothermal activity, 309
transport, 1 sediments, 324
turbulent transport, 89 waters, 309
interfacial transfer, 4 scavenged elements, 309
mass rate of accumulation, 69 diatom, 310
multiphase transport, 4 dichlorobenzene, 83, 86–87
retardation coefficient, 122 total concentration (TDCB(z)), 86
reverse osmosis, 2 diethyl phthalate (DEP), 56
transport processes, 2 diffusion, 4, 13–15, 17–18, 20–24, 26, 28,
transport with jets and plumes, 141 66–67, 78, 91–92, 94, 139, 178
transport, 1–2, 6, 8, 66 across a thin film, 15
processes, 2 across interfaces, 28
resistance, 6 boiling flux, 21
turbulent diffusion, 3 coefficient, 17, 22, 24, 26, 28, 92, 178
turbulent transport, 89 in gases, 26
chlorella residence time, 129 in liquids, 28
chloride, 129, 290 in solids, 28
chlorinated hydrocarbons, 360 concentrated, 20
chlorinated, 341, 357–360 dispersion, 23
aliphatic hydrocarbons, 341 equation development, 4, 20, 66–67, 78, 94
hydrocarbons (CHCs), 357–360 flux doubles, 17
biodegradation, 359 gaussian concentration profiles, 24
enhanced bioremediation, 359 mathematical description, 14
enhanced biostimulation, 359 permeance, 17
in situ chemical oxidation, 360 unsteady, 18
monitored natural attenuation, 359 diffusive flux rates, 67–68, 72
multiphase extraction, 358 diffusive transport with convection, 118
soil vapor extraction (SVE), 358 dimethyl phthalate (DMP), 56
VOCs, 358 dioxins, 54
chlorofluorocarbon (CFC), 305 chlorination, 54
circular jet, 142, 148 non-chlorinated, 54
Index 371
Dirac function, 20, 133 G

disinfection, 117, 129 Gas Exchange, 183
longitudinal, 117 gas/gaseous, 179, 183, 190
tracer determination, 129 exchange, smooth water surfaces,
dispersion, 3, 22–23, 114–115, 117, 121, 179, 183
123–125, 139 transfer, 190
coefficients, 23, 115, 117, 123 gas–liquid interface, 196
in groundwater flow, 121 Gaussian, 82, 145
in laminar flow, 115–117 distribution, 145
in rivers, 125 probability distribution, 82
in turbulent flow, 117–118 granular activated carbon (GAC), 350
Peclet number, 124 green chemistry, 42
plug flow, 115 groundwater, 80, 341, 347, 351–352,
dissolved oxygen, 242, 262 357, 359
drinking water pollution, 133 chlorinated aliphatic hydrocarbons, 357
dynamic(s) roughness, 100, 109 contamination, 80, 341
multiphase extraction, 351
pumping, 347, 351
E
eddy, 96, 101, 118, 188
correlation, 188 H
diffusion coefficient, 101 harmful algal bloom (HAB), 293
diffusivity, 118 heat conduction equation, 261
velocity, 96 Henry’s law constants, 6, 8, 33, 35,
empiric parameterization, atmosphere-water 227, 342
exchange, 190 hexachlorocyclohexane, 44
endocrine disrupting compound (EDC), 58 hydrodynamic, 238
energetics, 362 hydrogen sulfide, 29
environment/environmental, 19, 236 hydrolysis, 225
engineering, 19 hydrophobic compound, 75, 77
fluid dynamics code (EFDC) model, 236 hydrosphere, 176
equation, 245 hypoxic water, 310
Escherichia coli, 293
ester, 56
estuary/estuarial, 299 I
phytoplankton productivity, 299 in situ chemical oxidation, 356
redox processes, 299 infinite couple, 29
ether, 55 inland waterbody, 243
Eulerian-integral model, 164 inorganic chemicals, physicochemical
properties, 344
interface/interfacial mass transfer, 30, 31, 196
F
Fick’s law, 15–16, 21, 68, 165, 202
fish, 272–273 J
habitat projections under future climate jets, 142, 150, 155, 159, 163
scenarios, 273 deflection, 159
habitat simulation program, 272 effect of boundaries, 150
fluid density interface, 96 velocity, 163
flux equations, 32
Fourier transformation, 84
Freundlich isotherm, 76 K
Froude number, 168 kinematic viscosity, 90
Furan, 54 Knudsen diffusion, 27
372 Index
L M
Lagrangian coordinate system, 119 magnesium, 290, 300
lake, 242–243, 245–246, 248–257, 259, Manning equation, 228–229
262–264, 266, 269–271, marine, 203, 309
273–274, 277 phytoplankton carbonic anhydrase, 309
acid rain, 248 snow, 203
aging process, 242 mass balance, control volume, 70, 73, 245
atmospheric input, 248 mass conservation equation, 66
biological processes, 250 mass transfer, 29–31, 33, 35–39
biology, 245 coefficients, 31, 33, 35–37, 39
biomanipulation, 277 of ammonia, 37
chemical kinetics, 251 of oxygen, 35–36
chemistry, 245 interfacial, 30
density- or wind-driven currents, 255 mass transport equation, 81, 92–95, 120
dissolved oxygen, 266 metalloids in marine systems, 320–321
eutrophication, 242, 253, 256 metal, 363–365
external material inputs, 246 contaminants, 363
geometry ratios, 255, 274 geochemical fixation, 363
hydraulic residence time, 252 stabilization, 364
hypolimnia, 274 excavation, 365
ice-cover model, 257 monitored natural attenuation, 365
internal nutrient loading, 250 phytoextraction, 364
lake bottom diffusive oxygen flux, 262 phytoremediation, 364
lake sediment, 251, 256, 259, 269 phytostabilization, 364
air bubble plume, 251 phytovolatilization, 364
heat fluxes, 259 stabilization in soils, 364
of Cisco habitat, 269 Michaelis–Menten formula, 227
of water quality, 256 microbacteria, 5
lethal temperature (LT) isotherm, 270 microscale wave breaking, 186
management, 277 MINLAKE numerical simulation
manipulation of water quality, 251 model, 257
model simulations, 264 mixing length theory, 103
nutrients from watersheds, 248 molecular diffusion coefficient, 178
particle settling, 249 mono-aromatics, 52
phytoplankton concentration, 263 multimedia compartment (MMC)
radiation attenuation, 263 model, 206
resuspension, 249 multiphase jets, 164
stratification, 254 multiple jets, 163
sustainability, 242
sustainable cold-water fish habitat, 273
sustaining fish habitat, 266 N
vertical mixing, 254 natural organic matter (NOM), 44, 292
water temperature, 259, 266 Navier–Stokes equations, Reynolds-averaged,
year-round dissolved oxygen model, 262 144, 145, 147
year-round water temperature, 257 nitrification–denitrification processes, 11
Laplace transform, 84 nitrogen gas, 8–9
Lavoisier mass balance, 207 nonaqueous phase liquid, 66
lentic waterbody, 245
Lewis–Whitman laws, 206
limnology, 243 O
lindane, 51 ocean waters, 311–312, 314
liquid kinematic viscosity, 34 aluminium cycling, 311–312
logarithmic velocity profile, 100 cadmium cycling, 312–314
Index 373
manganese cycling, 311–312 permethrin, 58

mercury cycling, 314–320 persistent bioaccumulative and toxic (PBT), 43
silver cycling, 314–320 persistent organic pollutant, 43, 47, 50–51,
ocean/oceanic, 48, 287, 289, 294, 297–306, 290, 361
309–310, 314, 318, 321, 324–326 chemical classes, 51
anthropogenic inputs, 310, 318 incineration, 361
atmospheric input, 294, 301, 310 Stockholm convention, 50
atmospheric particulate concentrations, 325 vitrification, 361
chemical distributions, 306–310 pesticide, 57, 361
chemicals, 326 petroleum hydrocarbon (PHC), 353–354
chemistry, 289 biodegradation, 353
circulation tracers, 48, 306 bioventing, 354
cycling, 302, 321–326 enhanced bioremediation, 354
of chemicals, 302 monitored natural attenuation (MNA), 353
of organic chemicals, 321–326 phenol, 57
“deep conveyor belt” circulation, 309 photolysis, 226
export of particulate material, 298 photosynthesis/photosynthetic(ally), 263
fate, 287 phthalate acid esters (PAEs), 56
geochemical cycling, 326 phytoplankton, 263, 310
input, 298–300 phytoremediation, 355, 360
from hydrothermal vents, 299 phytovolatilization, 360
from the sediments, 300 piston velocity, 179
from the terrestrial environment, 298 plane plume, 145
organic chemicals, 297, 302 plane jets, 142, 155
organic contaminants, 325 plug flow reactor model, 115
pollutant elements, 327 plumes, 142–143, 145, 150, 155, 160, 161, 165
riverine inputs, 301, 310 coflow, 155
sinking particles, 305 crossflow, 159
“thermohaline” circulation, 303 density, 145
water, 303–304, 311–312, 314, 324 effect of boundaries, 150
column distributions, 325 in stagnant environment, 143
sinking, 304 pollutant mass flux, 147
upwelling, 303 polyaromatic hydrocarbon (PAHs), 53
wind-driven circulation, 303 polybrominated diphenyl ether (PBDE), 55
octanol, 44, 77, 321 polychlorinated biphenyl (PCB), 5, 49, 53, 341
octanol-water partitioning coefficient, 77 polycyclic aromatic hydrocarbon (PAH), 49
Ohm–Kirchhoff law, 196, 205 pore water biodiffusion, 211
Ohm’s law, 205 potassium, 290
open ocean, 293, 296–297 Prandtl’s mixing length, 100, 110
anthropogenic metals, 296
crustal metals, 297
transport of chemicals, 293 R
organic, 44, 343 redox-active metals, 363
carbon (OC), 44 Reichardt’s hypothesis, 163
chemicals, 42 reservoir, 243–244, 246, 248–253, 255–256,
contaminants properties, 343 266, 277
organochlorine (OC), 58 limnology, 244
oxygen concentration in water, 36 nutrients from watersheds, 248
sediment bed, 250
flux of solutes, 250
P sediment transport, 244
partial differential equation (PDE), 76 sustainability, 243
Peclet number, 120 retardation coefficient, 129
374 Index
Reynolds, 34, 90, 110, 148, 166 sediment-water interface

averaging, 110 Lavoisier mass balance, 208
equations, 148 molecular diffusive transport
number, 34, 90, 166 process, 202
rhizodegradation, 355, 360 multimedia compartment (MMC), 208, 210
Richardson number, 154 flux equation, 208
river flow impacts on water quality, 232 model, 210
river, 219–223, 228–230, 232–234, 236, 238 net flux, 201
advection, 230 resistance-in-series (RIS) model, 210
basin, 220 simulation methods, 208
characteristics, 220 the interface compartment (IC)
contaminants, 222, 238 model, 210
dispersion, 230 transport process theories, 202
dissolved oxygen, 233 selenoprotein, 320, 321
fate, 220 simple jet, 143
flow, 228, 232 simple plume, 145
hydrodynamic model, 229 skimmer pump, 348–349
mathematical formulations, 222 slurry jets, 168–169
modeling, 234, 238 sodium, 290
pathogen pollution, 221 soil, 74–75, 349
point source pollution, 220 transport rates, 75
pollutant loadings, 233 vapor extraction (SVE), 349
quality, 229 solute chemical, 69
residence time, 212 sink rates, 69
sediment siltation, 221 source, 69
transport, 219, 228, 236 spatial mean velocity, 114
waterbody, 222 Stokes–Einstein equation, 28
stream, 221
subsurface, 335, 339–340, 344, 346–347, 360
S contaminants transport, 340
Schmidt number, 178, 184, 190 contaminants sustainable remediation,
sediment/sedimentary, 74–75, 106, 204, 346–347
250, 262 energetic compounds, 362
adsorption, 74 explosive, 362
desorption, 74 fate, 335
oxygen demand (SOD), 250, 262 petroleum hydrocarbons, 347–357
transport, 106, 204 remediation, 339, 344–345
sediment–water interface, 195–196, 200–203, sustainable remediation, 345
205–206, 208–210, 212, 215 transport mechanisms, 340
biodiffusion coefficient, 209 transport of chemicals, 335
Brownian transport, 209 sulfate, 300
calculated fluxes, 210 sulfur dioxide, 29
chemical transport, 196 surface jet, 152–155
colloid Brownian diffusivity, 209 suspended sediment, 104, 106
color images, 196
compartment, 201, 206, 215
concentration, 206 T
model, 215 theory of mass transfer, 183
Darcian water advection, 208 thermography, 188
environmental models, 206 toxic(s), organic chemicals, 41–43
flux, 203, 205, 212 characteristics, 43–48
behavior, 212 properties, 43–48
equation, 203, 205 toxin diffusion in a biofilm, 38
Index 375
transfer, 179 V
coefficient, 179 velocity fluctuating component, 92
resistance, 179 volatile organic compounds
transport mass, 91 (VOCs), 349
transport in the environment, 65 volatilization, 227
transverse diffusive transport, 117 von Kármán constant, 178
trichlorobiphenyl, 204
trichloroethylene (TCE), 66, 133
turbulence/turbulent, 10, 90–92, 95, 97, 114, W
139, 142, 184–185 wall jets, 150
diffusion coefficient, 10, 90, 92–97, water quality, 225, 235, 245, 248
114, 139 management, 235
diffusivity, 97, 184 mechanistic studies, 245
eddies, 91, 95 mineralization, 225
flow Prandtl’s mixing length hypothesis models, 225
flow, 91, 97–101 precipitation, 248
jets, 142 studies, 245
transport in the environment, 90
viscosity, 97
wave dissipation term, 185 Y
turbulent jets in stagnant environment, 143 Young’s modulus, 24

John S. Gulliver (Auth.), John S. Gulliver (Eds.) - Transport and Fate of Chemicals in The Environment - Selected Entries From The Encyclopedia of Sustainability Science and Technology-Springer-Verlag

Uploaded by

Copyright:

Available Formats

John S. Gulliver (Auth.), John S. Gulliver (Eds.) - Transport and Fate of Chemicals in The Environment - Selected Entries From The Encyclopedia of Sustainability Science and Technology-Springer-Verlag

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

John S. Gulliver (Auth.), John S. Gulliver (Eds.) - Transport and Fate of Chemicals in The Environment - Selected Entries From The Encyclopedia of Sustainability Science and Technology-Springer-Verlag

Uploaded by

Copyright:

Available Formats

Transport and Fate of Chemicals

Transport and Fate

ISBN 978-1-4614-5730-5 ISBN 978-1-4614-5731-2 (eBook)

# Springer Science+Business Media New York 2012

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

1 Transport and Fate of Chemicals in the Environment,

12 Oceanic Fate and Transport of Chemicals . . . . . . . . . . . . . . . . . . . 287

Concentration The quantity of a compound or chemical per unit volume, unit

J.S. Gulliver (ed.), Transport and Fate of Chemicals in the Environment: 1

Fig. 1.1 Illustration Diffusion

Fig. 1.2 Illustration

Liquid and air bubbles are

Compressor Air is transported along with

where C is the concentration of a chemical; t is time; u, v, and w represent the temporal

Table 1.1 Examples of short-term and long-term fate

Chemical fate is the eventual short-term or long-term disposition of chemicals,

The Importance of Mixing

Mixing is a rate-related parameter in that most rates of reaction or transport are

Rate of process = Mixing parameter

Figure 1.4 gives an example of the adsorption of a compound to suspended

Volatilization of spills Turbulent mixing and dispersion !0

Fig. 1.4 Adsorption analogy Compound R1 R2 Sorbed

to diffusive transport, and is considered as a second resistance, R2, that acts in

Fig. 1.5 Air–water transfer Volatile Nonvolatile

Volatile compounds: Nonvolatile compounds:

Nitrogen gas would be a volatile compound because the equilibrium is strongly

Fig. 1.6 Transport to Primary

Convection Mass transfer effected by flow due to applied forces

J.S. Gulliver (ed.), Transport and Fate of Chemicals in the Environment: 13

Definition of the Subject

Diffusion is mixing without stirring. It is mixing caused by Brownian motion, that

The basic mathematical description of diffusion is Fick’s Law, suggested by

Fig. 2.1 Concentration

Diffusion Across a Thin Film

ðmass accumulationÞ ¼ ðmass diffusing in outÞ

This restates the assumption of steady-state diffusion, independent of time.

This is subject to two boundary conditions:

z¼0 c1 ¼ c10 (2.5)

This is enough to solve this important problem.

where A is an integration constant. Integrating a second time gives:

where H is a partition coefficient, the ratio at equilibrium of the concentration inside

Other Important Cases

Fig. 2.2 Concentration

t¼0 all z c1 ¼ c11 (2.17)

t¼0 z¼0 c1 ¼ c10 (2.18)

The concentration profile in this case is:

This is the key result for this case.

In these conditions, M is the total solute injected, A is the cross-sectional area,

In this Gaussian concentration profile, the quantity in square brackets is the

½Total flux ¼ ½Diffusion flux þ ½Convective flux

We now turn to an environmentally important problem mathematically similar to

Increasing diffusion increases dispersion. If there is a small, laminar flow of

Dispersion, now independent of diffusion, is due to the coupled turbulent

where d is an equivalent thickness of an absorbent and D0 is the diffusion coefficient

Diffusion is an important process because it is slow, and diffusion coefficients thus

Table 2.3 Variations of diffusion coefficients

½Volume of one molecule ¼

where s is the molecular diameter. Equations 2.41–2.43 can be combined to find:

½Total flux ¼ ½Diffusion flux þ ½Convective flux

½Volume of one molecule ¼

Force = ½Coefficient of friction f Velocity v1 (2.46)

½Accumulation in the air ¼

½Accumulation in dV ¼ ½CO2 Flow in out

c1 ðgasÞ ¼ 30 104 mol/‘ (2.67)

13 v 0:67 D0:50