2016-Castellazzi-Assessing Groundwater Depletion and Dynamics Using GRACE and InSAR Potential and Limitations
2016-Castellazzi-Assessing Groundwater Depletion and Dynamics Using GRACE and InSAR Potential and Limitations
2016-Castellazzi-Assessing Groundwater Depletion and Dynamics Using GRACE and InSAR Potential and Limitations
Abstract
In the last decade, remote sensing of the temporal variation of ground level and gravity has improved our
understanding of groundwater dynamics and storage. Mass changes are measured by GRACE (Gravity Recovery
and Climate Experiment) satellites, whereas ground deformation is measured by processing synthetic aperture
radar satellites data using the InSAR (Interferometry of Synthetic Aperture Radar) techniques. Both methods
are complementary and offer different sensitivities to aquifer system processes. GRACE is sensitive to mass
changes over large spatial scales (more than 100,000 km2 ). As such, it fails in providing groundwater storage
change estimates at local or regional scales relevant to most aquifer systems, and at which most groundwater
management schemes are applied. However, InSAR measures ground displacement due to aquifer response to
fluid-pressure changes. InSAR applications to groundwater depletion assessments are limited to aquifer systems
susceptible to measurable deformation. Furthermore, the inversion of InSAR-derived displacement maps into
volume of depleted groundwater storage (both reversible and largely irreversible) is confounded by vertical and
horizontal variability of sediment compressibility. During the last decade, both techniques have shown increasing
interest in the scientific community to complement available in situ observations where they are insufficient. In
this review, we present the theoretical and conceptual bases of each method, and present idealized scenarios to
highlight the potential benefits and challenges of combining these techniques to remotely assess groundwater
storage changes and other aspects of the dynamics of aquifer systems.
Introduction
1 Corresponding author: Institut national de la recherche scien- Groundwater systems play a central role in sustain-
tifique, Centre Eau, Terre et Environnement, Université du Québec, ing ecosystems and providing humanity with high-quality
490 rue de la Couronne, Québec, QC, Canada G1K 9A9; (418) freshwater (Taylor et al. 2013). The reliance on ground-
570-3630; fax: (418) 654-2600; [email protected] water will probably intensify under climate change, as
2
Institut national de la recherche scientifique, Centre Eau,
Terre et Environnement, Université du Québec, 490 rue de la it affects precipitation patterns, and as adaptation strate-
Couronne, Québec, QC, Canada G1K 9A9; (418) 654-2683; fax: gies generally rely on groundwater resources (Kundzewicz
(418) 654-2600; [email protected] et al. 2007; Green et al. 2011), pointing out the necessity
3
United States Geological Survey, Water Science Field
of adequate observation tools for management.
Team – West, 5957 Lakeside Boulevard, Indianapolis, IN 46278-
0000; (916) 801-2040; [email protected] As a diffusive media storing water, groundwater
4 Géosciences Rennes, UMR CNRS 6118. Université Rennes 1, systems respond to natural and human-induced changes
Campus Beaulieu. 35042 Rennes Cedex, France; (33) 223 236 546; in external and internal boundary fluxes (recharge and
[email protected] discharge) by changes in storage and fluid pressure
5 Geological Survey of Canada, Natural Resources Canada, 490,
rue de la Couronne, Quebec, QC, Canada G1K 9A9; (418) 654-2688; (head), which in turn affect flow. Groundwater depletion
fax: (418) 654-2615; [email protected] is one of the many objective and subjective factors used
[Correction added on 12 September 2016, after first online to determine the sustainability of groundwater resources.
publication: section GRACE: Principle and Application, Theoretical Overexploited aquifer systems typically exhibit ground-
Scenarios, and Fig. 2].
Received March 2016, accepted July 2016. water depletion, that is, a long-term decrease in the
© 2016, National Ground Water Association. volume of stored groundwater (Konikow 2015). Ground-
doi: 10.1111/gwat.12453 water extraction modifies the groundwater flows to satisfy
(B)
Figure 1. (A) Storativity vs. radius of influence according to the Cooper-Jacob approximation; (B) storativity vs. drawdown
in an observation well located 50 m distant from a pumping well according to the Theis nonequilibrium equation. Typical
values of storativity are highlighted in blue for confined conditions, and in green for unconfined conditions. Transmissivity is
100 m2 /d, 100 d of pumping at a constant rate of 200 m3 /d.
ideal aquifer response (drawdown, or h) to pumping The radius of influence, that is, the horizontal effect
at a specified rate in both types of aquifers of infinite of head loss, increases strongly when storativity decreases
extent (assuming similar hydraulic conductivity and other (Figure 1A). The cone of depression is deeper near the
aquifer properties) results in a larger extension/volume pumping well and spreads more widely for confined
of the cone of depression for confined aquifers. In other vs. unconfined conditions. While the total mass change
words, the change in storage (GWS) needed to supply is equivalent for both cones of depression, the mass
the pumping volume is derived from a larger volume changes nearby the pumping center are larger for the
of the aquifer material for confined vs. unconfined unconfined aquifer compared with those for the confined
conditions. This analysis applies to porous media only. aquifer. This partially explains good correlations between
water levels and local microgravity measurements in areas
near hydraulic stresses and in unconfined aquifers (Pool
Hydraulic Head Response and Eychaner 1995; Pool 2008). For confined aquifer
The Cooper-Jacob solution is an approximation of conditions, the correlation is poorer (Pool 2008), because
the Theis (1935) nonequilibrium method for the calcula- mass changes are less concentrated in the near-field region
tion of the radius of influence of a steadily pumping well around the pumping well.
as a function of storativity and time in an ideal confined
aquifer (Cooper and Jacob 1946; Dragoni 1998). It relies
on the assumptions of horizontal isotropic, homogeneous Aquifer System Compaction
aquifers of infinite extent, and fully penetrating wells. It Groundwater depletion can lead to ground sub-
can be used for unconfined aquifers with proper correc- sidence. Analysis and simulation of aquifer-system
compaction have been addressed primarily using two
tion. Figure 1 shows the radius of influence (Figure 1A)
approaches: one based on the conventional groundwater
and the drawdown (h 0 − h; where h 0 is the initial head
flow theory (Jacob 1940) and one based on the linear
prior to the imposed pumping stress) as a function of
poro-elasticity theory (Biot 1941). The former approach is
aquifer storativity (S ) using the Theis (Figure 1B) and
a special case of the latter, and both approaches are based
the Cooper-Jacob (Figure 1B) solutions. Note that (1) the
on the principle of effective stress (Terzaghi 1925). For
Cooper-Jacob solution should not be used for unconfined
the discussions here, we follow the approach of Hoffmann
conditions without Jacob’s correction for partial dewater-
et al. (2003b) based on the conventional groundwater flow
ing of water-table aquifers (e.g., transmissivity changes;
theory. Assuming incompressible solid grains, and only
see Halford et al. 2006) and (2) the Theis solution can be
vertical effective stresses and vertical strains, the effective
used for unconfined conditions only when the late-time
stress principle can be expressed (Equation 7):
segment of the Theis well function is considered (Theis
1935; Van der Kamp 1985; Kruseman and De Ridder
σt = σe + p (7)
1991), that is, when the delayed water-table response can
be overlooked.
Germany) are popular (see http://gracetellus.jpl.nasa of the observation system in the hydrogeologist’s toolbox
.gov/). Other solutions are available online at The Institute (Alley and Konikow 2015). For further application and
of Theoretical Geodesy and Satellite Geodesy (Graz Uni- resolution of scientific challenges, the main limitations
versity of Technology, Austria) and The Research Group of GRACE are the limited spatial resolutions as well
for Space Geodesy (GRGS, French National Space Cen- as the limited length of the available observations
ter) websites. The University of Colorado and GRGS cre- (see e.g., Scanlon et al. 2015). A recent consensus on
ated interactive portals to compute GRACE time-series the science and user needs for future satellite gravity
over a region, a country, or a watershed (http://geoid observing systems has been synthetized in Pail et al.
.colorado.edu/grace and http://thegraceplotter.com/). (2015). Waiting for the next generation of satellite gravity
The use of GRACE in hydrogeology is still limited to missions, the GRACE follow-on mission is planned
the largest systems, for which it can be considered as one for 2017.