Environmental Modeling & Assessment

https://doi.org/10.1007/s10666-020-09720-y

Impacts of Sea Level Rise on Seawater Intrusion in Cubatão

River, Brazil
Fábio Paiva da Silva 1 & José Rodolfo Scarati Martins 1 & Fábio Ferreira Nogueira 1

Received: 17 October 2019 / Accepted: 8 June 2020

# Springer Nature Switzerland AG 2020

Abstract
Estuarine systems are very sensitive environments to sea level rise as a consequence of climate changes, which can enhance
seawater intrusion and affect multiple water uses. The seawater intrusion under sea level scenarios in an estuarine river by
applying the one-dimensional hydrodynamic and water quality model HEC-RAS 5.0.5 was studied. The study was carried out at
the estuarine reach of Cubatão River, in São Paulo, Brazil. Considering sea level rise scenarios of ΔH = 0.25 m, 0.50 m, and
1.0 m combined with constant freshwater discharge conditions for Cubatão River (16 m3/s, mean annual discharge and 8 m3/s,
dry season discharge), the model results showed that seawater intrusion moves significantly upstream the river in all cases and the
maximum seawater intrusion length may reach 10 km in the worst scenario (ΔH = 1.0 m and 8 m3/s freshwater discharge), 70%
higher than the current sea level and the mean discharge. At the local water abstraction point for urban supply, salinity concen-
tration may reach 12 g/kg, making conventional water treatment unfeasible. Sea level rise may threaten water supply facilities and
require water resource management solutions, such as water abstraction restricted times when salinity concentration is low;
higher freshwater reservation; new water abstraction locations, farther the present ones; or higher water discharges in Cubatão
River from a local hydroelectric power plant, which can cause water resource management conflicts.

Keywords Estuarine system . Cubatão River . Water quality modeling . Seawater intrusion modeling . Sea level rise

1 Introduction on estuarine conservation and sustainable development [1, 3,

12].
Estuaries are of major importance for human development and One of the most important physical processes in estuarine
can provide innumerable benefits such as water for agriculture, areas is the seawater intrusion, which can be described as the
power supply, industrial plants, and urban consumption. These seawater incursion into the estuary as an effect of the tides and
water bodies have some general features, such as the presence of buoyancy. Seawater intrusion in estuaries is a complex phe-
port facilities and essential fishery resources, and they are also the nomenon and depends directly on many environmental con-
final destination of industrial and urban wastewater. ditions, such as the estuary geometry, tidal processes, and
Additionally, estuaries are among the most biologically produc- freshwater inputs. This process may cause high salinity con-
tive systems due to the presence of high nutrient concentrations, centration in upstream parts of estuarine rivers, affecting mul-
which intensify the biological productivity [1–7]. tiple uses of water [8–11].
Estuaries are very sensitive and vulnerable to human activ- Considering that each estuary has specific salinity patterns,
ities, but the environmental impacts were not considered sig- it is crucial to develop a better understanding on its local
nificant until the half of the nineteenth century. After this, behavior, principally in estuaries with high density population
there was an extensive growth of industrial and agricultural or industrial areas, where water resources demand is decisive
activities, fishery and port facilities, and dam constructions for for several activities [9, 12].
hydroelectric purposes, which caused an increasing concern Sea level rise (SLR) and extreme weather events are the
major consequences of climate changes in coastal environ-
ments, which strongly impact their shapes and dynamics.
* Fábio Paiva da Silva
Estuarine systems are particularly sensitive to SLR and are
[email protected] among the most exposed environments to the climate changes
consequences, as it changes not only the sea level but also the
1 hydrological cycles and the river flows. Seawater intrusion
School of Engineering, University of São Paulo, Sao Paulo, Brazil
 da Silva F.P. et al.

resulting from SLR is of great concern for estuarine systems as phenomenon to investigate in order to improve local water
a consequence of climate change. As the combination of tidal management [24–26].
activity and freshwater inputs are the predominant forcing As sea level rise geographic distribution is not uniform,
conditions for seawater intrusion patterns, a major effect of regional or local historical data are needed to determine sea
SLR in estuarine system is the increase on salinity concentra- level trends. In São Paulo state coast and specifically in Santos
tion due to higher seawater intrusion length, which can bay, SLR presents different patterns over time. From 1944 to
affect upstream freshwater that was not previously af- 1989, the SLR in Santos was estimated in 0.1132 cm/year, but
fected or increase salinity to intolerable values for in- considering the period from 1940 to 2014, the SLR pattern
dustrial or urban purposes [6, 13–18]. was of 0.33 cm/year, which follows the global trend of aggra-
The global mean rate of SLR from 1993 to 2010 is estimat- vation of SLR scenarios for the next decades. So, by 2100, the
ed between 2.8 and 3.6 mm/year, which is higher than the mean SLR from 1940 to 2100 will be 1.1 m, but during the
previous estimation of 1.5–2.0 mm/year from 1971 to 2010. next decades, this estimative can be higher. Considering the
It indicates that the rate of SLR is increasing with time and current sea level, a recommended SLR scenario contemplate
predictions must attempt to this behavior [17–20]. the range of 0.5–1.0 m by the year of 2100 [3, 27].
The environmental monitoring followed by numerical In this context, this research proposes the salinity monitor-
modeling are essential tools not only for understanding the ing and the application of numerical modeling techniques to
hydrodynamic and water quality processes but also for devel- investigate seawater intrusion behavior under sea level rise
oping plans for water resources and environmental manage- scenarios for Cubatão River.
ment under SLR scenarios [8, 11, 21, 22].
There are a significant number of recent applications of
hydrodynamic and water quality models to study salinity be- 2 Study Area
havior in estuarine systems [5, 6, 9, 10, 14, 20, 23]. In the past
few years, several studies have been conducted to investigate Cubatão is situated 12 km away from Port of Santos. It is part
estuarine responses to climate change effects, such as alter- of the Metropolitan Region of Baixada Santista and its popu-
ation of freshwater discharge and SLR, and their relation with lation is estimated to be 128,749 (2017) in an area of
seawater intrusion. 142.88 km2 [28]. Cubatão River is part of the Santos estuarine
Cubatão, State of São Paulo, is an emblematic and histor- system, a partially mixed estuary formed by a complex net-
ical example of an industrial city in Brazil which has a vast work of rivers and channels (Fig. 1), characterized by a semi-
experience with environmental issues (Fig. 1). It is located at diurnal tide with diurnal inequalities, resulting in two high
Santos estuarine system and the Cubatão River crosses the tides and two low tides at approximately 6-h intervals.
urban area being an indispensable source of water for human The urban reach of Cubatão River has approximately
and industrial supply. So, water quality is an important aspect 11.4 km and divides the city into two different environments.
of its management and seawater intrusion process is a crucial On its left bank, there is a great amount of industries such as

Fig. 1 Santos estuarine system

and Cubatão River location
Impacts of Sea Level Rise on Seawater Intrusion in Cubatão River, Brazil

oil refinery, petrochemical industry and processing industry; generating energy. HB water discharges are of major impor-
and on its right bank, there is a consolidated urban area [26]. tance for water management in Cubatão River, because it has
The climate pattern in Cubatão is Tropical with some local an impact on water quality parameters and salinity concentra-
variations due to topographic characteristics: near the Serra do tion along the urban reach of the river [25, 30].
Mar foothill, a coastal mountain chain, there is a more intense As a consequence of the urban and industrial water abstrac-
precipitation compared to the lowlands. There is a significant tion and the HB discharge in Cubatão River, its mixing pro-
rainfall along the year, and even during dry periods, there are cess is strongly determined by freshwater discharge; thus, it
considerable rain events. The average annual rainfall is may present different mixing patterns compared to Santos
2541 mm, with higher precipitation events during the wet estuarine system, in which tides are the governing mixing
period, mainly from January to March; and less precipitation mechanism and freshwater discharges play a small role.
events during the dry period, mainly from June to August.
This seasonal variation has a direct effect on the hydrolog-
ical regime, in which the rivers can reach peak discharges 3 Data and Field Survey
during significant precipitation events. Considering that those
rivers are located at the Santos estuarine system, the tidal Field surveys were conducted in two different periods for
activity also plays a significant role in the discharge variation primary environmental data acquisitions—salinity concentra-
during the year. The principal river basins in Cubatão are tion and water level data. The first field survey was conducted
Cubatão River basin (177 km2), Mogi River basin (68 km2), for approximately 4 months, from 03/11/2016 to 04/03/2017
and Perequê River basin (64 km2) [26]. and data were acquired at P1 (Fig. 2). The second field survey
Cubatão industrial area is one of the most important eco- was conducted for approximately 2 weeks, from 14/05/2019
nomic regions in the country, hosting industries, refining oil, to 30/05/2019, and data were acquired at P1 and P2 (Fig. 2).
steel mills, and fertilizers. From 1955 to 1975, Cubatão expe- Salinity was measured in three different depths in the water
rienced an intensive industrial growth, in which many impor- column: near the surface (approximately 0.5 m below the sur-
tant industries started their activities. This urban-industrial face), in the middle (approximately 3.0 m below the surface),
growth led to an increasing natural resources demand and, and near the bottom of the river. (approximately 4.5 m below
consequently, higher water consumption and water resources the surface). It was used at each depth a HOBO U24-002-C
management concerns [24, 25, 29]. Water Conductivity Data Logger with a low range from 100 to
In Cubatão River, the water abstractions are destined to 10.000 μS/cm and a high range from 5.000 to 55.000 μS/cm,
different purposes, such as urban water supply and significant over the range of 5 °C to 35 °C. The salinity accuracy is 5% of
industrial uses, with special consideration for the Companhia the reading and the temperature accuracy is 0.1 °C [31]. The
de Saneamento Básico do Estado de São Paulo (SABESP), the water level was measured at P1 every 10 min using the HOBO
utility in charge of urban water supply, and Presidente Water Level Logger-U20L-01 (Operational Range 0 to 9 m of
Bernardes oil refinery, run by the Brazilian oil company water depth at sea level (maximum error ± 2 cm of water).
PETROBRAS, with a major industrial water consump- Secondary data acquired were tidal elevation at Port of
tion [25, 26]. Santos, HB water discharge, and Cubatão River upstream wa-
An important aspect of Cubatão River water management ter level. The astronomical tidal elevation was obtained from a
is the Henry Borden Hydroelectric Power Plant (HB), operat- tide table for the Port of Santos, in Torre Grande (23.949°S,
ed by Empresa Metropolitana de Águas e Energia SA 46.308°W), approximately 13 km from the confluence of
(EMAE). It is located on the foot of Serra do Mar with two Cubatão and Mogi River, computed upon 45 tidal harmonic
high head (720 m) power houses equipped with 14 groups of constituents [32]. The time series of the astronomical tidal
generators, totalizing 889 MW of power for a maximum flow level was calculated based on the 45 tidal harmonic constitu-
rate of 157.0 m3/s. The water used by the power plant in- ents. Cubatão River upstream water level was taken from
creases the Cubatão River freshwater, which is diverted from Sistema de Alerta a Inundações de São Paulo (SAISP), a vast
rivers (Tietê and Pinheiros) located in the highlands stored at network of monitoring stations operated by the Hydraulic
intermediate reservoirs and driven to the estuary by pipelines. Laboratory of the University of Sao Paulo [33]. Table 1 sum-
However, in 1992, the operation of this system was consider- marizes the primary and secondary data. Figure. 2 illustrates
ably reduced due to water quality problems found in the men- their location on the map.
tioned rivers. In fact, the full reversion can only be done dur-
ing flood events. This operational limitation decreased in ap-
proximately 75% the energy produced in the facility and, con- 4 Model Description
sequently, its original freshwater discharge to Cubatão River.
Nowadays, only 6.0 m3/s is conveyed from HB to keep a The hydrodynamic and water quality model adopted in this
minimum operational flow when the power plant is not study to simulate seawater intrusion in Cubatão River under
 da Silva F.P. et al.

Fig. 2 Primary (P1 and P2) and

secondary data

different forcing conditions was the 1D HEC-RAS in its ver- For unsteady flow, HEC-RAS solves the full 1D Saint-
sion 5.0.5. The U.S. Army Corps of Engineers River Analysis Venant equation using an implicit finite difference method
System (HEC-RAS) is a free software developed by the US (Eqs. 1–4):
Army Hydrologic Engineering Center that allows the user to
perform one-dimensional steady flow, one and two-
∂A ∂ϕQ ∂ð1−ϕÞQ
dimensional unsteady flow and water quality simulations. þ þ ¼0 ð1Þ
All those components are linked and the software uses a ∂t ∂xc ∂x f
unique geometric data and hydraulic routine. The unsteady

 !
flow simulation is adequate to simulate one-dimensional un- ∂Q ∂ ϕ2 Q2 ∂ ð1−ϕÞ2 Q2
steady flow in a network of channels and floodplains; and þ þ
∂t ∂xc Ac ∂x f Af
perform subcritical, supercritical or mixed flow regimes cal-



culations [34]. As mentioned, HEC-RAS incorporates a water ∂z ∂z
quality module that allows the calculation of conservative and þ gAc þ S c þ gA f þ Sf
∂xc ∂x f
non-conservative constituents. These features can be coupled
to the alternate flow directions that are observed within an ¼0 ð2Þ
estuary during incoming tides.
In contrast to 3D models, HEC-RAS does not account
Kc A5=3
multi-directional flows or variations in fluid density. ϕ¼ K¼ ð3Þ
Kc þ K f nP2=3
However, considering the geometry of an estuarine river,
HEC-RAS was expected to reasonably represent the hydraulic
and the depth averaged salinity behavior [34].
ϕ 2 Q 2 nc 2 ð1−ϕÞ2 Q2 n f 2
Sc ¼ Sf ¼ ð4Þ
Rc 4=3 Ac 2 R f 4=3 A f 2

Table 1 Acquired data during field surveys where Q is the total flow, Ac is the cross-sectional area in
the channel, Af is the cross-sectional area in the floodplain, xc
Primary data Salinity (P1 and P2)
is the distance along the channel, xf is the distance along the
Cubatão River downstream water level (P1)
floodplain, P is the wet perimeter, R is the hydraulic radius, n
Secondary data Tidal elevation (Port of Santos)
is the Manning’s roughness value and S is the friction slope. ϕ
Cubatão River upstream water level
indicates how flow is divided between the channel and the
HB water discharge
floodplain and kc and kf are the conveyances.
Impacts of Sea Level Rise on Seawater Intrusion in Cubatão River, Brazil

On the water quality module, HEC-RAS can perform a The water quality module initial conditions were set as null
constituent transport modeling using the QUICKEST- salinity concentration at all cells and the downstream bound-
ULTIMATE explicit numerical scheme to solve 1D ary condition was set with a constant salinity of 32 g/kg, a
advection-dispersion equation. The model can appropriately typical value for Santos Bay, in which salinity can vary from
describe the transport of an arbitrary conservative constituent, 29 to 35 g/kg. A 7-day spinout time was applied to stabilize
e.g., salinity [34]. the model and flush the initial conditions out. The time step for
all simulated scenarios was set of 30 s.

5 Model Setup
6 Model Calibration and Validation
Model topography was taken from a previous study in
Cubatão River, using planimetric and altimetric data and pho- A set of observed data collected during the second field survey
togrammetric survey in order to create contour lines and use (May 2019) was used to calibrate the hydrodynamic and the
them in hydrodynamic and water quality models. The avail- water quality models. As the Manning coefficient (n) strongly
able data were combined to compose a digital elevation model influences the tidal propagation, the hydrodynamic model cal-
(DEM) used to create the model geometry [26]. ibration was carried out by adjusting this coefficient for each
The available topography data was only for the river reaches; river reach. Water surface elevation data at P1 was used to
thus, there was a lack of bathymetric data for the whole estuary and calibrate the hydrodynamic model. The longitudinal dispersion
the open sea. Since in most cases the downstream boundary con- coefficient (Dx) is the key calibration parameter for the water
dition is the open sea with a constant salinity concentration value, it quality module, so salinity concentration data at P1 and P2 were
would be challenging to set up a hydrodynamic and water quality used to calibrate the water quality model by adjusting Dx for
model with this limitation. A 30-km idealized estuary was applied each river reach [6, 9, 11, 35].
downstream the final reach of Cubatão River to represent the The calibration procedure was performed by comparing the
Santos Estuary, so it was possible to set the tidal elevation and simulated data with the observed data in the same period. For
the constant salinity concentration as downstream boundary con- both calibration and validation procedures, the model was
ditions (Fig. 3). assessed using the skill model:
The hydrodynamic model is forced by tides in the down- n
stream boundary condition and by river inflows in the up- ∑ jS−Oj2
i¼1
stream boundary conditions. The hydroelectric power plant skill ¼ 1− n    2 ; ð5Þ
   
discharge was set as a lateral inflow to Cubatão River (Fig. ∑ S−O þ O−O
3). Table 2 summarizes the boundary and initial conditions. i¼1

Fig. 3 Open boundary conditions

 da Silva F.P. et al.

Table 2 Boundary forcing and

initial conditions Boundary Hydrodynamic model WQ model (salinity)

Upstream boundary Cubatão River 10 m3/s Null

Perequê River 4 m3/s Null
Mogi River 6 m3/s Null
Lateral boundary HB 6 m3/s Null
Downstream boundary Idealized estuary mouth Water surface elevation 32 g/kg
(Port of Santos tide station)

where n is the number of observation-simulation pairs, S and be used as tool to predict seawater intrusion behavior in
is the simulated value, and O is the observed value. A Cubatão River.
skill value of 1.0 indicates a perfect performance of the
model, excellent for skill between 0.65 and 1.0, very
good for skill in a range of 0.5 to 0.65, good for skill
in the range of 0.2 to 0.5, and poor for skill less than 7 Investigated Scenarios
0.2 [5, 6, 11, 20].
The best adjustment between computed and observed The simulated scenarios were defined based on the two most
data was achieved for Manning coefficients of 0.06 s/ important forcing factors in an estuarine river: river discharges
m1/3 in the idealized estuary and 0.03 s/m1/3 in Cubatão and tides. The scenarios were investigated considering the
River reach. This last value was also adopted for the SLR impacts on maximum seawater intrusion length in
two left bank tributaries, Mogi and Perequê. Figure 4 Cubatão River considering different flows. The SLR consid-
illustrates the comparison between simulated and ob- ered were ΔH = 0.25 m, 0.5 m, and 1.0 m and the Cubatão
served stage data at P1. River flows (considering the natural river flow and HB dis-
For the water quality model, the Dx coefficients var- charge) were 8.0 m3/s, 16 m3/s, 20 m3/s, 40 m3/s, 60 m3/s,
ied from 370 to 580 m2/s along the idealized estuary 80 m3/s, and 100 m3/s.
and the river sections. Figure 5 illustrates comparison For all scenarios, the horizontal salinity pattern of the
between simulated and observed salinity at P1 and P2 Cubatão River was considered under well-mixed conditions.
in the same period. So, it is not considered salt-wedge stratification or significant
The model validation was performed by comparing vertical salinity stratification. The flows of Perequê and Mogi
the simulated data with the observed data in a different rivers were set as constant discharges of 4.0 m3/s and 6.0 m3/s,
period to certify the calibrated n and Dx coefficients. respectively, and the HB water discharge was set of 6 m3/s (its
The observed data was used from the first field survey, operational constant discharge). In all scenarios, the Petrobras
conducted in 2016 and 2017. Figure 6 illustrates the dam, an in-line level control structure existing in the studied
hydrodynamic model validation, Fig. 7 the water quality river reach, was considered not operating during the whole
model validation. simulation period (Fig. 2). This study focused just on the
The hydrodynamic and water quality models performed SLR effects on seawater intrusion in Cubatão River with no
simulations in agreement with the observed data, with skill considerations related to the changes in runoff or the atmo-
values for both calibration and validation procedures higher spheric forcing conditions like wind and heat flux. All seawa-
than 0.75, which indicates an excellent model performance ter intrusion analyses consider the maximum seawater intru-
(Table 3). So, the model was considered capable of sion length in Cubatão River and the corresponding salinity
performing simulations under different forcing conditions concentration at the farther water abstraction location (Fig. 2).

Fig. 4 Hydrodynamic calibration

at P1
Impacts of Sea Level Rise on Seawater Intrusion in Cubatão River, Brazil

Fig. 5 Water quality calibration

at P1 and P2

8 Results and Discussion enabled the application of the model for seawater intru-
sion investigation with fair results.
In 1D hydrodynamics and water quality models, it is assumed In estuarine systems, the main forcing conditions that im-
that the estuary is fully mixed at the cross-sections during the pact seawater intrusion length are freshwater discharge and
whole simulation period. It means that different mixing con- sea surface height, and a consequence of SLR in estuarine
ditions and stratification cannot be represented, such as the systems is that it impacts the mean water depth and the dis-
salt wedge and the partially mixed estuary types. On the other charge of tidal flow. The higher seawater volume moves sea-
hand, 1D models require fewer input data than 2D or 3D water intrusion upstream the estuary, reaching parts of the
models, making them suitable for studies with few observed estuary that were predominately dominated by freshwater.
data but capable to deliver sufficient preliminary results for Additionally, larger river discharges push the seawater back
decision-making processes in water quality management. to the estuarine mouth, while lower freshwater discharges al-
The model setup presented excellent calibration and vali- low increased seawater intrusion.
dation results considering the downstream boundary assump- The numerical results revealed that typical dry seasons
tion in which an idealized estuary with artificial Manning combined with higher sea surface height caused by SLR leads
coefficient was imposed to reproduce the effects of the estua- to serious concerns about the seawater intrusion at the up-
rine cone. For the hydrodynamic and water quality calibration stream parts of Cubatão River estuarine reach, where strategic
procedures, the minimum skill value was 0.813, and for the water abstraction facilities are placed. This phenomenon can
validation procedures, it was 0.790. Both procedures pre- affect the local water supply system which may lead to severe
sented excellent performances, which indicate that for 1D economic and social impacts in the future.
modeling, even with a lack of bathymetric and observed In this study, the results showed that the seawater intrusion
water quality data, the adoption of an idealized stretch of increases significantly as a consequence of SLR in São Paulo
the water body may not offer a great obstacle and it coastal region. In the reference scenario (Cubatão River

Fig. 6 Hydrodynamic validation

at P1
 da Silva F.P. et al.

Fig. 7 Water quality validation at

P1

discharge = 10 m3/s combined with HB discharge = 6 m3/s), the Santos estuary itself, this estuarine river dynamics strongly
the maximum seawater intrusion length is 5.9 km, but under depends on meteorological forcing conditions for efficient water
SLR conditions, the seawater intrusion goes further the management decision-making processes (Fig. 9).
Cubatão River, reaching the farther water abstraction location Considering a SLR of 0.25 m, the maximum sweater intru-
(Fig. 2) with higher values of salinity concentration. sion profile presented similar behavior to the ΔH = 0.0 m sce-
For the reference discharge, the maximum seawater intru- nario, with a slight increase in salinity concentration at the end
sion profile for a SLR of 0.25 m presented similar behavior to of the river. However, SLR of 0.5 m and 1.0 m would impact
the ΔH = 0.0 m, but both SLR of 0.50 m and 1.0 m caused substantially seawater intrusion patterns in Cubatão River.
prominent changes in seawater intrusion profile. Due to 0.5 m Due to 0.5 m SLR, the salinity concentration at the end of
SLR, the salinity concentration at the downstream end of Cubatão River would be 26.0 g/kg and the seawater intrusion
Cubatão River was 21.0 g/kg and the seawater intrusion length length would reach 7.9 km and, at the water abstraction loca-
moved 600 m upstream, from 7.4 to 8.0 km. For the 1.0 m SLR, tion, the salinity concentration reaches almost 5.0 g/kg.
the maximum salinity at the end of the river would reach almost For the 1.0 m SLR in the dry scenario (8 m3/s), the worst case,
27.0 g/kg, a very saline environment and the seawater intrusion the maximum salinity at the end of the river reached more than
length would increase 2.6 km, from 7.4 to 10.0 km (Fig. 8). 29.0 g/kg and the seawater intrusion length increased to 10 km,
The water abstraction location is 6.5 km upstream the end of which means that almost the whole urban reach of Cubatão River
Cubatão River, which would be strongly affected by a 1 m SLR would be with salinity higher than 1.0 g/kg. During the maxi-
scenario, considering the current mean river discharge. During mum seawater intrusion length, the salinity concentration at the
the maximum seawater intrusion length in spring tides, the salin- water abstraction would reach 12.0 g/kg, a value that makes
ity concentration would reach almost 7.0 g/kg. This high salinity conventional water treatment unfeasible for urban consumption
value strongly impacts water treatment for urban distribution, (Fig. 9).
requiring specific treatment processes; otherwise, it would harm SLR impacts on seawater intrusion for higher freshwater
the treatment plant facilities or the water distribution pipelines. river discharges were also investigated: 20 m3/s, 40 m3/s,
For the dry season scenario (8 m3/s), the maximum seawater 60 m3/s, 80 m3/s, and 100 m3/s. In such a way, it was possible
intrusion profile without considering SLR is already significant. to estimate the maximum seawater intrusion versus the river
While in the reference scenario, the maximum seawater intrusion discharge for each SLR scenario (Fig. 10).
length is 5.9 km, in the dry scenario, it increases 1.5 km, reaching
7.4 km, beyond the water abstraction location. In the studied
Cubatão River reach, seawater intrusion is very sensitive to fresh-
water discharge variation, which indicates that, differently from

Table 3 Skill values (calibration and validation)

Parameter Station Skill Performance

Calibration
Water stage (m) P1 0.987 Excellent
Salinity (g/kg) P1 0.897 Excellent
P2 0.813 Excellent
Validation
Water stage (m) P1 0.790 Excellent
Salinity (g/kg) P1 0.869 Excellent Fig. 8 Maximum seawater intrusion profile for each SLR scenario
(reference scenario. Cubatão River discharge = 16 m3/s)
Impacts of Sea Level Rise on Seawater Intrusion in Cubatão River, Brazil

Fig. 11 Minimum freshwater discharge to maintain S < 0.5 g/kg at the

water abstraction location
Fig. 9 Maximum seawater intrusion profile for each SLR scenario (dry
season. Cubatão River discharge = 8 m3/s)
Similarly, the increasing SLR would increase the discharge
requirements to keep low salinity values at the water abstrac-
Some relevant facts can be posed from the investigated sce-
tion location in Cubatão River. For the reference scenario, for
narios. With the increasing SLR, it would be necessary higher
example, the minimum discharge to keep salinity concentra-
river discharge to maintain the maximum seawater intrusion
tion below 0.5 g/kg is 18 m3/s; it increases to 20 m3/s for a
length compared to the ΔH = 0.0 m scenario. For example,
0.25 m SLR; to 30 m3/s for a 0.50 m SLR; and to 40.0 m3/s for
the maximum seawater intrusion length for a 40 m3/s discharge
1.0 m SLR. Thus, a 1.0-m SLR scenario would increase the
is approximately 3.8 km; in the 1.0 m SLR scenario, the re-
discharge requirements in more than 2 times (Fig. 11).
quired discharge to maintain this seawater intrusion length is
SLR increases the longitudinal salinity gradient, which causes
77.0 m3/s.
higher seawater intrusion in the estuarine system. The increase in
Increased freshwater discharge pushes the limits of salt
Cubatão River discharge required to keep salinity concentration
intrusion towards the river mouth. On the other hand, when
values lower than 0.5 g/kg at the abstraction location due to
the river discharge is decreased because of hydrological pat-
0.25 m, 0.5 m, and 1.0 m SLR is 2 m3/s, 12 m3/s, and 22 m3/s,
terns, e.g., dry periods, the salinity intrusion length increases.
respectively.
A serious outcome of more freshwater required to keep
In a future climate change scenario, in which river dis-
seawater intrusion length in acceptable values under SLR sce-
charge patterns may be strongly affected, water management
narios is that, in future scenarios of climate changes, there will
in Cubatão River may require different strategies and actions
be a considerable variation of precipitation patterns and, con-
to minimize the effects of seawater intrusion enhanced by
sequently, river flows may present unexpected behavior. In
SLR. Those strategies can be (1) water abstraction during high
Cubatão, in which both city and industries concur for water
discharge events, when the salinity concentration at the water
abstraction, this abstraction restriction may pose a major water
abstraction location is reduced; (2) higher freshwater reserva-
management concern.
tion for eventual dry periods when the abstraction can be im-
practicable due to high salinity; (3) selection of future water
abstraction location where the salinity intrusion is negligible
even during low freshwater discharges; (4) higher HB dis-
charges in Cubatão River, which can provoke water resources
management conflicts, because the hydroelectric power plant
and the water treatment facilities are operated by different
companies.

9 Conclusion

The impacts of SLR rise on seawater intrusion in Cubatão

River were studied by applying a 1D hydrodynamic and water
Fig. 10 Variation of the seawater intrusion length versus freshwater
quality numerical model. It was the first investigation on SLR
discharge under SLR scenarios (ΔH = 0.00 m, 0.25 m, 0.50 m, and impacts on seawater intrusion in Cubatão River specifically;
1.00 m) thus, it was a preliminary study which combined a local field
 da Silva F.P. et al.

survey to apply numerical modeling methods. For its survey 2. Bhuiyan, M. J. A. N., & Dutta, D. (2012). Assessing impacts of sea
level rise on river salinity in the Gorai river network, Bangladesh.
limitation, the model considered just the main environmental
Estuarine, Coastal and Shelf Science, 96(1), 219–227. https://doi.
forcing mechanisms: freshwater discharges and tides. org/10.1016/j.ecss.2011.11.005.
The first finding related to the seawater intrusion modeling 3. Alfredini, P., & Arasaki, E. (2013). Engenharia Portuária (2nd
refers to the applied method, in which the estuary was simplified ed.). São Paulo: Edgard Blucher.
4. Rich, V. I., & Maier, R. M. (2015). (2015). Aquatic Environments.
to an idealized mixing zone. A 30-km idealized reach with an
In I. L. Pepper, C. P. Gerba, & T. J. Gentry (Eds.), Environmental
artificial n = 0.06 s/m1/3 worked adequately and the calibration microbiology (3rd. ed., pp. 111–138). London: Elsevier Inc..
and validation procedures delivered excellent results. It showed 5. Vargas, C. I. C., Vaz, N., & Dias, J. M. (2017). An evaluation of
that this adaptation may be tested on different studies in which climate change effects in estuarine salinity patterns: application to
Ria de Aveiro shallow water system. Estuarine, Coastal and Shelf
there is a lack of water quality or bathymetric observed data.
Science, 189, 33–45. https://doi.org/10.1016/j.ecss.2017.03.001.
The model also presented a satisfactory response to SLR 6. Wang, J., Li, L., He, Z., Kalhoro, N. A., & Xu, D. (2019).
scenarios, in which seawater intrusion length increases with Numerical modelling study of seawater intrusion in Indus River
the increasing mean sea level. As many coastal regions, Estuary, Pakistan. Ocean Engineering, 184(November 2018), 74–
84. https://doi.org/10.1016/j.oceaneng.2019.05.029.
Cubatão River water management may face severe restrictions
7. Wolanski, E., Day, J. W., Elliott, M., & Ramachandran, R. (2019).
due to increased seawater intrusion length. In some scenarios, Coasts and estuaries: The future. Orxford: Elsevier.
for example, seawater intrusion length may duplicate for the 8. Martin, J. L., & McCutcheon, S. C. (1999). Hydrodynamics and
same river discharge as a response to SLR. In the 1.0-m SLR transport for water quality modeling. Lewis Publishers.
scenario, the freshwater discharge requirement to maintain the 9. Etemad-Shahidi, A., et al. (2015). Effects of sea level rise on the
salinity of Bahmanshir estuary. International Journal of
current salinity concentration at the water abstraction point Environmental Science and Technology. Springer Berlin
may more than duplicate. Heidelberg, 12(10), 3329–3340. https://doi.org/10.1007/s13762-
For further studies, it may be suggested to validate and 015-0761-x.
apply the modeling method (considering an idealized estuary) 10. Haddout, S., Igouzal, M., & Maslouhi, A. (2016). Analytical and
numerical study of the salinity intrusion in the Sebou river estuary
in different estuarine systems; it may also consider more en- (Morocco) - effect of the “Super Blood Moon” (total lunar eclipse)
vironmental forcing conditions, such as wind stress and the of 2015. Hydrology and Earth System Sciences, 20(9), 3923–3945.
effects of non-astronomical tides (meteorological tides) on the https://doi.org/10.5194/hess-20-3923-2016.
model hydrodynamics. 11. Ji, Z.-G. (2017). Hydrodynamics and water quality. Modeling
Rivers, Lakes and Estuaries. 2nd edn. John Wiley and Sons.
Other studies may make progress on investigating seawater 12. de Miranda, L. B., de Castro, B. M., & Kjerfve, B. (2012).
intrusion scenarios in a numerical model with higher spatial res- Princípios de Oceanografia Física de Estuários (2nd ed.). Edusp:
olution (2D or 3D). It may find significant results on salt wedge Edited by L. B. de Miranda. São Paulo.
dynamics in Cubatão River and establish more precisely the 13. Prandle, D. (2009). Estuaries: Dynamics, Mixing, Sedimentation
and Morphology New York: Cambridge University Press. https://
seawater intrusion length under different forcing conditions. doi.org/10.15713/ins.mmj.3.
Expanded scenarios may be explored, such as the seawater 14. Chua, V. P. & Xu, M. (2014). Impacts of sea-level rise on estuarine
intrusion and salinity stratification response under different HB circulation: an idealized estuary and San Francisco Bay Journal of
operational scenarios, as it has a complex discharge water behav- Marine Systems. Elsevier B.V., 139, pp. 58–67. https://doi.org/10.
1016/j.jmarsys.2014.05.012.
ior in Cubatão River. It may also be relevant to investigate the 15. Liu, W. C., & Liu, H. M. (2014). Assessing the impacts of sea level
seawater intrusion response considering operational scenarios of rise on salinity intrusion and transport time scales in a tidal estuary,
Petrobras dam and how it impacts the salinity behavior. Taiwan. Water (Switzerland), 6(2), 324–344. https://doi.org/10.
3390/w6020324.
Acknowledgments We acknowledge FCTH (Fundação Centro 16. Yang, Z., Wang, T., Voisin, N., & Copping, A. (2015). Estuarine
Tecnológico de Hidráulica) and EMAE (Empresa Metropolitana de response to river flow and sea-level rise under future climate change
Águas e Energia). and human development. Estuarine, Coastal and Shelf Science.
Elsevier Ltd, 156(1), 19–30. https://doi.org/10.1016/j.ecss.2014.
08.015.
Funding Information This study was financed by CAPES (Coordenação 17. Kuang, C., Chen, W., Gu, J., Su, T. C., Song, H., Ma, Y., & Dong,
de Aperfeiçoamento de Pessoal de Nível Superior–Brasil) through the Z. (2017). River discharge contribution to sea-level rise in the
Finance Code 001 and MOMA-SE Project (CAPES-ANA). Yangtze River Estuary, China. Continental Shelf Research.
Elsevier, 134(December 2016), 63–75. https://doi.org/10.1016/j.
csr.2017.01.004.
18. Palmer, K., Watson, C. & Fischer, A. (2019). Non-linear interac-
References tions between sea-level rise, tides, and geomorphic change in the
Tamar estuary, Australia. Estuarine, Coastal and Shelf Science.
1. Jeong, S., Yeon, K., Hur, Y., & Oh, K. (2010). Salinity intrusion Elsevier, 225, p. 106247. https://doi.org/10.1016/j.ecss.2019.
characteristics analysis using EFDC model in the downstream of 106247.
Geum River. Journal of Environmental Sciences. The Research 19. IPCC. (2014). Climate change 2014. Synthesis report. Contribution
Centre for Eco-Environmental Sciences, Chinese Academy of of working groups I, II and III to the fifth assessment report of the
Sciences, 22(6), 934–939. https://doi.org/10.1016/S1001- intergovernmental panel on climate change, climate change 2014:
0742(09)60201-1. synthesis report. Contribution of working groups I, II and III to the
Impacts of Sea Level Rise on Seawater Intrusion in Cubatão River, Brazil

fifth assessment report of the intergovernmental panel on climate 28. Instituto Brasileiro de Geografia e Estatística – IBGE. (2017).
change. Edited by C. W. Team, R. K. Pachauri, and L. A. Meyer. Cubatão. Available at: https://cidades.ibge.gov.br/brasil/sp/
Geneva: IPCC. https://doi.org/10.1017/CBO9781107415324. cubatao/panorama.
20. Chen, W., Chen, K., Kuang, C., Zhu, D. Z., He, L., Mao, X., Liang, 29. Gutberlet, J. (1996). Cubatão. Desenvolvimento, Exclusão Social,
H., & Song, H. (2016). Influence of sea level rise on saline water Degradação Ambiental. São Paulo: Edusp.
intrusion in the Yangtze River Estuary, China. Applied Ocean re- 30. Gonçalves, G. R. (2013). Henry Borden: urbanização e
search, 54, 12–25. https://doi.org/10.1016/j.apor.2015.11.002. industrialização. São Paulo: II Simpósio Internacional
21. James, A. (1993). An Introduction to Water Quality Modelling. 2nd Eletrificação e Modernização Social.
edn. Edited by A. James. Chichester: John Wiley and Sons. 31. ONSET. (2019). HOBO U24–002-C data logger. United States.
22. Chapra, S. (1997). Surface water-quality modeling (2nd ed.). Available at: <https://www.onsetcomp.com/products/data-
McGraw-Hill. loggers/u24-002-c>.
23. Saran, S. H., Rahman, A. and Yunus, A. (2018). Simulation of flow 32. Brasil. (2019). Porto de Santos - Torre Grande (Estado de São
and salinity in Rupsha-Passur River system. Proceedings of the 4th Paulo). Available at: https://www.marinha.mil.br/chm/sites/www.
International Conference on Civil Engineering for Sustainable marinha.mil.br.chm/files/dados_de_mare/santos_torre_grande_
Development (ICCESD 2018). 2019.pdf.
24. Pinto, C. de S. (2005). Cubatão. História de uma cidade industrial. 33. SAISP. (2019). Sistema de Alerta a Inundações de São Paulo. São
Edited by C. de S. Pinto. Cubatão. Paulo. Available at: https://www.saisp.br/estaticos/sitenovo/home.
25. Couto, J. M. (2012). Industrialização, Meio Ambiente e Pobreza. O xmlt.
Caso do Município de Cubatão/SP. Maringá: Editora da 34. Brunner, G. W. (2016). HEC-RAS River analysis system user’s
Universidade Estadual de Maringá. manual. Davis.
26. Fundação Centro Tecnológico de Hidráulica -FCTH. (2017). 35. UNESCO. (1991). Guidelines on the study of seawater intrusion
Controle de inundações das áreas urbanas e industriais situadas into rivers. United Nations Educational, Scientific and Cultural
nas bacias dos Rios Cubatão. São Paulo: Perequê e Mogi. Organization: Edited by V. der H. Tuin. Paris.
27. Harari, J., & Camargo, R. Tides and mean sea level variabilities in
Santos (SP), 1944 to 1989. (1995). Relatório Interno do Instituto Publisher’s Note Springer Nature remains neutral with regard to jurisdic-
Oceanográfico, v. 36, n. February, p. 1–15. tional claims in published maps and institutional affiliations.