Blin 2022

Journal of Hydrology: Regional Studies 39 (2022) 100996
Contents lists available at ScienceDirect
Journal of Hydrology: Regional Studies

journal homepage: www.elsevier.com/locate/ejrh
Potential impacts of climate change on an aquifer in the arid

Altiplano, northern Chile: The case of the protected wetlands of
the Salar del Huasco basin
Nicole Blin a, *, Mark Hausner b, Sarah Leray a, Christopher Lowry c,
Francisco Suárez a, d, e
a
Pontificia Universidad Católica de Chile, Santiago, Chile
b
Desert Research Institute, NV, USA
c
University at Buffalo, NY, USA
d
Centro de Desarrollo Urbano Sustentable (CEDEUS), Santiago, Chile
e
Centro de Excelencia en Geotermia de los Andes (CEGA), Santiago, Chile
A R T I C L E I N F O A B S T R A C T
Keywords: Study region: The Salar del Huasco is an undeveloped basin in the arid Chilean Altiplano.
Climate change Study focus: This study aims to investigate potential impacts of climate change on the aquifer of
Groundwater modeling the Salar del Huasco basin to better manage groundwater resources in the Chilean Altiplano. With
Chilean Altiplano
this purpose, precipitation and temperature projections from multiple general circulation models
Statistical downscaling
Arid regions
were statistically downscaled using the hybrid delta approach to produce ensemble scenarios.
Groundwater recharge Groundwater recharge was estimated for each scenario based on a rainfall-runoff model that was
used to drive a numerical groundwater model, whose results were used to evaluate the future
behavior of the aquifer under each scenario from 2020 to 2100.
New hydrological insights for the study region: This study found that higher precipitation would
contribute to larger groundwater recharge rates. Additionally, results showed that the aquifer is
able to withstand the impacts of climate change through the balance of its water components,
maintaining thus stable levels in areas of ecological value. Since the study site is one of the few
remaining undeveloped groundwater basins, the results obtained represent the natural response
of the system to climate change without interference from anthropogenic factors and thus serve as
a baseline basin that could be used for comparative purposes. This contributes to a better
quantification of aquifer dynamics in the Altiplano, allowing the evaluation of future protective
measures to ensure survival of similar groundwater-dependent ecosystems in the area.
1. Introduction
Groundwater is a vital global resource that provides one third of the planet’s fresh water for domestic, agricultural and industrial
uses (Bloomfield et al., 2019; Taylor et al., 2013; Velis et al., 2017). Predicted climate change poses a threat for groundwater resources,
most of which are non-renewable on meaningful timescales for humans and ecosystems (Kløve et al., 2013). Future projections predict
an increase in the mean global temperatures as well as an increase in the magnitude and frequency of extreme precipitation events
* Corresponding author.
E-mail address: [email protected] (N. Blin).
https://doi.org/10.1016/j.ejrh.2022.100996
Received 21 December 2020; Received in revised form 7 January 2022; Accepted 8 January 2022
Available online 12 January 2022
2214-5818/© 2022 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
N. Blin et al. Journal of Hydrology: Regional Studies 39 (2022) 100996
(IPCC, 2014; Nkhonjera and Dinka, 2017). On one hand, Central Chile (latitude 30ºS) has experienced a drying trend since the
mid-20th century, which is expected to continue in the future, leading to a reduction in average annual precipitation of up to 40%
relative to current values (Garreaud et al., 2020). This region has experienced a significant decrease in rainfall since the late 1970s,
which has been evidenced by more frequent drought events (Rivera et al., 2020). On the other hand, increasing trends have been found
in average annual precipitation in the Bolivian Altiplano (Torres-Batlló and Martí-Cardona, 2020), and in annual precipitation in
northern Chile (DGA, 2017; Souvignet et al., 2012). However, there is still great uncertainty concerning future precipitation changes
across the Andes (Urrutia and Vuille, 2009), while there is general agreement on future increasing temperature trends in the region
(DGA, 2017). Given these projections, understanding the effects of climate change on groundwater becomes of utmost importance.
Different studies have evaluated potential consequences of climate change on groundwater. Decreased snowpack, given a reduction
in snow accumulation due to more precipitation in the form of rain, is likely to reduce river flow in summer months and greater
evaporation rates will increase water demand of crops, thus increasing groundwater extraction and leading to its depletion (Alam et al.,
2019; Amanambu et al., 2020; Nkhonjera and Dinka, 2017; Taylor et al., 2013). Water shortage problems derived from changes in
perception lead to groundwater depletion, which is particularly critical in arid and semi-arid environments where groundwater
discharge dominates the water balance (Rossman et al., 2018; Velis et al., 2017), and becomes even more crucial since population and
groundwater demand in arid regions are expected to increase (Huang et al., 2015; Nkhonjera and Dinka, 2017; Scanlon et al., 2006;
Taylor et al., 2013). Additionally, arid zones are projected to expand to cover up to half of the terrestrial surface by the end of the
century, which will diminish the natural retention (or sequestration) of carbon (Huang et al., 2015; Koutroulis, 2019; Prăvălie, 2016;
Prăvălie et al., 2019).
General circulation models (GCMs) have become the primary tool for the study of projected climate change impacts on hydrological
systems (Persaud et al., 2020). GCMs simulate the behavior and interaction of flow systems in the biosphere, hydrosphere, cryosphere,
atmosphere and geosphere in the climate system (Green et al., 2011; Huang et al., 2015; Scanlon et al., 2006; Taylor et al., 2013).
Climate change scenarios established in the fifth report of the Intergovernmental Panel on Climate Change (IPCC) are expressed in
terms of Representative Concentration Pathways (RCP), which provide quantitative information of the concentration in time of
contaminants associated to climate change. However, the coarse resolution of GCM outputs prevents them from being used directly
with local (basin-scale) models because they do not represent processes taking place at scales finer than their own grids, which are
generally of the order of ~200 km (Fowler et al., 2007; Green et al., 2011; Persaud et al., 2020; Wilby et al., 1998). Therefore,
downscaling techniques are required to bring the GCM’s raw data into the local scale for them to be applicable in basin scale models.
Previous studies have used GCM projections to evaluate potential effects of climate change on hydrogeological systems through
different approaches. Studies by Crosbie et al. (2011, 2013), Hartmann et al., (2017), Hashemi et al. (2015), Holman (2006), McKenna
and Sala (2018), Rossman et al. (2018), Tillman et al. (2016) and Zhang et al. (2019) focused on groundwater recharge or unsaturated
zone processes, but differed in the number of GCMs and scenarios considered for future projections, as well as in the methodology
applied for estimating recharge. Other studies have assessed groundwater recharge uncertainty under future conditions (Goderniaux
et al., 2015; Shen et al., 2018; Xie et al., 2018). Groundwater modeling has been applied to investigate future impacts of climate change
on aquifers and their interaction with surface waters (Persaud et al., 2020; Scibek et al., 2007; Scibek and Allen, 2006; Taheri Tizro
et al., 2019; Usman et al., 2020), and remote sensing has been used to study historical climate change impacts on groundwater re
sources (e.g., Usman et al., 2020; Xu and Su, 2019). Nevertheless, there will always be uncertainty associated with climate models,
particularly in mountainous regions such as the Central Andes where the topography changes abruptly, hindering its correct repre
sentation in models with large grids (Vuille et al., 2008). Therefore, studies that incorporate multiple scenarios to investigate hy
drological and hydrogeological systems in arid catchments with complex topography, such as the Chilean Altiplano, are still scarce.
Recent studies have investigated the hydrological processes occurring in the Andean Chilean altiplano to evaluate groundwater
recharge (Herrera et al., 2016, 2021; Uribe et al., 2015; Urrutia et al., 2019), to assess evaporation (de la Fuente et al., 2021;
Hernández-López et al., 2014; Mosre and Suárez, 2021; Suárez et al., 2020); water table fluctuations (Viguier et al., 2019a, 2019b) and
wetland dynamics (de la Fuente and Meruane, 2017; de la Fuente et al., 2021). Additionally, numerical groundwater modeling has
been used to study the impacts of anthropogenic exploitation (Marazuela et al., 2020; Marazuela et al., 2019b, 2019a; Samuel et al.,
2020; Scheihing and Tröger, 2018; Scheihing et al., 2017). Nonetheless, to the best of the authors’ knowledge, there are no studies that
have used numerical groundwater modeling to study and quantify the impacts of climate change in the Chilean Altiplano, despite this
being essential to manage water resources in a sustainable way and to ensure the survival of groundwater-dependent ecosystems.
The main objective of this research is to develop a methodology that allows a proper management of groundwater resources in the
Chilean Altiplano by: (1) accounting for future uncertainty of climatic variables through ensemble scenarios, and (2) assessing po
tential impacts of climate change on groundwater recharge, groundwater levels, and groundwater discharge fluxes. To achieve this
objective, this research uses statistical downscaling techniques to process GCM outputs, from different RCPs, to the local scale in an
arid region in northern Chile. The downscaled outputs are used as input to a hydrological model for groundwater recharge estimation,
which is then used in a hydrogeological model to simulate the aquifer’s response to future climate change scenarios defined in this
study. Our analysis considers two main time spans: a reference period (1981–2010) against which a future period (2020–2100) is
evaluated. Therefore, even when the incidence of climate is important in long-term aquifer recharge in the central Andes during the
Holocene (Herrera et al., 2018; Sáez et al., 2016), our work focuses on time scales on the order of decades. Furthermore, the focus of
this research is to evaluate potential changes particularly in groundwater resources due to climate change in terms of water table
changes and flow balance of the hydrogeological system. Therefore, even though recharge is a fundamental process related to
groundwater, its particular aspects, such as timing shifts and intersessional behavior of sub-processes taking place in the vadose zone,
are not the main focus of this research.
The Salar del Huasco basin, located in the Chilean altiplano, was chosen as our study site. The basin lies in the presence of
2
groundwater-fed wetlands that sustain the life of unique flora and fauna (de la Fuente et al., 2021; Johnson et al., 2010). The wetlands
are located in the boundaries of the salt flat nucleus, where groundwater springs form a shallow lake of 5–15 cm deep (de la Fuente and
Meruane, 2016; Lobos-Roco et al., 2021). The Salar del Huasco is ecologically important because it is the home of three out of the six
species of flamingos that exist in the world (the Andean flamingo, the Chilean flamingo and the James’s flamingo), which come to the
salt flat to create their nests (CONAF, 2005). It also shelters unique species such as Andean foxes, lamas, vicuñas, pumas, among others.
As a result of this diversity, the salt flat was declared a RAMSAR zone of protected wetlands in 1996 and a National Park by the Chilean
government in 2020. Given the importance of these wetlands in the study area, the achieving of our objective is focused on the
assessment of groundwater changes that could threaten future life in these groundwater-dependent ecosystem due to climate change.
Groundwater resources in the Salar del Huasco basin have had no human intervention, and therefore all results obtained in this study
represent the natural response of the system to the environmental forcing. Consequently, this basin could be considered as a baseline
against which to compare other basins in the area where intervention exists, and thus be able to quantify the effects of such in
terventions in contrast with the natural response of the Salar del Huasco. Additionally, we expect that our results could be further
extended and be a helpful reference for groundwater assessment in similar areas worldwide.
2. Study area
The Salar del Huasco is an endorheic basin located in the Andean plateau of northern Chile (latitude 20.28◦ S and longitude
68.85◦ W) (Fig. 1). The basin extends over an area of 1470 km2 and it has a mean elevation of 4165 m ASL. Its shallow lake is at the end
of a well-developed alluvial system draining from the north of the basin, and constitutes the main water source that sustains life of
unique flora and fauna species under extreme arid conditions (de la Fuente et al., 2020; Suárez et al., 2020).
Fig. 1. (a) General location of the study site; (b) the Salar del Huasco basin (outlined in black) with the location of fluviometric and meteorological
stations, observation wells and springs. Nearby basins are also indicated in (b). Background source: Esri, Maxar, Earthstar Geographics, CNES/
Airbus DS, USDA, USGS, AeroGRID, IGN, and the GIS User Community. A figure with the terrain’s altitude is available in the supplemen
tary material.
3
2.1. Geological settings
The Salar del Huasco basin is bordered to the west by the Altos de Pica plateau, which represents the water divide from Pampa del
Tamarugal (Viguier et al., 2019a, 2019b), and by the Andes mountain range to the east that separates it from the Bolivian basins of
Coipasa and Empexa. The Salar del Huasco basin was formed by volcanic activity that caused the emergence of Mesozoic and Paleozoic
rocks to form the depressed basement area on which alluvial and lacustrine sedimentary materials were deposited as basin fill.
Subsequently, Miocene rhyolitic tuffs with high degrees of welding originated from volcanic eruptions and deposited in the basin,
being the Huasco ignimbrite the most important geological formation due to its extension and volume (Gardeweg and Sellés, 2015;
Tapia et al., 2020). This volcanic activity gave rise to the Andes mountain range, closing the basin to the east, which continued to fill
with volcanic debris and pyroclastic materials (DGA, 2009).
In general, the basin is considered a closed basin with little to no hydrologic connection to the neighboring basins. Previous studies,
using isotope and geochemical information, have discarded a possible connection between the Salar del Huasco and the aquifer of Pica
(Risacher et al., 1999; Uribe et al., 2015). However, Scheihing et al. (2017) discarded this connection arguing that water from Salar del
Huasco would have to travel a long way under the mountain range, passing through geothermal deposits that would raise the tem
perature of water that discharge to Pica. Since that is not the case, they concluded that such connection does not exist. Consequently, in
this study the Salar del Huasco is considered a hydrogeologically closed basin until there is more evidence to confirm otherwise.
2.2. Climate
Two types of climates can be identified in the Salar del Huasco, according to Köppen classification system: the Tundra climate due
to high altitudes of the Andes and the Cold Desert climate. The former is in the northeastern part of the basin while the latter is mainly
in the southwestern part. Mean daily temperatures are around 4–14ºC depending on the altitude, with great thermal oscillation be
tween day and night. In winter months the lowest temperatures reach − 20ºC (DGA, 2009). Evaporation is the main water discharge
mechanism of the basin, with a mean annual potential evaporation of 1200 mm/year (de la Fuente and Meruane, 2016). Precipitation
in this part of the Altiplano occurs in the austral summer (November through March) as heavy convective storms originated by the
intense surface heating that destabilizes the local troposphere and by the establishment of upper-level easterly winds that transport
moist air from the interior of the continent (Sáez et al., 2016, Garreaud, 2009; Garreaud et al., 2003). Interannual precipitation is
highly variable, ranging from 11 to 400 mm/year (Garreaud et al., 2003; Hernández-López et al., 2014). Although marked wet and dry
periods can be identified on the millennial time-scales (Sáez et al., 2016), the El Niño-Southern Oscillation (ENSO) phenomenon is the
major source of the interannual variability in the Altiplano (Garreaud, 2009). It directly affects the coasts of Ecuador, Perú and
northern Chile, causing indirect effects on most of subtropical South America and higher latitudes (Garreaud et al., 2009). Given the
large-scale nature of this phenomenon in terms of its spatial and temporal extents, its assessment is beyond the scope of this study.
However, note that GCMs used to study climate change do include these large-scale phenomena in their formulation (Stocker, 2014).
2.3. Hydrology
The Collacagua river is the main surface watercourse in the basin (Fig. 1). It runs from north to south with daily flow rates that vary
throughout the year in response to precipitation. Despite this, the average monthly flow rates do not vary significantly throughout the
year. The upper section of the river, at the Río Piga en Collacagua fluviometric station, has a daily average flow rate of 0.13 m3 /s while
the downstream section, at the Río Collacagua en Peñablanca fluviometric station, has an average of 0.15 m3 /s. The river between these
two fluviometric stations gains water during most of the year. However, downstream Río Collacagua en Peñablanca, the water from the
river infiltrates as it gets near the salt flat nucleus. The seasonal river flow dynamics are shown in the supplementary material (section
A).
In the topographic lowest point of the basin, a salt flat nucleus that originated due to long-term phreatic evaporation extends over
50–60 km2, the flow paths of mineral-enriched groundwater converge and are discharged to the atmosphere through evaporation
(Corenthal et al., 2016). The Salar del Huasco aquifer discharges mainly into four springs that feed the wetlands located in the borders
of the salt flat nucleus (Fig. 1). Records at gauging stations at these four springs provided by the Chilean National Water Division
(Dirección General de Aguas, DGA) show relatively stable discharges throughout the year, with annual averages ranging from 8 × 10− 3
to 23 × 10− 3 m3 /s for the lowest and highest discharges, respectively (DGA, 2009).
3. Methods
The general methodology followed in this study is based on three major steps. The first step is to downscale temperature and
precipitation outputs from multiple GCMs so they can be used at the basin scale. The projected scenarios used for climate change
impacts assessment are also established in this step to represent different future conditions up to the end of the century. The second step
is to calculate groundwater recharge using the downscaled meteorological data for each projected climate scenario and the hydrologic
model developed by Uribe et al. (2015). In the third step, the groundwater recharge rates are used as inputs in the hydrogeological
model of the Salar del Huasco aquifer. The model then simulates the behavior of groundwater over time under each projected scenario.
4
3.1. Climate projections
Simulations of projected climate scenarios were carried out using the hybrid delta (HD) approach with precipitation and tem
perature offsets based on the Coupled Model Intercomparison Project Phase 5 (CMIP5) (Taylor et al., 2012), multi model ensemble
output. Ensemble models include all four RCPs scenarios. The HD method combines two well-established downscaling techniques: bias
correction and spatial downscaling (BCSD) (Reclamation, 2014) and the delta method (DM) approach (Tohver and Hamlet, 2014).
Here, we briefly describe both techniques and how they are applied in the HD statistical downscaling.
As a first step of the HD approach, the BCSD method is applied. In the BCSD technique, statistical properties of the cumulative
distribution function (CDF) of climatic simulations from the GCMs are adjusted according to the statistical properties of the observed
CDF in a baseline period. Thus, systematic bias of the GCM relative to the observed probability distribution is removed (Tohver and
Hamlet, 2014; Wood et al., 2002).
The second step consists of applying the DM. In the DM, monthly changes of variables projected by the GCMs are determined and
applied to monthly means of observations. To do this, the future is divided into time windows of equal length as the baseline
observational period. For that purpose, the Collacagua station was chosen given it was the only one with enough precipitation and
temperature records, in fact, from 1980 to 2010 there is a complete temperature record, and only one month (February 1992) with no
precipitation records. In this case, 30-year windows were defined according to the available data in the station: 1981–2010 (baseline
period), 2011–2040 (near future), 2041–2070 (mid-future) and 2071–2100 (far future). These windows were defined so there are no
time gaps between the baseline and the distant future. Therefore, the downscaling results are continuously applied in the hydrological
processes. The changes or deltas are calculated for temperature and precipitation, according to the following equations (Camici et al.,
2014; Hamlet et al., 2010; Tohver and Hamlet, 2014):
( )
TDM fut = Tobs + TGCM fut − TGCM base (1)
( )
PGCM fut
PDM fut = Pobs ∙ (2)
PGCM base
where T is temperature and P is precipitation. The subindices in Eqs. (1) and (2) are as follows: obs corresponds to the observed
variables, DM is related to the variables corrected by that method, and GCM base and GCM fut refer to the variables simulated by the
climate models in the baseline period and in the future, respectively. The overbar reflects that the variables are monthly means. For a
step-by-step description of the HD method see Hamlet et al. (2010) and Tohver and Hamlet (2014).
The final step in the HD approach is the temporal disaggregation, from which future daily series of the downscaled variables are
obtained. Daily time series are then used as input to the hydrological model. This temporal disaggregation was performed according to
Tohver et al. (2014) and Vormoor and Skaugen (2013):
( )
Tfut daily = Tobs daily + Tfut monthly − Tobs monthly (3)
( )
Pfut monthly
Pfut daily = Pobs daily ∙ (4)
Pobs monthly
where Tfut daily and Pfut daily are the daily future variables; Tobs daily and Pobs daily are the observed daily temperature and precipitation
data; Tfut monthly and Pfut monthly are the future monthly values obtained by the HD method; and Tobs monthly and Pobs monthly are the observed
monthly temperature and precipitation data.
Consequently, the HD method perturbs the entire CDF, as opposed to the traditional use of the DM that only perturbs the monthly
means of the observed variable (Tohver and Hamlet, 2014). Thereby, future time series of temperature and precipitation represent the
changes projected by the GCM while preserving at the same time its characteristics and behavior (e.g. duration, location, interarrival
time of storms) given by the observed data at a local scale (Hamlet et al., 2010).
In this study, the HD approach was applied to 219 climate models that include emission scenarios RCP2.6, RCP4.5, RCP6.0 and
RCP8.5, to cover a larger range of climate projections and in the absence of a justification for choosing a particular RCP scenario
(Hausner et al., 2014). Subsequently, the 10th, 50th and 90th percentiles of the downscaled variables were calculated to generate
hybrid scenarios of temperature and precipitation that are not subjected to a particular radiative forcing but instead are a compound of
different emission scenarios simulated by different models (Hausner et al., 2016).
Table 1
Future scenarios configuration.
Temperature percentile Precipitation (P) percentile Scenario
Tmin Pmin Extreme-low

Tmax Pmax Extreme-high
Tmed Pmed Median
Tmin Pmax Cross-wet
Tmax Pmin Cross-dry
5
To have a broader spectrum of possible future scenarios, all precipitation and temperature percentiles were combined in the
scenario definition, as shown in Table 1, where the 10th percentile sub index was replaced by “min”, 50th by “med” and 90th by “max”,
for simplicity.
In doing so, the hybrid scenarios take into consideration the large uncertainty that comes along with projected precipitation in the
Altiplano due to the coarse resolution of GCMs that hinders an adequate representation of climatic gradients given by complex to
pographies (Urrutia and Vuille, 2009).
3.2. Groundwater recharge estimations
The semi-distributed hydrological model developed by Uribe et al. (2015) for the Salar del Huasco catchment was used to predict
groundwater recharge under the climate change scenarios previously defined. The model uses as inputs the downscaled mean tem
perature and precipitation. Estimates of reference evaporation (ETo) at a daily basis are also used to obtain groundwater recharge. ETo
(mm/day) was calculated using the FAO-modified Blaney-Criddle (1950) equation (Doorenbos and Pruitt, 1977):
ETo = a + b[c(0.46Tmean + 8.13)] (5)
where a and b are fitting parameters, c is the mean annual percentage of daylight hours, and Tmean is the monthly mean temperature.
The fitting parameters (a and b) were estimated using observed monthly pan evaporation data at Collacagua station in the period
between January 1981 and December 2010, corrected by a pan factor of 0.7 as suggested by DGA (2009). Eq. (5) was selected to
estimate future projections of ETo since it is a simple approach that uses Tmean as an input variable. In climate ensembles, Tmean is more
reliable than other variables that are required for estimating ETo with more sophisticated models, e.g., the Penman-Monteith,
Thornwaite or Hargreaves equations (Allen et al., 2011).
The rainfall-runoff model used in this study was previously developed and validated by Uribe et al. (2015) for the Salar del Huasco
basin and it is based on the Soil Moisture Accounting routine (SMA) of the HEC-HMS model (Feldman, 2000; Hydrologic Engineering
Center US, 2001). For its development, they defined five hydrologic response units (HRU) (Fig. 2b)) according to the hydrogeological
characteristics of the basin, whose connectivity was set using a 90 m resolution digital model elevation (DEM) from the Shuttle Radar
Topography Mission (SRTM). HRU R1 was defined as rocks with very low effective porosity and without fractures, R2 represents
fractured rocks with significant effective porosity; S1 represents lacustrine and evaporite deposits, S2 corresponds to alluvial and
colluvial deposits, and S3 represents fluvial deposits. A set of 15 parameters were defined for representing the hydrological processes at
each HRU, which were calibrated and validated using stream flow records between 1978 and 2010 at station Río Piga en Collacagua.
Here, we provide a brief description of this model. For a detailed description of the model structure the reader is referred to section B of
the supplementary material, and to Uribe et al. (2015) for details on the model configuration (such as HRU definition), the parameters
involved and their calibration, long-term recharge characterization and model performance.
Fig. 2. a) Diagram of the rainfall-runoff model structure that shows a conceptual description of the hydrological processes considered in each
hydrologic response unit (HRU) in the Salar del Huasco basin. In the current work, the groundwater reservoir is represented by the groundwater
model. The variables that are inside the parenthesis indicate that the hydrological process is dependent on these variables, e.g. evapotranspiration
depends on temperature: ET(T). b) Map of the model configuration that shows the five HRUs defined by Uribe et al. (2015) and the 21 subbasins that
contribute to the main drainage network of the basin.
6
The rainfall-runoff model simulates the hydraulic processes occurring at the surface and soil in HRUs through mass balance re
lationships in four reservoirs that represent the different levels of the system at daily time steps (Fig. 2a)). The reservoirs simulate: (1)
canopy storage, (2) surface detention, (3) unsaturated soil processes and (4) the saturated zone. The basic mechanism behind the
balance in each reservoir is that the incoming water can change its storage, evaporate, become runoff, or percolate into the next
reservoir (Uribe et al., 2015). In the canopy storage reservoir, precipitation intercepted by vegetation’s foliage evaporates while the
remaining precipitation that was not captured passes as input to the next reservoir. Since vegetation species that develop in this arid
zone have little or no foliage, canopy storage is almost negligible. Modifications made by Uribe et al. (2015) to adjust the model to the
climatic conditions of the Salar del Huasco consisted in the simulation of freezing at the unsaturated zone by introducing a temporal lag
if the daily mean temperature is below 0 ºC, and the simulation of groundwater routing using a series of linear reservoirs to better
represent the generated hydrograph. The former modification was meant to represent the general behavior of the basin at low tem
peratures. Sublimation or a more detailed representation of snow melting are not included in this model. To justify the use of a
simplified mechanism instead of a more complex representation of snow-related processes, we used Landsat images, acquired from
USGS’s Earth Explorer (earthexplorer.usgs.gov/), to estimate the number of pixels with snow in the months in which it occurs in the
baseline period. To do this, we calculated the Normalized Difference Snow Index (NDSI) and estimated the percentage of area that the
pixels with snow represent of the total basin. We found that the average snow coverage between 1986 and 2010 is of 1.52% of the total
basin and the median value is of 0.46%, which supports the simplification of Uribe et al. (2015) already validated model.
The mass balance in the vadose zone reservoir depends on the soil’s storage capacity, the percolation rate to the next reservoir, and
the meteorological conditions of the day being analyzed. Additionally, a temporal lag in evaporation and infiltration processes is
represented when temperatures are too low. The outputs of the balance in the saturated zone are the deep percolation and the
baseflow, the latter calculated as a series of linear reservoirs to simulate the routing hydrograph through the basin. Groundwater
recharge is defined as the total amount of water entering the saturated zone, which can become deep percolation or baseflow (Uribe
et al., 2015). The usefulness of the saturated zone reservoir is that it allows simulating the processes of deep percolation and baseflow
separately, allowing the latter to be used to calibrate the parameters of the hydrological model.
In the SMA routine of the HEC-HMS model (Hydrologic Engineering Center US, 2001), the rate of percolation between the
soil-profile storage and the saturated zone (i.e. groundwater recharge), depends on the volume in the source and receiving reservoirs. It
is calculated as a function of a maximum percolation rate and the contents of the storage from which percolation is occurring (Bennett
and Peters, 2004). The maximum percolation rate and maximum storage capacity of each reservoir are parameters that are different for
each HRU of the model. Their values are determined after calibrating the model, and future simulations of groundwater recharge are
performed under the assumption that they remain constant in time. Since changes in vegetation would impact soil parameters (due to
root uptake and transpiration), it is assumed that variations in vegetation patches in the basin does not change dramatically in the
future.
Actual evaporation from the subsurface (Ea) is among the calculations performed by the hydrological model to obtain groundwater
recharge. This value is further used to calculate how much water remains from ETo that is still available to be evaporated in the aquifer,
which we define as remnant evaporation (Erem).
Note that recharge resulting from the hydrological model is used as input to the groundwater flow model and is therefore
considered as actual recharge as it reaches the water table (de Silva and Rushton, 2007; Scanlon et al., 2002).
Fig. 3. Configuration of the hydrogeological model developed in Modflow using ModelMuse. The “z” axis shows discretization into layers of three
main aquifers: an upper aquifer made of recent sedimentary deposits and Huasco Ignimbrite, a middle aquifer of clay and Collacagua formation
(sequence of gravel lacustrine deposits, silt, clay and sand (Acosta, 2004)) and a lower aquifer of Collacagua formation and Huasco Ignimbrite. A
geologic fault develops at the bottom of the lower aquifer, represented in the model as a zone of higher hydraulic conductivity.
7
3.3. Groundwater flow model development
The groundwater flow model used to simulate the behavior of the aquifer under each scenario was developed in MODFLOW
(Harbaugh et al., 2000), which is a widely used numerical model that solves the groundwater flow equation through a porous media in
three dimensions, and can represent groundwater and surface water interactions. The main focus of our research is to assess the effects
of climate change by analyzing changes in future simulations with respect a baseline period rather than analyzing absolute values of
potentiometric surfaces, and therefore, effects of variable density were not considered since they are not expected to affect such
changes. Furthermore, given the ecological importance that exists in the surroundings of the wetlands, we gave especial attention to
this area.
The model was built using the open-source interface ModelMuse (Winston, 2009). The geometry of the model consists of 270 rows,
253 columns and 10 layers (Fig. 3) that discretize three main aquifers. The vertical discretization of the three aquifers into 10 layers
was defined to avoid convergence problems associated with cell dimensions, which was done using the uniform spacing option in
ModelMuse (Winston, 2009). No-flow boundary condition were set at the edges of the model according to previous hydrogeological
and geochemical studies (Risacher et al., 1999; Scheihing et al., 2017). The model was run under steady state conditions for the
baseline period and under transient conditions for the future projections. In the latter, monthly stress periods were used from 2020 to
2100. The spatially distributed monthly groundwater recharge projections, obtained from the hydrological model, were incorporated
in the groundwater flow model using MODFLOW’s Recharge package (RCH) (Harbaugh et al., 2000). More specifically, monthly
recharge rates calculated with the rainfall-runoff model (described in Section 3.2) were imported to ModelMuse using a shapefile to
distribute them throughout the model domain, according to the sub-basin spatial distribution Fig. 2b). The Collacagua river was
represented in the groundwater model using the River package (RIV) (Harbaugh et al., 2000). An average river depth of 10 cm and
conductance of 0.06–0.12 m2/s were assigned along the course of the Collacagua river. These values were defined after the steady state
calibration to represent the flux interacting with the aquifer (Acosta, 2004).
The four springs located near the wetlands were represented in the model using the Drain package (DRN) (Harbaugh et al., 2000).
The stage and conductance of each drain were defined and calibrated using DGA’s records of fluviometric stations at the springs, which
are available at www.dga.cl. Groundwater evaporation was modeled with the Evapotranspiration package (EVT) (Harbaugh et al.,
2000), and was defined near the wetlands and streambed of the Collacagua river, where the water table is shallow enough to allow
groundwater evaporation. A potential evaporation of 6.3 mm/day was defined in the steady state simulation, according to pan
evaporation records in Collacagua meteorological station, between April 1994 and May 2007 (Acosta and Custodio, 2008; Uribe,
2012). In the transient simulations, monthly evaporation rates were introduced, corresponding to the simulated Erem, which represents
the water available for evaporation in the aquifer. The extinction depth was set to 3 m beneath the surface, typical for arid saline soils
(Shanafield et al., 2015), and in agreement with ET measurements performed in the basin with the portable chamber method (DGA,
2009; Johnson et al., 2010). To avoid confusion, in the rest of the document we use the term "groundwater evaporation" to refer to this
Fig. 4. Observed and simulated groundwater levels obtained in the steady state calibration. The map on the left shows the level of adjustment of the
simulated heads in wells distributed throughout the basin, where blue circles denote simulated levels lower than observed and red circles denote
simulated levels higher than observed. The size of the circles is proportional to the difference between the observed and simulated level. The chart
on the right shows the adjustment of simulated levels to observations, where the same color notation as in the left map is used, with red and blue
markers that represent simulated levels higher and lower than observed, respectively.
8
discharge, while the term “groundwater recharge” refers to the output of the hydrological model, which includes the evaporation at the
subsurface (i.e., Ea).
3.4. Groundwater flow model calibration and simulations
A steady state simulation was performed to represent the long-term historical (last 40 years) behavior of the aquifer. To ensure the
adequate representation of conceptual model inflows and outflows of the groundwater system according to previous studies (Acosta
and Custodio, 2008; DGA, 2009; Johnson et al., 2010; Uribe et al., 2015), a manual calibration of hydraulic conductivity was per
formed using observed levels at 44 monitoring wells (Fig. 4), using the minimum of the mean absolute error (MAE) as objective
function (Bennett et al., 2013). An ideal model would achieve a MAE of 0, nevertheless, the model is considered accurate if the MAE is
less than 5% of the largest difference of observed data (SEA, 2012). The performance of the model was then evaluated with the root
mean square error (RMSE) and the coefficient of determination (R2 ) indicators (Bennett et al., 2013).
Due to the state of conservation that protects the Salar del Huasco basin from human intervention, there is little data on aquifer
levels over the baseline period, which are required to perform a transient calibration. Thus, hydraulic conductivities were calibrated
through a steady state calibration and storage values in the aquifer were assigned based on previous studies (Acosta and Custodio,
2008; DGA, 2009). Despite not having long time series of groundwater level information over the baseline period, records from 2005
and 2014 (SEA, 2019), along with isolated records from November 2018 (Suárez et al., 2020) were used to validate the model. The
available groundwater levels are presented in section A of the supplementary material.
The groundwater model was used to simulate the aquifer’s behavior in the baseline period: January 1981- December 2010; and in
different future scenarios starting in the near future until the end of the century: January 2020 – December 2100. The results of this
simulation characterized the aquifer behavior in a reference period against which future climate change simulations were compared.
4. Results
4.1. Climate projections
Future precipitation and temperature scenarios (10th percentile: denoted by subscript min, 50th percentile: med, and 90th
percentile: max) projected with the HD approach are compared at a daily basis with observations in the baseline period (Table 2).
Scenarios percentile from min to max are chosen to set guidelines for determining the system sensitivity to climate change. Future
temperature scenarios, in all three-time windows, project a daily increasing trend in maximum, mean and median temperatures in
comparison with the baseline period. Differences can be noted between scenarios though, where projected values increase from Tmin to
Tmax .
Unlike temperatures, projected precipitation behaves differently in each scenario. Substantial differences are observed in the mean
and maximum statistics, where in all three future windows, Pmax ’s mean and maximum exceed those in the baseline period, while in
Pmin the opposite occurs. Increasing and decreasing trends can also be easily distinguished between scenarios at monthly basis (Fig. 5),
where all temperature scenarios project a total increase, while precipitation varies throughout the scenarios defined in Table 1.
4.2. Groundwater recharge
4.2.1. Baseline period

The hydrological model was run first for the baseline period to simulate historical groundwater recharge. The model took
approximately seven years to stabilize, and after this spin-up period groundwater recharge is relatively stable in time, with an esti
mated value of 42.9 ± 27.9 mm/yr (mean ± standard deviation) (Fig. 6) over the model domain. Groundwater recharge varies with
seasons with peaks occurring mostly during summer months, which means that summer precipitation is transformed almost
Table 2
Comparison of projected future variables in the Near, Mid- and far future periods with observations in the baseline period at a daily basis.
Baseline Near future Mid-future Far future
1981–2010 2011–2040 2041–2070 2071–2100
T (◦ C) Obs Tmin Tmed Tmax Tmin Tmed Tmax Tmin Tmed Tmax
minimum -8 -5.50 -5.07 -4.57 -4.90 -4.01 -2.85 -4.61 -3.04 -0.60
maximum 14.1 14.66 15.13 15.90 15.18 16.28 17.70 15.43 17.09 20.09
mean 4.46 5.04 5.44 5.96 5.61 6.44 7.68 5.78 7.16 9.88
median 4.4 4.91 5.31 5.84 5.48 6.32 7.55 5.66 7.01 9.77
standard deviation 3.56 3.35 3.32 3.27 3.34 3.26 3.25 3.30 3.20 3.19
P (mm) Obs Pmin Pmed Pmax Pmin Pmed Pmax Pmin Pmed Pmax
minimum 0 0 0 0 0 0 0 0 0 0
maximum 31 28.5 34.87 65.94 28.5 41.35 81.68 28.5 39.69 87.82
mean 0.36 0.17 0.30 0.53 0.16 0.29 0.58 0.13 0.32 0.64
median 0 0 0 0 0 0 0 0 0 0
standard deviation 1.82 0.96 1.63 3.03 0.92 1.64 3.43 0.78 1.77 3.75
9
Fig. 5. Temporal evolution of temperature, T, (top three charts) and precipitation, P, (bottom three charts) for the minimum, median and maximum
scenario at monthly basis. In all charts, the dashed dark blue line represents the observations between 1981 and 2010 registered in Collacagua
meteorological station.
Fig. 6. Groundwater recharge estimated for the baseline period presented as monthly averages. Monthly observed precipitation data are also
presented in right axis.
immediately into groundwater recharge.
4.2.2. Future projections

Scenarios that project an increase in recharge represent those with the highest precipitation input (Pmax ), corresponding to the
10
cross-wet and extreme-high scenarios (Fig. 7). Both scenarios differ in their temperature inputs, which greatly affects seasonal vari
ability of recharge. Higher temperatures in the extreme-high scenario allow for higher percolation in winter months in comparison
with the cross-wet scenario, reducing the differences between summer and winter recharge as opposed to what occurs in the cross-wet
scenario. However, an overall greater recharge is generated in the cross-wet since the higher temperatures of the extreme-high scenario
produce higher evaporation rates in the surface, reducing the amount of water that infiltrates and transforms into recharge.
Temperature effects can also be observed in scenarios with the lowest precipitation input (Pmin ). As mentioned before, higher
temperatures mean that groundwater recharge does not decrease as much in winter months due to the relatively warmer temperatures.
As a result, the cross-dry scenario has an overall greater recharge in comparison with extreme-low, which is the scenario that projects
the lowest recharge rates.
Table 3 shows projected changes in groundwater recharge for the mid- and far-future periods with respect to the baseline period
(the near-future period was not included in Table 3 since future simulations start in 2020). Positive changes are projected in both
periods in scenarios with increasing recharge, while negative changes are projected otherwise.
Finally, Fig. 8 shows seasonal changes projected per each scenario by the end of the century in terms of monthly mean recharge
throughout the year over the far-future period (2071 – 2100) in comparison with the baseline simulation (blue dashed line). Fig. 8
allows assessing groundwater recharge changes not only in magnitude, but also in terms of its behavior between seasons. Taking that
into account, the groundwater recharge behavior in the cross-wet scenario is the one that most closely resembles that of the baseline,
but with higher magnitudes. Regarding magnitudes, recharge in the median and cross-dry scenario are closest to the baseline in winter
months (May through August), while in summer recharge in the extreme-high scenario is the closest. Also, the effect of temperature in
recharge in winter in scenarios with higher temperatures (cross-dry and extreme-high) mentioned above can also be observed in Fig. 8,
where less variations between winter and summer can be observed.
4.3. Groundwater model
4.3.1. Steady state simulation

The MAE in the steady state simulation was of 6.4 m, which corresponds to 4.2% of the largest observed difference. RMSE and R2
values, of 9.55 m and 0.98 m, respectively, indicate a good model performance, with an overall good fit (simulated levels close to the
1:1 line). The map in Fig. 4 shows the spatial distribution of the calibration performance at each well given by the size of the circular
marker, that increases along with the difference between observations and simulations. It can be thus noted that groundwater levels
Fig. 7. Groundwater recharge estimated for the different scenarios evaluated in this study. The Median scenario (black line) is plotted in both charts
in order to compare median conditions with both cross and extreme scenarios.
11
N. Blin et al.
Table 3
Comparison of groundwater balance components for the mid-future (2041–2070) and far-future (2071–2100) periods with respect to the baseline period (1981–2010) in terms of fluxes percentiles.
Baseline period shows absolute values, while future periods show the projected changes (also known as deltas, estimated as the values obtained in the future scenarios minus the values estimated in the
baseline period of each percentile). The numbers in parenthesis are the percentages of change that each delta represents. Recharge, Groundwater evaporation, River discharge and Spring discharge flows
are expressed in mm/yr over the area of the entire basin.
Recharge (mm/yr) Groundwater evaporation River discharge (mm/yr) Spring discharge (mm/yr)
(mm/yr)
Simulation 25th 50th 75th 25th 50th 75th 25th 50th 75th 25th 50th 75th
Flow rate Baseline 15.30 22.61 34.19 0.0 0.0 14.5 20.3 20.7 21.8 3.0 3.1 3.1
Projected Mid-future
changes Median -4.02 -10.33 -21.53 0.0 0.0 -2.49 -13.28 -13.45 -14.30 -0.21 -0.26 -0.29
(− 26.3%) (− 45.7%) (− 63%) (− 17.2%) (− 65.4%) (− 65.1%) (− 65.6%) (− 7.1%) (− 8.5%) (− 9.2%)
Extreme- 16.2 10.29 0.17 (0.5%) 0.0 0.0 7.59 7.36 (36.2%) 7.23 (35.0%) 6.51 (29.8%) 0.39 0.30 (9.6%) 0.29 (9.4%)
High (105.9%) (45.5%) (52.3%) (12.9%)
Extreme- -9.95 -16.81 -27.19 0.0 0.0 -7.64 -21.32 -21.52 -22.56 -0.39 -0.48 -0.48
12
Low (− 65.0%) (− 74.4%) (− 79.5%) (− 52.7%) (− 105.0%) (− 104.2%) (− 103.4%) (− 12.9%) (− 15.4%) (− 15.4%)
Cross-Dry -3.7 -10.41 -20.19 0.0 0.0 -4.85 -15.59 -15.73 -16.75 -0.28 -0.37 -0.37
(− 24.2%) (− 46.1%) (− 59.0%) (− 33.5%) (− 76.8%) (− 76.1%) (− 76.8%) (− 9.3%) (− 11.9%) (− 12.0%)
Cross-Wet 23.5 18.69 9.31 (27.2%) 0.0 0.0 14.45 16.59 (81.6%) 16.45 (79.6%) 15.73 (72.1%) 0.39 0.32 0.29 (9.4%)
(153.7%) (82.6%) (99.6%) (12.9%) (10.3%)
Far-future
Median -2.10 -8.98 -19.91 0.0 0.0 -1.85 -13.23 -13.37 -14.30 -0.17 -0.26 -0.29
(− 13.7%) (− 39.7%) (− 58.2%) (− 12.8%) (− 65.1%) (− 64.7%) (− 65.6%) (− 5.7%) (− 8.5%) (− 9.2%)
Extreme- 20.27 15.34 5.71 (16.7%) 0.0 0.0 5.87 8.00 (39.4%) 7.87 (38.1%) 7.15 (32.8%) 0.39 0.32 0.30 (9.8%)
Journal of Hydrology: Regional Studies 39 (2022) 100996

High (132.5%) (67.8%) (40.5%) (12.9%) (10.3%)
Extreme- -11.86 -18.61 -29.37 0.0 0.0 -8.93 -21.21 -21.43 -22.45 -0.36 -0.45 -0.47
Low (− 77.5%) (− 82.3%) (− 85.9%) (− 61.6%) (− 104.4%) (− 103.7%) (− 103.0%) (− 12.1%) (− 14.7%) (− 15.1%)
Cross-Dry -4.40 -10.91 -20.69 0.0 0.0 -5.71 -15.38 -15.73 -16.66 -0.28 -0.36 -0.37
(− 28.7%) (− 48.3%) (− 60.5%) (− 39.4%) (− 75.7%) (− 76.1%) (− 76.4%) (− 9.3%) (− 11.5%) (− 12.0%)
Cross-Wet 29.30 26.59 16.91 0.0 0.0 17 (117%) 17.87 (88.0%) 17.74 (85.9%) 17.02 (78.0%) 0.41 0.33 0.32
(191.6%) (117.6%) (49.5%) (13.6%) (10.7%) (10.2%)
Fig. 8. Mean monthly recharge over the far-future period (2071 – 2100). The blue dashed line shows the recharge rate simulated in the baseline
period (1981 – 2010).
are being better represented at lower altitudes in the basin, which includes the area of the wetlands and their surroundings. Using the
groundwater levels registered in nine observation wells in the November 2018 field campaign, a MAE of 9 m (which corresponds to
5.8% of the largest observed difference) and a R2 of 0.90 were obtained.
4.3.2. Baseline period

Water discharge from the springs and the river are the fluxes that remain most stable in time, while evaporation is the discharge
with the largest variation (Fig. 9). The median evaporation rate is null, signifying half of the time water is not being discharged through
this process, but when evaporation occurs its magnitude varies greatly, reaching up to 1.8 m3 /s (38.6 mm/yr). Zero groundwater
evaporation means that the water table is below the extinction depth of evaporation and consequently the aquifer is not discharging by
evaporation in the stress periods when that occurs. Thus, despite the high magnitudes that can be reached by evaporation, its average is
below that of the groundwater recharge.
4.3.3. Future projections

The groundwater model was used to simulate future behavior of the system using the projected recharge for each scenario. The
obtained results are analyzed in terms of mass balance components and equipotential contours.
Groundwater recharge corresponds to the only (external) source of water entering the aquifer. Groundwater-surface water in
teractions vary along the river course, starting from the northern-most elevated alluvial parts of the basin, where the river has a gaining
reach, until the river reaches the lowest elevations near the salt flat nucleus, where it becomes a losing river. However, the net
interaction between the river and the aquifer results in a groundwater discharge towards the river (referred to from here on as river
discharge) in most scenarios.
Changes projected in mass balance components with respect to the baseline period in the mid- and far-future periods are presented
in Table 3, which shows 25th, 50th and 75th percentiles of flow changes. Note that the 25th and 50th percentiles of groundwater
evaporation are zero, which means that most of the time evaporation is not from the aquifer, but it is only occurring at ground surface
both in the baseline and future scenarios. Positive changes in the 75th percentile result from the scenarios that project an increase in
recharge, while negative changes are observed in the ones where recharge decreases (see section C of the supplementary material).
Additionally, despite the slight projected changes reported in Table 3, spring discharge is the most stable flux in both future periods,
with the lowest percentage of change with respect the baseline period. Scenarios with increasing recharge are the ones that project
greater changes in all discharge fluxes, showing the buffering behavior of the basin (Table 3).
Fig. 9. Boxplot of mass balance components of the baseline period simulation (1981–2010). Medians correspond to the red lines. Groundwater
recharge component corresponds to an input of the MODFLOW model estimated with the rainfall-runoff model.
13
Fig. 11 shows the simulated equipotential contours of climate change scenarios by the end of the year 2100, as well as the contours
resulting from the baseline simulation by the end of 2010, in the sector of the wetlands. Looking at the shape of this equipotential line,
it is possible to observe that the 2100 equipotential map of the extreme-high scenario is the one that most closely resembles that of
2010 in the baseline simulation, while the rest of the future maps show changes in the shape of this contour.
5. Discussion
5.1. Groundwater recharge: future projections
Precipitation occurs mainly at higher altitudes and thus infiltration is greater in higher areas (higher subbasins), contributing thus
to the overall increase in the total groundwater recharge. Temperature also plays a role, although minor, on groundwater recharge
since the process of evaporation from the soil depends on it. When higher temperatures are projected at higher elevations (where
temperatures are lowest within the basin) more water is evaporated in the root zone, increasing the space available in the soil reservoir
for the incoming infiltration. This increases the potential infiltration rate, enhancing infiltration in areas where low temperatures
normally limit the process. Nevertheless, the effect of temperature on groundwater recharge is smaller than that of precipitation. More
details on the groundwater recharge produced in higher and lower subbasins and their contribution to the total recharge can be found
in section D of the supplementary material. Other studies have also concluded that warmer temperatures reduce the amount of ground
frost, allowing more water infiltration that leads to increased groundwater recharge in cold regions (Jyrkama and Sykes, 2007;
Kovalevskii, 2007).
Evaporation depends on two aspects: the amount of precipitation entering the system and the ET o defined for each scenario
(calculated with Eq. (5) for the three temperature percentiles). As expected, scenarios with higher precipitation, i.e., cross-wet and
extreme-high, produced the largest Ea rates, since not all precipitation can be transformed into groundwater recharge in the same time
step.
5.2. Groundwater model: steady state simulation
Elevated parts of the basin, as shown in Fig. 4, is where the greatest differences between observed and simulated heads occur. The
steady state simulation underestimates the observed head in these elevated areas. Consequently, river discharge is likely to be
underestimated in the elevated areas.
5.3. Groundwater model: future projections
Groundwater evaporation is the main discharge mechanism that balances recharge inputs, being thus a major contributor in the
Fig. 10. Box plot of mass balance components of (a) groundwater recharge (estimated with the rainfall-runoff model), (b) spring discharge; (c) river
discharge and (d) groundwater evaporation, for the entire simulated period for each future scenario. The blue box indicates where the 50% of the
results are concentrated, the red line is the median, whiskers extend to minimum and maximum values.
14
equilibrium of the basin. Consequently, when recharge increases groundwater evaporation increases due to a higher water table,
eliminating the excess of water entering to the system. However, as shown in Fig. 10d, across all model runs the median evaporation
rate was zero, signifying that during half the simulation period no groundwater evaporation was occurring, but during the other half of
the period groundwater evaporation is occurring with greater variability than the other mass balance discharge components.
Therefore, although groundwater evaporation reaches the highest discharge, half of the time it has a null value. When evaporation is
zero, river discharge is the main mechanism with which the basin stabilizes the recharge input.
Results show that river and spring discharge are strongly related to the amount of recharge entering the system in each scenario
(Fig. 10). Groundwater recharge is higher in the most elevated parts of the basin, where river discharge occurs. Therefore, when
groundwater recharge increases, the aquifer discharges more water into the river to maintain its balance. Consequently, scenarios with
higher groundwater recharge project an increase in the median of this flux in comparison with the baseline period, whereas river
discharge decreases as a result of a reduction in groundwater recharge. However, only in the extreme-low scenario does the river
becomes a water source for groundwater, thus maintaining the balance in the system. Spring discharge shows a similar behavior, but
with much less variation between scenarios. This result implies that groundwater evaporation and river discharge balance ground
water recharge without significant changes in springs discharge, thus protecting this area from potentially big impacts. Table 3 shows
that scenarios projecting greater changes in recharge are the ones that also project greater changes in all discharge fluxes, evidencing
thus the buffering capacity of the groundwater system.
Future changes in the water table at the wetlands are analyzed in detail given the ecological interest of this area (de la Fuente et al.,
2021; Suárez et al., 2020). By analyzing how the shape of the 3780.6 m ASL equipotential line changes between scenarios (Fig. 11),
direct comparisons can be made of their forecasts in terms of wetter or drier conditions in the wetlands. This equipotential line
represents the minimum value of the water table simulated in all scenarios and in the baseline period. Note the surface enclosed by this
equipotential line (referred to herein as water table surface of lowest level) is not necessarily the lake surface since topography might
be higher. In fact, a larger surface enclosed by this contour represents a lowering of the water table across the wetlands leading to a
drier surface condition across the study site. On the contrary, when the surface enclosed by this contour is smaller, the water table is on
average higher across the wetlands leading to wetter surface conditions and an increase in lake levels. The changes in the extension of
this surface will surely affect the size of the wetlands and therefore the fauna and plant communities that depend on groundwater. As
our focus is related to potential impacts of climate change on aquifer dynamics (i.e., impacts on potentiometric surface and water
balance), it is critical to monitor how ecosystems can be impacted as the wetlands size changes, we refer the reader to the work of
others (Fu et al., 2020; Mao et al., 2020; Mu et al., 2020; Peng et al., 2020; Shen et al., 2019; Weise et al., 2020; Wu and Zheng, 2020).
Low precipitation scenarios (Pmin and Pmed ) project an increase in the water table surface of lowest level, i.e. deeper water table,
Fig. 11. Simulated equipotential contours, i.e. contour maps, near the wetlands for the last stress period in the future projections (i.e., December
2100) and the baseline period (December 2010).
15
forecasting slightly drier surface conditions in the wetlands. In comparing these scenarios, the 2100 equipotential map that most
closely resembles that of 2010 is the extreme-high map, followed closely by the cross-wet map. This means the water table in the
aquifer would remain stable (with slight increases) in future scenarios that project 38% increase in precipitation, while decreases in
precipitation of 15% onwards would result in lower future groundwater levels in the Salar del Huasco.
Although there are changes in groundwater levels in the wetlands area, these changes are slight, and the overall behavior of the
spring discharge is stable over time (Fig. 10). These projections are in accordance with the stable behavior of the water levels observed
in the basin between 2000 and 2016 (Fig. A1 in Supplementary material), from which the buffering capacity maintained its balance
despite the climatic changes (i.e. increasing temperature trends mostly). This is consistent with other studies such as that of Cuthbert
et al. (2019), who found that groundwater fluxes in arid regions, and particularly groundwater discharge to wetlands in dry land
scapes, are unaffected by climate change, showing lower response to climate variability than humid regions.
Regarding the groundwater modeling approach, our model does not consider variable density flow. Therefore, in reality, the
absolute value of the potentiometric surface is likely to be affected in the salt flat margin, where a fresh-salt water interface develops
(Marazuela et al., 2019c). However, as this occurs in a system with no anthropogenic perturbation, for the timescales investigated in
this research the projected changes between the baseline and future simulations, which are the focus of assessment in our research,
should not be affected.
As a final remark, model results are subject to the imposed boundary conditions, as well as to the mechanisms defined to represent
the system’s sinks and sources through which water exchange between the aquifer and its surroundings takes place. The quality of the
representation of sinks and sources will depend on the available information in the study area and the objective of the modeling. In this
case the main purpose of the model was to quantify the behavior of the groundwater system in the basin and its potential future
changes. Therefore, further improvements may consider using additional remote sensing resources to fill the scarce available in-situ
information in the study site to better represent such elements that simulate important water balance components such as evaporation,
river infiltration and river discharge.
5.4. Future work
As part of future work regarding future groundwater recharge, it would be valuable to assess not only changes in its magnitude but
also in its interannual variability and potential timing shifts. Additionally, it would be interesting to assess potential impacts of climate
change using the new future scenarios of CMIP6.
6. Conclusions
This study investigated potential impacts of climate change by the end of the century in the aquifer of the Salar del Huasco basin,
located in the high mountains of the Chilean Altiplano, characterized by its arid climatic conditions. The main conclusions of this work
are:
• Future climate scenarios in the Salar del Huasco basin project an increasing trend in mean temperature, while precipitation shows
varying trends between scenarios. However, all scenarios agree that there will be an increase in the magnitude of extreme events.
• Precipitation is the main driver of larger groundwater recharges resulting in higher groundwater evaporation rates. Additionally,
the expected increase in future temperatures would likely lead to recharge with less seasonal variability.
• Groundwater evaporation is the main discharge mechanism that counterbalances the recharge inputs, being thus a key element in
the system’s balance.
• The aquifer responded to the different external conditions represented in each scenario by discharging and storing water
accordingly, which results in different potentiometric levels. In scenarios where a lower groundwater recharge is projected, results
in lower groundwater levels and drier surface conditions near the wetlands; wetter surface conditions are expected when larger
groundwater recharge is projected. However, these changes in the potentiometric surface are almost negligible in the area of the
wetlands, where the groundwater-dependent systems are located, as a result of the balance between recharge and discharge fluxes,
which manages to keep the water table generally stable in this area.
Results from this research demonstrate the Salar del Huasco aquifer would be able to buffer the impacts of climate change by
balancing the inputs with outputs, maintaining the levels in the lake and wetlands, thus ensuring the conservation of unique eco
systems in the Chilean Altiplano. It is important to note, however, that the resilient capacity of the aquifer may be due to its protected
status, and different results could be obtained if anthropogenic factors were interfering with the response of the system. Therefore, the
advantages of being able to carry out this type of study in a protected area of the arid Altiplano are that (1) it could contribute to better
manage the exploitation of other Altiplano groundwater-dependent ecosystems and even to define if any protection measures are
necessary; (2) it could serve as a baseline of systems without anthropogenic action, allowing it to be compared with developed basins
such as those used in the mining industry to distinguish the effects purely associated with climate change and those intensified by
anthropogenic effects; and finally, (3) the results obtained could be generalized to regions of similar characteristics, which would
contribute to a wider knowledge of what will happen in these areas and thus be able to manage groundwater resources more
efficiently.
16
CRediT authorship contribution statement
Nicole Blin: Conceptualization, Methodology, Software, Formal analysis, Visualization, Investigation, Writing – Original Draft,
Writing – Review & Editing, Francisco Suárez: Conceptualization, Methodology, Resources, Visualization, Funding acquisition,
Writing – Original Draft, Writing – Review & Editing, Mark Hausner: Methodology, Resources, Writing – Original Draft, Writing –
Review & Editing, Sarah Leray: Conceptualization and Validation, Writing – Original Draft, Christopher Lowry: Conceptualization
and Validation, Writing – Original Draft.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.
Acknowledgments
We acknowledge funding from the Agencia Nacional de Investigación y Desarrollo (ANID) through grant ANID/FONDECYT/
1210221, as well as the Centro de Desarrollo Urbano Sustentable (CEDEUS – ANID/FONDAP/15110020) and the Centro de Excelencia
en Geotermia de los Andes (CEGA – ANID/FONDAP/15090013). We also thank the contribution of the reviewers (Dr. Alberto Sáez and
an anonymous reviewer) and editors, which greatly improved the presentation of our results.
Appendix A. Supporting information
Supplementary data associated with this article can be found in the online version at doi:10.1016/j.ejrh.2022.100996.
References
Acosta, O., & Custodio, E. (2008). Impactos ambientales de las extracciones de agua subterránea en el Salar del Huasco (norte de Chile). Retrieved from 〈http://
bibliotecadigital.ciren.cl/bitstream/handle/123456789/6525/HUM2–0035.pdf?sequence=1〉.
Acosta, Orlando, 2004. Impactos de las extracciones de agua subterránea en el Salar del Huasco (norte de Chile). Universidad Politéctica de Catalunya.
Alam, S., Gebremichael, M., Li, R., Dozier, J., Lettenmaier, D.P., 2019. Climate change impacts on groundwater storage in the Central Valley, California. Clim. Change
157 (3–4), 387–406. https://doi.org/10.1007/s10584-019-02585-5.
Amanambu, A.C., Obarein, O.A., Mossa, J., Li, L., Ayeni, S.S., Balogun, O., … Ochege, F.U. (2020). Groundwater system and climate change: Present status and future
considerations. https://doi.org/10.1016/j.jhydrol.2020.125163.
Bennett, N.D., Croke, B.F., Guariso, G., Guillaume, J.H., Hamilton, S.H., Jakeman, A.J., Land, C., 2013. Characterising performance of environmental models q.
Environ. Model. Softw. 40, 1–20. https://doi.org/10.1016/j.envsoft.2012.09.011.
Bennett, T.H., & Peters, J.C. (2004). Continuous soil moisture accounting in the Hydrologic Engineering Center Hydrologic Modeling System (HEC-HMS). Joint
Conference on Water Resource Engineering and Water Resources Planning and Management 2000: Building Partnerships, 104. https://doi.org/10.1061/40517
(2000)149.
Bloomfield, J.P., Marchant, B.P., Mckenzie, A.A., 2019. Changes in groundwater drought associated with anthropogenic warming. Hydrol. Earth Syst. Sci. 23,
1393–1408. https://doi.org/10.5194/hess-23-1393-2019.
Camici, S., Brocca, L., Melone, F., Moramarco, T., 2014. Impact of climate change on flood frequency using different climate models and downscaling approaches.
J. Hydrol. Eng. 19 (8), 04014002 https://doi.org/10.1061/(ASCE)HE.1943-5584.0000959.
CONAF. (2005). Ficha informativa de los Humedales de Ramsar (FIR).
Corenthal, L.G., Boutt, D.F., Hynek, S.A., Munk, L.A., 2016. Regional groundwater flow and accumulation of a massive evaporite deposit at the margin of the Chilean
Altiplano. Geophys. Res. Lett. 43 (15), 8017–8025. https://doi.org/10.1002/2016GL070076.
Crosbie, R.S., Dawes, W.R., Charles, S.P., Mpelasoka, F.S., Aryal, S., Barron, O., Summerell, G.K., 2011. Differences in future recharge estimates due to GCMs,
downscaling methods and hydrological models. Geophys. Res. Lett. 38 (11) https://doi.org/10.1029/2011GL047657.
Crosbie, R.S., Scanlon, B.R., Mpelasoka, F.S., Reedy, R.C., Gates, J.B., Zhang, L., 2013. Potential climate change effects on groundwater recharge in the High Plains
Aquifer, USA. Water Resour. Res. 49 (7), 3936–3951. https://doi.org/10.1002/wrcr.20292.
Cuthbert, M.O., Gleeson, T., Moosdorf, N., Befus, K.M., Schneider, A., Hartmann, J., Lehner, B., 2019. Global patterns and dynamics of climate–groundwater
interactions. Nat. Clim. Change 9 (2), 137–141. https://doi.org/10.1038/s41558-018-0386-4.
DGA. (2009). Sistema Piloto I Región: Salar del Huasco. In Levantamieno hidrogeológico para el desarrollo de nuevas fuentes de agua en áreas prioritarias de la zona
norte de Chile, Regiones XV, I, II y III.
DGA. (2017). Actualización del Balance Hídrico Nacional, SIT No 417. Santiago, Chile.
Doorenbos, J, Pruitt, W.O, 1977. Crop water requirements. FAO irrigation and drainage paper 24. Land Water Dev. Div. FAO 144.
Feldman, A.D., 2000. Hydrologic Modeling System HEC-HMS: Technical Reference Manual. US Army Corps of Engineers, Hydrologic Engineering Center,.
Fowler, H.J., Blenkinsop, S., Tebaldi, C., 2007. Linking climate change modelling to impacts studies: recent advances in downscaling techniques for hydrological
modelling. Int. J. Climatol. 27 (12), 1547–1578. https://doi.org/10.1002/joc.1556.
de la Fuente, A., & Meruane, C. (2016). Investigación en el Salar del Huasco: Entendiendo los procesos fundamentales que mantienen a los Salares y que explican la
vida en torno a estos ecosistemas altiplánicos.
de la Fuente, A., Meruane, C., 2017. Spectral model for long-term computation of thermodynamics and potential evaporation in shallow wetlands. Water Resour. Res.
53 (9), 7696–7715. https://doi.org/10.1002/2017WR020515.
de la Fuente, A., Meruane, C., Suárez, F., 2020. Long-term spatiotemporal variability in high Andean wetlands in northern Chile. Sci. Total Environ., 143830 https://
doi.org/10.1016/j.scitotenv.2020.143830.
de la Fuente, A., Meruane, C., Suárez, F., 2021. Long-term spatiotemporal variability in high Andean wetlands in northern Chile. Sci. Total Environ. 756, 143830
https://doi.org/10.1016/j.scitotenv.2020.143830.
Fu, J., Liu, J., Wang, X., Zhang, M., Chen, W., Chen, B., 2020. Ecological risk assessment of wetland vegetation under projected climate scenarios in the Sanjiang Plain,
China. J. Environ. Manag. 273, 111108 https://doi.org/10.1016/j.jenvman.2020.111108.
17
Gardeweg, M., & Sellés, D. (2015, October). Estratigrafía y evolución estructural del área Collacagua-Rinconada, alta cordillera de Iquique, 20 00′ -20 30′ S–Región de
Tarapacá. XIV Congreso Geológico Chileno.
Garreaud, R.D., 2009. The Andes climate and weather. Adv. Geosci. 22, 3–11. https://doi.org/10.5194/adgeo-22-3-2009.
Garreaud, René D., Boisier, J.P., Rondanelli, R., Montecinos, A., Sepúlveda, H.H., Veloso-Aguila, D., 2020. The Central Chile Mega Drought (2010–2018): a climate
dynamics perspective. Int. J. Climatol. 40 (1), 421–439. https://doi.org/10.1002/joc.6219.
Garreaud, R., Vuille, M., Clement, A.C., 2003. The climate of the Altiplano: observed current conditions and mechanisms of past changes. Palaeogeogr. Palaeoclimatol.
Palaeoecol. 194 (1–3), 5–22. https://doi.org/10.1016/S0031-0182(03)00269-4.
Garreaud, René D., Vuille, M., Compagnucci, R., Marengo, J., 2009. Present-day South American climate. Palaeogeogr. Palaeoclimatol. Palaeoecol. 281 (3–4),
180–195. https://doi.org/10.1016/j.palaeo.2007.10.032.
Goderniaux, P., Brouyère, S., Wildemeersch, S., Therrien, R., Dassargues, A., 2015. Uncertainty of climate change impact on groundwater reserves - Application to a
chalk aquifer. J. Hydrol. 528, 108–121. https://doi.org/10.1016/j.jhydrol.2015.06.018.
Green, T., Taniguchi, M., Kooi, H., Gurdak, J., … D. A.-J. of, & 2011, U. (2011). Beneath the surface of global change: Impacts of climate change on groundwater.
Elsevier. Retrieved from: 〈https://www.sciencedirect.com/science/article/pii/S0022169411002988〉.
Hamlet, A., Salathé, E., & Carrasco, P. (2010). Statistical downscaling techniques for global climate model simulations of temperature and precipitation with
application to water resources planning studies. Retrieved from 〈https://digital.lib.washington.edu/researchworks/bitstream/handle/1773/38428/2010–5.pdf?
sequence=1〉.
Harbaugh, A., Banta, E., Hill, M., & McDonald, M. (2000). MODFLOW-2000, The US Geological Survey Modular Ground-Water Model-User Guide to Modularization
Concepts and the Ground-Water Flow Process. In Open-file Report 00–92. wipp.energy.gov. Retrieved from 〈http://wipp.energy.gov/library/CRA/2009_CRA/
references/Others/Harbaugh_Banta_Hill_and_McDonald_2000_MODFLOW_2000_Open_File_Report_00_92.pdf〉.
Hartmann, A., Gleeson, T., Wada, Y., & Wagener, T. (2017). Enhanced groundwater recharge rates and altered recharge sensitivity to climate variability through
subsurface heterogeneity. https://doi.org/10.1073/pnas.1614941114.
Hashemi, H., Uvo, C.B., Berndtsson, R., 2015. Coupled modeling approach to assess climate change impacts on groundwater recharge and adaptation in arid areas.
Hydrol. Earth Syst. Sci. 19, 4165–4181. https://doi.org/10.5194/hess-19-4165-2015.
Hausner, M.B., Wilson, K.P., Gaines, D.B., Suárez, F., Scoppettone, G.G., Tyler, S.W., 2014. Life in a fishbowl: prospects for the endangered Devils Hole pupfish (
Cyprinodon diabolis) in a changing climate. Water Resour. Res. 50 (8), 7020–7034. https://doi.org/10.1002/2014WR015511.
Hausner, M.B., Wilson, K.P., Gaines, D.B., Suárez, F., Scoppettone, G.G., Tyler, S.W., 2016. Projecting the effects of climate change and water management on Devils
Hole pupfish (Cyprinodon diabolis) survival. Ecohydrology 9 (4), 560–573. https://doi.org/10.1002/eco.1656.
Hernández-López, M.F., Gironás, J., Braud, I., Suárez, F., Muñoz, J.F., 2014. Assessment of evaporation and water fluxes in a column of dry saline soil subject to
different water table levels. Hydrol. Process. 28 (10), 3655–3669.
Herrera, C., Custodio, E., Chong, G., Lambán, L.J., Riquelme, R., Wilke, H., Lictevout, E., 2016. Groundwater flow in a closed basin with a saline shallow lake in a
volcanic area: Laguna Tuyajto, northern Chilean Altiplano of the Andes. Sci. Total Environ. 541, 303–318. https://doi.org/10.1016/j.scitotenv.2015.09.060.
Herrera, C., Gamboa, C., Custodio, E., Jordan, T., Godfrey, L., Jódar, J., Sáez, A., 2018. Groundwater origin and recharge in the hyperarid Cordillera de la Costa,
Atacama Desert, northern Chile. Sci. Total Environ. 624, 114–132. https://doi.org/10.1016/j.scitotenv.2017.12.134.
Herrera, C., Godfrey, L., Urrutia, J., Custodio, E., Jordan, T., Jódar, J., Barrenechea, F., 2021. Recharge and residence times of groundwater in hyper arid areas: the
confined aquifer of Calama, Loa River Basin, Atacama Desert, Chile. Sci. Total Environ. 752, 141847 https://doi.org/10.1016/j.scitotenv.2020.141847.
Holman, I.P., 2006. Climate change impacts on groundwater recharge-uncertainty, shortcomings, and the way forward? Hydrogeol. J. 14 (5), 637–647. https://doi.
org/10.1007/s10040-005-0467-0.
Huang, J., Yu, H., Guan, X., Wang, G., Change, R.G.-N.C., 2016, U., 2015. Accelerated dryland expansion under climate change. Nature. 〈https://www.nature.com/
articles/nclimate2837〉.
Hydrologic Engineering Center (US). (2001). The Hydrologic Modeling System (HEC-HMS).
IPCC. (2014). IPCC, 2014: climate change 2014:synthesis report. Contribution of Working Groups I. II and III to the Fifth Assessment Report of intergovernmental
panel on Climate Change. In Team, Core Writing Pachauri, Rajendra K. Meyer, L. A. Geneva, Switzerland.
Johnson, E., Yáñez, J., Ortiz, C., Muñoz, J., 2010. Evaporation from shallow groundwater in closed basins in the Chilean Altiplano. Hydrol. Sci. J. 55 (4), 624–635.
https://doi.org/10.1080/02626661003780458.
Jyrkama, M.I., Sykes, J.F., 2007. The impact of climate change on spatially varying groundwater recharge in the grand river watershed (Ontario). J. Hydrol. 338
(3–4), 237–250. https://doi.org/10.1016/j.jhydrol.2007.02.036.
Kløve, B., Ala-Aho, P., Bertrand, G., Gurdak, J.J., Kupfersberger, H., Kvaerner, J., … Pulido-Velazquez, M. (2013). Climate change impacts on groundwater and
dependent ecosystems. https://doi.org/10.1016/j.jhydrol.2013.06.037.
Koutroulis, A.G., 2019. Dryland changes under different levels of global warming. Sci. Total Environ. 655, 482–511. https://doi.org/10.1016/j.
scitotenv.2018.11.215.
Kovalevskii, V.S., 2007. Effect of climate changes on groundwater. Water Resour. 34 (2), 140–152. https://doi.org/10.1134/S0097807807020042.
Lobos-Roco, F., Hartogensis, O., Vilà-Guerau de Arellano, J., de la Fuente, A., Muñoz, R., Rutllant, J., Suárez, F., 2021. Local evaporation controlled by regional
atmospheric circulation in the Altiplano of the Atacama Desert. Atmos. Chem. Phys. Discuss. 1–38. https://doi.org/10.5194/acp-2020-1300.
Mao, D., Tian, Y., Wang, Z., Jia, M., Du, J., Song, C., 2020. Wetland changes in the Amur River Basin: differing trends and proximate causes on the Chinese and
Russian sides. J. Environ. Manag., 111670 https://doi.org/10.1016/j.jenvman.2020.111670.
Marazuela, M.A., Vázquez-Suñé, E., Ayora, C., García-Gil, A., 2020. Towards more sustainable brine extraction in salt flats: learning from the Salar de Atacama. Sci.
Total Environ. 703, 135605 https://doi.org/10.1016/j.scitotenv.2019.135605.
Marazuela, M.A., Vázquez-Suñé, E., Ayora, C., García-Gil, A., Palma, T., 2019a. Hydrodynamics of salt flat basins: the Salar de Atacama example. Sci. Total Environ.
651, 668–683. https://doi.org/10.1016/J.SCITOTENV.2018.09.190.
Marazuela, M.A., Vázquez-Suñé, E., Ayora, C., García-Gil, A., Palma, T., 2019b. The effect of brine pumping on the natural hydrodynamics of the Salar de Atacama:
the damping capacity of salt flats. Sci. Total Environ. 654, 1118–1131. https://doi.org/10.1016/j.scitotenv.2018.11.196.
Marazuela, M.A., Vázquez-Suñé, E., Ayora, C., García-Gil, A., Palma, T., 2019c. The effect of brine pumping on the natural hydrodynamics of the Salar de Atacama: the
damping capacity of salt flats. Sci. Total Environ. 654, 1118–1131. https://doi.org/10.1016/J.SCITOTENV.2018.11.196.
McKenna, O.P., Sala, O.E., 2018. Groundwater recharge in desert playas: current rates and future effects of climate change. Environ. Res. Lett. 13 (1), 014025 https://
doi.org/10.1088/1748-9326/aa9eb6.
Mosre, J., Suárez, F., 2021. Actual evapotranspiration estimates in arid cold regions using machine learning algorithms with in situ and remote sensing data. Water 13
(6), 870. https://doi.org/10.3390/w13060870.
Mu, S., Li, B., Yao, J., Yang, G., Wan, R., Xu, X., 2020. Monitoring the spatio-temporal dynamics of the wetland vegetation in Poyang Lake by Landsat and MODIS
observations. Sci. Total Environ. 725, 138096 https://doi.org/10.1016/j.scitotenv.2020.138096.
Nkhonjera, G.K., & Dinka, M.O. (2017). Significance of direct and indirect impacts of climate change on groundwater resources in the Olifants River basin: A review.
https://doi.org/10.1016/j.gloplacha.2017.09.011.
Peng, K., Jiang, W., Deng, Y., Liu, Y., Wu, Z., Chen, Z., 2020. Simulating wetland changes under different scenarios based on integrating the random forest and CLUE-S
models: a case study of Wuhan Urban Agglomeration. Ecol. Indic. 117, 106671 https://doi.org/10.1016/j.ecolind.2020.106671.
Persaud, E., Levison, J., MacRitchie, S., Berg, S.J., Erler, A.R., Parker, B., Sudicky, E., 2020. Integrated modelling to assess climate change impacts on groundwater and
surface water in the Great Lakes Basin using diverse climate forcing. J. Hydrol. 584, 124682 https://doi.org/10.1016/j.jhydrol.2020.124682.
Prăvălie, R. (2016). Invited review Drylands extent and environmental issues. A global approach. https://doi.org/10.1016/j.earscirev.2016.08.003.
Prăvălie, R., Bandoc, G., Patriche, C., Sternberg, T., 2019. Recent changes in global drylands: evidences from two major aridity databases. Catena 178, 209–231.
https://doi.org/10.1016/j.catena.2019.03.016.
18
Reclamation. (2014). Downscaled CMIP3 and CMIP5 Hydrology Projections Release of Hydrology Projections, Comparison with Preceding Information, and Summary
of User Needs Bureau of Reclamation Climate Analytics Group Climate Central. Retrieved from 〈http://gdo-dcp.ucllnl.org/downscaled_cmip_projections/〉.
Risacher, F., Alonso, H., & Salazar, C. (1999). Geoquímica de aguas en cuencas cerradas: I, II y III regiones-Chile.
Rivera, J.A., Arnould, G., & Leal, R. (2020). Evaluation of the ability of CMIP6 models to simulate precipitation over Southwestern South America: Climatic features
and long-term trends (1901–2014). https://doi.org/10.1016/j.atmosres.2020.104953.
Rossman, N.R., Zlotnik, V.A., Rowe, C.M., 2018. Using cumulative potential recharge for selection of GCM projections to force regional groundwater models: a
Nebraska Sand Hills example. J. Hydrol. 561, 1105–1114. https://doi.org/10.1016/j.jhydrol.2017.09.019.
Sáez, A., Godfrey, L.V., Herrera, C., Chong, G., Pueyo, J.J., 2016. Timing of wet episodes in Atacama Desert over the last 15 ka. The Groundwater Discharge Deposits
(GWD) from Domeyko Range at 25◦ S. Quat. Sci. Rev. 145, 82–93. https://doi.org/10.1016/j.quascirev.2016.05.036.
Samuel, A., Blin, N., Muñoz, J.F., Suárez, F., 2020. An unsaturated/saturated coupled hydrogeological model for the llamara salt flat, Chile, to investigate Prosopis
tamarugo survival. Geosciences 10 (1). https://doi.org/10.3390/geosciences10010001.
Scanlon, B.R., Healy, R.W., Cook, P.G., 2002. Choosing appropriate technique for quantifying groundwater recharge. Hydrogeol. J. 10, 18–39. https://doi.org/
10.1007/s10040-0010176-2.
Scanlon, B.R., Keese, K.E., Flint, A.L., Flint, L.E., Gaye, C.B., Edmunds, W.M., Simmers, I., 2006. Global synthesis of groundwater recharge in semiarid and arid
regions. Hydrol. Process. 20 (March), 3335–3370. https://doi.org/10.1002/hyp.
Scheihing, K.W., Moya, C.E., Tröger, U., 2017. Insights into Andean slope hydrology: reservoir characteristics of the thermal Pica spring system, Pampa del
Tamarugal, northern Chile. Hydrogeol. J. 25 (6), 1833–1852. https://doi.org/10.1007/s10040-017-1533-0.
Scheihing, K., Tröger, U., 2018. Local climate change induced by groundwater overexploitation in a high Andean arid watershed, Laguna Lagunillas basin, northern
Chile. Hydrogeol. J. 26 (3), 705–719. https://doi.org/10.1007/s10040-017-1647-4.
Scibek, J., Allen, D., Cannon, A., Whitfield, P., 2007. Groundwater–surface water interaction under scenarios of climate change using a high-resolution transient
groundwater model. J. Hydrol. 〈https://www.sciencedirect.com/science/article/pii/S0022169406004069〉.
Scibek, J., Allen, D.M., 2006. Modeled impacts of predicted climate change on recharge and groundwater levels. Water Resour. Res. 42 (11) https://doi.org/10.1029/
2005WR004742.
SEA. (2019). “Desarrollo de Infraestructura y Mejoramiento de Capacidad Productiva de Collahuasi” ADENDA. Elaborado por: SGA.
SEA, S. (2012). Guía para el uso de Modelos de Aguas Subterráneas en el SEIA.
Shanafield, M., Cook, P.G., Gutiérrez-Jurado, H.A., Faux, R., Cleverly, J., Eamus, D., 2015. Field comparison of methods for estimating groundwater discharge by
evaporation and evapotranspiration in an arid-zone playa. J. Hydrol. 527, 1073–1083. https://doi.org/10.1016/j.jhydrol.2015.06.003.
Shen, M., Chen, J., Zhuan, M., Chen, H., Xu, C.Y., Xiong, L., 2018. Estimating uncertainty and its temporal variation related to global climate models in quantifying
climate change impacts on hydrology. J. Hydrol. 556, 10–24. https://doi.org/10.1016/j.jhydrol.2017.11.004.
Shen, G., Yang, X., Jin, Y., Xu, B., Zhou, Q., 2019. Remote sensing and evaluation of the wetland ecological degradation process of the Zoige Plateau Wetland in China.
Ecol. Indic. 104, 48–58. https://doi.org/10.1016/j.ecolind.2019.04.063.
de Silva, C.S., Rushton, K.R., 2007. Groundwater recharge estimation using improved soil moisture balance methodology for a tropical climate with distinct dry
seasons. Hydrol. Sci. J. 52 (5), 1051–1067. https://doi.org/10.1623/hysj.52.5.1051.
Souvignet, M., Oyarzún, R., Verbist, K.M.J., Gaese, H., Heinrich, J., 2012. Hydro-meteorological trends in semi-arid north-central Chile (29-32◦ S): Water resources
implications for a fragile Andean region. Hydrol. Sci. J. 57 (3), 479–495. https://doi.org/10.1080/02626667.2012.665607.
Stocker, T., 2014. Climate Change 2013: The Physical Science Basis: Working Group I Contribution to the Fifth Assessment Report of the Intergovernmental Panel on
Climate Change. Cambridge university press.
Suárez, F., Lobos, F., de la Fuente, A., Vilà-Guerau de Arellano, J., Prieto, A., Meruane, C., Hartogensis, O., 2020. E-DATA: a comprehensive field campaign to
investigate evaporation enhanced by advection in the hyper-arid altiplano. Water 12 (3), 745. https://doi.org/10.3390/w12030745.
Taheri Tizro, A., Fryar, A.E., Pour, M.K., Voudouris, K.S., Mashhadian, M.J., 2019. Groundwater conditions related to climate change in the semi-arid area of western
Iran. Groundw. Sustain. Dev. 9, 100273 https://doi.org/10.1016/j.gsd.2019.100273.
Tapia, J., Schneider, B., Inostroza, M., Álvarez-Amado, F., Luque, J.A., Aguilera, F., Bravo, M., 2020. Naturally elevated arsenic in the Altiplano-Puna, Chile and the
link to recent (Mio-Pliocene to Quaternary) volcanic activity, high crustal thicknesses, and geological structures. J. S. Am. Earth Sci., 102905 https://doi.org/
10.1016/j.jsames.2020.102905.
Taylor, R.G., Scanlon, B., Döll, P., Rodell, M., van Beek, R., Wada, Y., Treidel, H., 2013. Ground water and climate change. Nat. Clim. Change 3 (4), 322–329. https://
doi.org/10.1038/nclimate1744.
Taylor, K.E., Stouffer, R.J., Meehl, G.A., 2012. An overview of CMIP5 and the experiment design. Bull. Am. Meteorol. Soc. 93, 485–498. https://doi.org/10.1175/
BAMS-D-11-00094.1.
Tillman, F.D., Gangopadhyay, S., Pruitt, T., 2016. Changes in groundwater recharge under projected climate in the upper Colorado River basin. Geophys. Res. Lett. 43
(13), 6968–6974. https://doi.org/10.1002/2016GL069714.
Tohver, I., Hamlet, A., 2014. Impacts of 21st-century Climate Change on Hydrologic Extremes in the Pacific Northwest Region of North America. Wiley Online
Library. https://doi.org/10.1111/jawr.12199. Retrieved from.
Torres-Batlló, J., Martí-Cardona, B., 2020. Precipitation trends over the southern Andean Altiplano from 1981 to 2018. J. Hydrol. 590, 125485 https://doi.org/
10.1016/j.jhydrol.2020.125485.
Uribe, J., 2012. Estimación de la recarga del acuífero del Salar del Huasco en base a un modelo de precipitación-escorrentía y una caracterización isotópica. Pontificia
Universidad Católica de Chile.
Uribe, J., Muñoz, J.F., Gironás, J., Oyarzún, R., Aguirre, E., Aravena, R., 2015. Assessing groundwater recharge in an Andean closed basin using isotopic
characterization and a rainfall-runoff model: Salar del Huasco basin, Chile. Hydrogeol. J. 23 (7), 1535–1551. https://doi.org/10.1007/s10040-015-1300-z.
Urrutia, J., Herrera, C., Custodio, E., Jódar, J., Medina, A., 2019. Groundwater recharge and hydrodynamics of complex volcanic aquifers with a shallow saline lake:
Laguna Tuyajto, Andean Cordillera of northern Chile. Sci. Total Environ. 697, 134116 https://doi.org/10.1016/j.scitotenv.2019.134116.
Urrutia, R., Vuille, M., 2009. Climate change projections for the tropical Andes using a regional climate model: temperature and precipitation simulations for the end
of the 21st century. J. Geophys. Res.: Atmos. 114 (D2).
Usman, M., Qamar, M.U., Becker, R., Zaman, M., Conrad, C., Salim, S., 2020. Numerical modelling and remote sensing based approaches for investigating
groundwater dynamics under changing land-use and climate in the agricultural region of Pakistan. J. Hydrol. 581, 124408 https://doi.org/10.1016/j.
jhydrol.2019.124408.
Velis, M., Conti, K.I., Biermann, F., 2017. Groundwater and human development: synergies and trade-offs within the context of the sustainable development goals.
Sustain. Sci. 12 (6), 1007–1017. https://doi.org/10.1007/s11625-017-0490-9.
Viguier, B., Daniele, L., Jourde, H., Leonardi, V., Yáñez, G., 2019a. Changes in the conceptual model of the Pampa del Tamarugal Aquifer: implications for Central
Depression water resources. J. S. Am. Earth Sci. 94, 102217 https://doi.org/10.1016/j.jsames.2019.102217.
Viguier, B., Jourde, H., Leonardi, V., Daniele, L., Batiot-Guilhe, C., Favreau, G., De Montety, V., 2019b. Water table variations in the hyperarid Atacama Desert: role of
the increasing groundwater extraction in the pampa del tamarugal (Northern Chile). J. Arid Environ. 168, 9–16. https://doi.org/10.1016/j.jaridenv.2019.05.007.
Vormoor, K., Skaugen, T., 2013. Temporal disaggregation of daily temperature and precipitation grid data for Norway. J. Hydrometeorol. 14 (3), 989–999. https://
doi.org/10.1175/JHM-D-12-0139.1.
Vuille, M., Francou, B., Wagnon, P., Juen, I., Kaser, G., Mark, B.G., Bradley, R.S., 2008. Climate change and tropical Andean glaciers: past, present and future. Earth-
Sci. Rev. 89 (3–4), 79–96.
Weise, K., Höfer, R., Franke, J., Guelmami, A., Simonson, W., Muro, J., Hilarides, L., 2020. Wetland extent tools for SDG 6.6.1 reporting from the Satellite-based
Wetland Observation Service (SWOS). Remote Sens. Environ. 247, 111892 https://doi.org/10.1016/j.rse.2020.111892.
Wilby, R.L., Wigley, T.M.L., Conway, D., Jones, P.D., Hewitson, B.C., Main, J., Wilks, D.S., 1998. Statistical downscaling of general circulation model output: a
comparison of methods. Water Resour. Res. 34 (11), 2995–3008. https://doi.org/10.1029/98WR02577.
19
Winston, R.B. (2009). ModelMuse: a graphical user interface for MODFLOW-2005 and PHAST.
Wood, A.W., Maurer, E.P., Kumar, A., Lettenmaier, D.P., 2002. Long-range experimental hydrologic forecasting for the eastern United States. J. Geophys. Res. D
Atmos. 107 (20) https://doi.org/10.1029/2001JD000659.
Wu, H. bing, Zheng, B. hui, 2020. Wetland area identification and waterbird protection management in consideration of lake topography and water level change.
Glob. Ecol. Conserv. 23, e01056 https://doi.org/10.1016/j.gecco.2020.e01056.
Xie, Y., Cook, P.G., Simmons, C.T., Partington, D., Crosbie, R., Batelaan, O., 2018. Uncertainty of groundwater recharge estimated from a water and energy balance
model. J. Hydrol. 561, 1081–1093. https://doi.org/10.1016/j.jhydrol.2017.08.010.
Xu, W., Su, X., 2019. Challenges and impacts of climate change and human activities on groundwater-dependent ecosystems in arid areas – a case study of the
Nalenggele alluvial fan in NW China. J. Hydrol. 573, 376–385. https://doi.org/10.1016/j.jhydrol.2019.03.082.
Zhang, H., Wang, B., Liu, D.L., Zhang, M., Feng, P., Cheng, L., Eamus, D., 2019. Impacts of future climate change on water resource availability of eastern Australia: a
case study of the Manning River basin. J. Hydrol. 573, 49–59. https://doi.org/10.1016/j.jhydrol.2019.03.067.
20

Blin 2022

Uploaded by

Copyright:

Available Formats

Blin 2022

Uploaded by

Document Information

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

Blin 2022

Uploaded by

Copyright:

Available Formats

Journal of Hydrology: Regional Studies 39 (2022) 100996

Contents lists available at ScienceDirect

Journal of Hydrology: Regional Studies

Potential impacts of climate change on an aquifer in the arid

2.1. Geological settings

3.1. Climate projections

Tmin Pmin Extreme-low

3.2. Groundwater recharge estimations

3.3. Groundwater flow model development

3.4. Groundwater flow model calibration and simulations

4.1. Climate projections

4.2. Groundwater recharge

4.2.1. Baseline period

immediately into groundwater recharge.

4.2.2. Future projections

4.3. Groundwater model

4.3.1. Steady state simulation

Journal of Hydrology: Regional Studies 39 (2022) 100996

4.3.2. Baseline period

4.3.3. Future projections

5.1. Groundwater recharge: future projections

5.2. Groundwater model: steady state simulation

5.3. Groundwater model: future projections

5.4. Future work

CRediT authorship contribution statement

Declaration of Competing Interest

Appendix A. Supporting information

You might also like