Uncorrected Proof

© 2023 The Author Journal of Hydroinformatics Vol 00 No 0, 1 doi: 10.2166/hydro.2023.062

Satellite-based investigation of water stress at the basin scale: an integrated analysis of

downscaled GRACE estimates and remotely sensed data

Behnam Khorrami
Department of GIS, The Graduate School of Natural and Applied Sciences, Dokuz Eylul University, Doğ uş Cad. 207/A Tınaztepe Yerleşkesi, Buca, Izmir 35390,
Türkiye
E-mail: [email protected]

BK, 0000-0003-3265-372X

ABSTRACT

Freshwater availability is a very determining issue, especially in semiarid and arid regions, for sustainable development and secured food
production. In this premise, the detection and assessment of water stress are of utmost importance. In this study, the satellite-based Poten-
tial Available Water Storage (PAWS) index is used to test its feasibility for a basin-scale analysis of water stress in the Western Mediterranean
Basin (WMB) in Türkiye. The coarse-resolution GRACE (Gravity Recovery And Climate Experiment) estimates were downscaled based on the
Random Forest (RF) model and then were integrated with fine-resolution precipitation data to derive fine-resolution PAWS values. The accu-
racy of the index was validated against the net water flux (NWF) and water storage deficit (WSD) values over the basin. The results revealed a
good performance for the PAWS index for a local scale evaluation of water stress. The PAWS variations turned out to be highly correlated with
the NWF (r ¼ 0.72) and WSD (r ¼ 0.66). The PAWS indicates that the WMB has suffered from a critical hydrological situation from 2003 to 2020
during which the basin has been under stress with the most critical situation in 2018 when the per capita water has fallen below 500 m3
suggesting an absolute water stress status.

Key words: GRACE, potential available water storage, remote sensing, water stress, Western Mediterranean Basin

HIGHLIGHTS

• The PAWS index is capable of revealing water stress status at local scales.
• The Western Mediterranean Basin (WMB) has a critical situation regarding its water resources.
• There is a descending trend in water storage variations of the basin.
• The per capita water availability of the WMB is less than 500 m3 in 2018.

This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY-NC-ND 4.0), which permits copying and
redistribution for non-commercial purposes with no derivatives, provided the original work is properly cited (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Downloaded from http://iwaponline.com/jh/article-pdf/doi/10.2166/hydro.2023.062/1243609/jh2023062.pdf

by [email protected]
 Uncorrected Proof

Journal of Hydroinformatics Vol 00 No 0, 2

GRAPHICAL ABSTRACT

1. INTRODUCTION
Climate change is one of the main challenges of human beings in recent history which invites a real menace to different
aspects of human life as well as nature. It has had perceptible ramifications for the agricultural sector in many regions of
the world which jeopardizes the food safety of the residents (Gregory et al. 2005). Climate change is fundamentally mani-
fested as changes in the established norms of climatic parameters such as precipitation (P), temperature (T ),
evapotranspiration (ET), etc. in a region (McCarty 2001; Dore 2005). The decreasing P and rising T and ET accompanied
by increased population put tremendous pressure on water resources (Richey et al. 2015). The deterioration of the hydrolo-
gical balance between inputs and outputs in a region due to natural and anthropogenic drivers engenders a gap between water
supply and demand which is defined as water scarcity (Hasan et al. 2019). Water stress is regarded as a potential driver of
regional insecurity that can contribute to regional unrest (Richey et al. 2015).
The spatiotemporal dynamics between human and natural water systems can be understood easily through water stress
analysis. To date, different approaches have been followed by researchers to characterize and analyze water stress which
yielded various water stress indicators. Water Resources Availability (Yang et al. 2003); Watershed Sustainability Index
(Chaves & Alipaz 2006); Green–Blue Water Scarcity (Rockström et al. 2009); Water Scarcity Function of Water Footprint
(Hoekstra et al. 2011); Water Use Availability Ration (Alcamo & Henrichs .2002); Local Relative Water Use and Reuse
(Vorosmarty et al. 2005); Water Poverty Index (Lawrence et al. 2002); and Water Supply Stress Index (McNulty et al.
2010). Although these indices have made good strides in assessing water stress status, their optimal use is shackled by
some restrictions mainly due to their inability to integrate all forms of ‘available water’, particularly, soil moisture and ground-
water due to data dearth (Hasan et al. 2019).
To overcome the aforementioned handicap, Hasan et al. (2019) proposed a new water stress index based on the Gravity
Recovery And Climate Experiment (GRACE)’s Terrestrial Water Storage (TWS) estimates and precipitation data from The

Downloaded from http://iwaponline.com/jh/article-pdf/doi/10.2166/hydro.2023.062/1243609/jh2023062.pdf

by [email protected]
 Uncorrected Proof

Journal of Hydroinformatics Vol 00 No 0, 3

Tropical Rainfall Measuring Mission (TRMM). The GRACE-observed TWS is a composite hydrological parameter consisting
of different water cycle components such as surface water, soil moisture, snow water, canopy water, and groundwater (Rodell
et al. 2018). Therefore, it encompasses all forms of water storage and its application for water stress assessment gives a clear
insight into the total water availability in a region (Hasan et al. 2019). In their recent study, Hasan et al. (2019) tested the
feasibility of the Potential Available Water Storage (PAWS) indicator to determine water stress in African countries and
reported good performance. They applied this approach over a large scale (country level) because of two reasons; first, the
coarse resolution (ranging from 200,000 km2 at low latitudes to 90,000 km2 near the poles) (Rodell et al. 2018) of the
GRACE data hinders its applications over smaller scales. Moreover, they validated their results against the water availability
statistics retrieved from the AQUASTAT dataset (a global water information system of the Food and Agriculture Organization
(FAO)) which is available only on country scales. The novelty of this study lies in its application of the PAWS index for a
smaller scale (basin level) which is more required for hydrological analysis and applications. In this premise, the GRACE-
derived TWS was downscaled to 10 km of spatial resolution based on Random Forest (RF) machine learning. The downscaled
TWS later was integrated with satellite precipitation estimates from the CHIRPS (Climate Hazards Group InfraRed Precipi-
tation with Station data) mission to derive the PAWS index.

2. SITE DESCRIPTION
The Western Mediterranean Basin (WMB) is located in the South-West of Türkiye (Figure 1). The surface water flow of the
basin discharges to the Aegean and the Mediterranean Sea. Encompassing four provinces of the country including Antalya,
Burdur, Denizli, and Muğla. The WMB’s surface area is about 21,224 km2 which accounts for about 2.72% of Türkiye’s land
(Çelekli & Lekesiz 2020). The dominant climate of the basin is of the Mediterranean type with dry and hot summer days and
mild and rainy winter days. The mean annual precipitation is approximately 876 mm (Çelekli & Lekesiz 2020). The popu-
lation of the basin, which was approximately 1.2 million in 2000, increased to 2.0 million in 2022. The average population
growth rate of the basin is 2.23%.

Figure 1 | The geographical location of the Western Mediterranean Basin (WMB) in the southwest of Türkiye.

Downloaded from http://iwaponline.com/jh/article-pdf/doi/10.2166/hydro.2023.062/1243609/jh2023062.pdf

by [email protected]
 Uncorrected Proof

Journal of Hydroinformatics Vol 00 No 0, 4

3. METHODOLOGY AND DATA

Several steps were followed in conducting the analysis and evaluation of the results. In the first step, data acquisition and
preprocesing were conducted using TWS from the GRACE satellites, hydro-meteorological data from the Famine Early Warn-
ing Systems Network Land Data Assimilation System (FLDAS) model, and population statistics. In the next step, fine-
resolution TWS data were estimated using the RF model. Then, the PAWS index was calculated. Finally, the results were vali-
dated against P, ET, and water storage deficit (WSD) values. Figure 2 details the link between each step and the procedure
which is detailed as follows.

3.1. GRACE TWS

The GRACE is a twin satellite remote sensing project, working in tandem to gather the gravitational signals of the Earth
(Rodell et al. 2007). The GRACE has been providing the public end-users with data on the monthly anomalies of TWS pro-
cessed by different centers (Khorrami et al. 2022). Mass Concentration (Mascon) blocks are the latest data format of the
mission, which are post-processed and ready to use. Moreover, the Mascon solutions are global and not tailored toward a
particular application, and hence, can be utilized for all scientific topics of interest (Save et al. 2016). Therefore, the latest
release of the GRACE/GRACE-FO JPL Mascon solutions with a spatial resolution of 50 km was utilized in this study. The
TWS data can be downloaded from https://earth.gsfc.nasa.gov/geo/data/grace-mascons.

3.2. FLDAS model

The FLDAS model is a large-scale hydrology model within which two Land Surface Models (LSMs) are used to simulate
different hydrometeorological variables through integrated in situ, modeled, and satellite-derived observations at different res-
olutions (Loeser et al. 2020). The FLDAS makes simulations under the Variable Infiltration Capacity (VIC), and the Noah
models (McNally et al. 2017) with a spatial resolution of 25 and 10 km, respectively. In this study, the different hydro-

Figure 2 | The schematic flow of the study analysis.

Downloaded from http://iwaponline.com/jh/article-pdf/doi/10.2166/hydro.2023.062/1243609/jh2023062.pdf

by [email protected]
 Uncorrected Proof

Journal of Hydroinformatics Vol 00 No 0, 5

meteorological parameters such as P, ET, T, soil moisture storage, snow water storage, and runoff were extracted from the
Noah model of the FLDAS mission via https://disc.gsfc.nasa.gov/datasets?keywords¼FLDAS.

3.3. CHIRPS precipitation

CHIRPS dataset is a relatively new quasi-global, high-resolution, daily/monthly precipitation dataset. The multisource
CHIRPS project combines three different datasets (global climatology, satellite estimates, and in situ observations) to generate
gridded precipitation data. Being known as one of the most potent gridded representatives of the rainfall field (Ray et al.
2022), CHIRPS is often used as an alternative source of precipitation measurements, especially where sufficient measure-
ments are not accessible (Paca et al. 2020). The CHIRPS monthly precipitation data with a resolution of 0.05 degrees can
be downloaded from http://chg.geog.ucsb.edu/data/chirps/.

3.4. MODIS ET
MOD16 is an algorithm used by MODIS (Moderate Resolution Imaging Spectroradiometer) mission to generate high-resol-
ution ET data by integrating MODIS observations and MERRA meteorological data (Souza et al. 2019). In this study, ET data
from the MOD16A2 products with a spatial resolution of 500 m and the temporal resolution of 8 days were used. MOD16A2
ET estimates are freely available at https://modis.gsfc.nasa.gov/data/dataprod/mod16.php.

3.5. Random forest

The coarse spatial resolution of the GRACE data remains a handicap for local studies with a spatial extent of less than the
GRACE resolution (Wang et al. 2019; Khorrami & Gunduz 2021a). Recent developments in downscaling techniques make
GRACE data suitable to assess the local scale variations in water storage (see e.g., Chen et al. 2019; Ali et al. 2021; Arshad
et al. 2022; Yin et al. 2022; Khorrami et al. 2023b). In this study, RF machine learning technique, which is proven to be a very
successful downscaling technique for the GRACE data (Chen et al. 2019), was applied to enhance the spatial footprint of the
GRACE TWS from 50  50 km to 10  10 km over the entire country. The RF benefits from the average values obtained
from decision trees generated based on a set of homogenous subsets of random predictors (Rahaman et al. 2019). The
spatial downscaling is implemented as follows: (i) All the input variables were aggregated to the spatial resolution of
the GRACE-JPL Mascon (50 km), and then the statistical associations between the TWS, geospatial variable
(elevation), and FLDAS-derived variables (soil moisture, snow water, rainfall, surface runoff, ET) at 10 km resolution
were evaluated by developing the RF model to predict TWS. Later, (ii) the residual values were calculated by deducing
the model-derived TWS from the GRACE-derived TWS. Afterward, (iii) the developed model was applied to the hydro-
logical and geospatial variables at a spatial resolution of 10 km to attain the estimated 10 km TWS. Finally, (iv) the
residual correction was performed at 10 km by adding residuals to the estimated TWS at 10 km to obtain the down-
scaled TWS.

3.6. Water storage deficit

WSD expresses the water surplus or deficit in terms of deviations of monthly TWS estimates from monthly climatology values
(Wang et al. 2020). The WSD was calculated by subtracting the climatology (long-running mean monthly) TWS from the
monthly TWS in each corresponding month (Equation (1)) (Khorrami & Gunduz 2021b) where negative WSD values indi-
cate water storage depletion and positive WSD suggests storage surplus.

WSDij ¼ TWSij  TWSj (1)

where TWSij defines the value of TWSA for the month j of the year i. The climatology value of each month is given by TWSj .

3.7. Potential available water storage index

The PAWS is a satellite-derived hydrological index proposed by Hasan et al. (2019) to investigate water stress status at large
scales. The PAWS is calculated as follows:
First, TWS change (DTWS) is estimated as the temporal derivative of the GRACE-observed TWS for 2 consecutive months
(Forootan et al. 2017) (Equation (2)).

DTWSiþ1 ¼ TWSiþ1  TWSi (2)

Downloaded from http://iwaponline.com/jh/article-pdf/doi/10.2166/hydro.2023.062/1243609/jh2023062.pdf

by [email protected]
 Uncorrected Proof

Journal of Hydroinformatics Vol 00 No 0, 6

Second, Internal Water Storage (IWS) is estimated based on the integrated DTWS and P. The IWS represents the water
availability in all forms (i.e. surface and subsurface storage). In this study, P values were used from the CHIRPS dataset.
Since there is a 3-month of temporal lag between TWS and P in the WMB, the monthly IWS was estimated in mm as the
difference between the P estimate of the month (i) and the ΔTWS of the consecutive month (i þ 3) both in mm of water
(Equation (3)).

IWSi ¼ Pi  DTWSiþ3 (3)

And finally, the PAWS was estimated (in m3 per capita) by dividing the basin-average monthly IWS by its population
density (PD) (Equation (4)).

X
1 n12
IWS
n i¼1
PAWS ¼ (4)
PD

The water stress level of the basin based on the PAWS threshold was determined according to the classification given in
Table 1.

4. RESULTS
4.1. Performance of downscaling model
The results of the downscaling model were verified through comparison of the basin-wise variations of the original and pre-
dicted (downscaled) TWS values. According to the results (Figure 3), with a correlation of coefficient (CC) of 0.87, and RMSE
of 2.3 mm, the RF model turned out to be successful in modeling the finer resolution TWS over the study area.

Table 1 | The water stress threshold (Falkenmark & Widstrand 1992)

Threshold/capita Stress level

. 1,700 Occasional or local water stress (no stress)

1,700–1,000 Regular water stress (vulnerable)
1,000–500 Chronic water shortage (stressed)
, 500 Absolute water scarcity (scarcity)

Figure 3 | The multiyear (2003–2020) temporal associations between the GRACE-TWS and downscaled RF-TWS.

Downloaded from http://iwaponline.com/jh/article-pdf/doi/10.2166/hydro.2023.062/1243609/jh2023062.pdf

by [email protected]
 Uncorrected Proof

Journal of Hydroinformatics Vol 00 No 0, 7

4.2. Fluctuations of TWS

To depict the temporal fluctuations of TWS over the WMB, the basin-wise values were extracted from the downscaled TWS.
According to the variations of the mean TWS (Figure 4) negative storage variations are seen from 2007 to 2010, 2013 to 2014,
2016, and 2018 to 2020. The maximum storage loss happened on August 2008 with a 13:80 cm of equivalent water height
which is equal to 2:93 km3 . On the other hand, positive variations in the water storage of the basin are observed from
2003 to 2006, 2011 to 2013, 2015, and 2019. The maximum storage surplus of 21.95 cm occurred in March 2012 which is
equivalent to 4:66 km3 of water storage. Overall, the long-term variations (Jan 2003–Dec 2020) of TWS indicate that the
basin has experienced decreasing water storage variations with an annual storage loss of 0:42 cm equal to 0:09 km3 . The
multiyear trend of the TWS (Figure 5) suggest that the water storage of the basin diminishes from the west to the east.

4.3. Associations between TWS and climatic parameters

As the main hydrological input and output forces, the variations of P and ET over the basin and their associations with the
TWS were investigated. The P and ET values were, respectively, extracted from the CHIRPS and MODIS datasets. The
annual ET and P of the basin were estimated as 8:52 and 17:44 km3 , respectively. To draw a rational analogy between
the variables, the P and ET anomalies were calculated based on the same baseline (mean of Jan 2004–Dec 2009) of the
GRACE mission (Moghim 2020). The results (Figure 6) indicated that the variations of P and ET were associated with
lagged TWS over the basin where there was a correlation of 0:57 between TWS and P and ET.

Figure 4 | The multiyear (2003–2020) temporal fluctuations of TWS.

Figure 5 | Multiyear (2003–2020) annual trend of TWS.

Downloaded from http://iwaponline.com/jh/article-pdf/doi/10.2166/hydro.2023.062/1243609/jh2023062.pdf

by [email protected]
 Uncorrected Proof

Journal of Hydroinformatics Vol 00 No 0, 8

Figure 6 | Multiyear (2003–2020) associations of TWS with P and ET anomalies.

4.4. Variations of IWS and PAWS

The basin-wise variations of the IWS and PAWS are given in Figures 7 and 8, respectively. The variations of IWS over the
WMB are associated with those of TWS with a correlation of 0.59. The annual IWS is also associated with the PAWS
with a correlation of 0.65. In Figure 8, the PAWS time series suggests that the amount of available water per capita is the
minimum in 2018 with 441:35 m3 of available water suggesting absolute water scarcity this year. The maximum water is avail-
able in 2003 with 1729:58 m3 of available water per capita which demonstrates no water stress in the basin.

4.5. Validation
To validate the accuracy of the results, the variations of PAWS were validated against the WSD and the net water flux (NWF)
in the basin. NWF is calculated as precipitation minus ET and describes the effective flux of water between the atmosphere
and the earth’s surface (Khorrami et al. 2023a) and therefore provides important information regarding the interaction of the
atmosphere with the land surface. This parameter is also important from a hydroclimatic perspective because it is a measure
of water stress (Gao & Giorgi 2008). Decreasing NWF signals water scarcity thus diminishing available water in a basin.

Figure 7 | Temporal variations of Internal Water Storage (IWS) and its associations with TWS.

Figure 8 | Temporal variations of PAWS and the annual IWS.

Downloaded from http://iwaponline.com/jh/article-pdf/doi/10.2166/hydro.2023.062/1243609/jh2023062.pdf

by [email protected]
 Uncorrected Proof

Journal of Hydroinformatics Vol 00 No 0, 9

Figure 9 | Associations among the annual net water flux (NWF), Water Storage Deficit (WSD), and Potential Available Water Storage (PAWS).

Figure 10 | Water stress of the WMB according to the long-term (2003–2020) mean PAWS.

Downloaded from http://iwaponline.com/jh/article-pdf/doi/10.2166/hydro.2023.062/1243609/jh2023062.pdf

by [email protected]
 Uncorrected Proof

Journal of Hydroinformatics Vol 00 No 0, 10

Therefore, higher NWF corresponds to a condition where water is more available for other compartments of the hydrological
cycle. In this study, the annual ET was subtracted from the annual P to estimate the NWF over the study area.
The results (Figure 9) demonstrate a high association between the PAWS index variations and those of NWF and WSD
from 2003 to 2020 with a correlation of 0.72 and 0.66, respectively. The high interactions between PAWS and NWF demon-
strate the great influence of the climatic situation on the availability of water during the study period in the basin. Where
increasing NWF implies more water availability and thus more freshwater accessibility for use and vice versa. The WSD vari-
ations also indicate that the basin has experienced a harsh hydrological situation with negative WSD in most of the years
suggesting the WSD. The highest storage deficit of the basin is seen in 2008 with a 1:79 km3 water storage.

4.6. Spatial evolution of water stress

The spatial status of the water stress is given in Figure 10. According to the results, it can be stated that the basin has had an
overall water scarcity from 2003 to 2020. Excluding 2003 and 2004 for which partial stressless areas detected over the basin,
during the remaining years the basin has been stressed out hydrologically. The situation is more exacerbated in 2016 and 2018
when the WMB suffered from absolute water scarcity, especially in 2018 when the majority of the basin was stressed out.

5. DISCUSSIONS
The WMB is one of the most susceptible basins to drought in Türkiye (Özüpekçe 2021). The recent water stress status of the
basin was investigated based on the PAWS index. According to the results, The WMB has suffered from diminishing water
storage from 2003 to 2020 with a storage loss of 0.09 km3 per annum which is equal to 1.62 km3 of total water storage loss
during the last 18 years. Since the GRACE signals are more sensitive to agricultural and hydrological droughts (Okay Ahi &
Jin 2019), the results can be interpreted taking the drought analysis results into account. The critical storage loss of the basin
in 2007, 2008 2018, and 2020 can be ascribed to the harsh country-wise drought incidents (Türkeş et al. 2009; Kurnaz 2014;
Özüpekçe 2021; Khorrami & Gündüz 2022). The hydrologic status of the WMB has been in particular very critical in 2008
during which the basin lost 1.79 km3 of its water storage. This is justified by taking the drought analysis results of the region
into consideration. Özüpekçe (2021) reported extremely dry conditions for the basin in 2008 and stated that the wetlands of
the basin had been in peril of complete disappearance due to the drought side effects as well as anthropogenic impacts.
The basin-wise PAWS results reveal that except in 2003, the WMB has had a critical status with a per capita available water
of less than 1,200 m3. Since the water stress threshold is 1,700 m3, it can be inferred that the basin has been under great
pressure from 2004 onward. Although the GRACE estimates reflect the impacts of both natural and anthropogenic forces
on the variations of water storage (Mandea et al. 2020), in light of the high associations between the PAWS and NWF
over the basin, it can be inferred that the variations of water storage and thus water stress in this basin are fundamentally
ascribed to climate change consequences.

4. CONCLUSIONS
The prevalent impacts of climate change and population increase have culminated to the detriment of the hydrological bal-
ance in many regions of the world. As a result, water resources were put under exorbitant pressure challenging the availability
of freshwater, especially in the regions with semiarid to arid climates. Since water availability plays a crucial role in sustaining
the world food supply, and socio-economic development, investigation of water availability and analysis of water stress are
very important. Remote sensing observations provide a good alternative for the traditionally used field data with their
global coverage, affordability, and acceptable accuracy. The PAWS is a new index developed based on the GRACE estimates
and satellite precipitation data to investigate water stress. In this study, downscaled GRACE data and fine-resolution precipi-
tation from the CHIRPS mission were integrated to evaluate water stress at a local scale. The results suggested a very good
performance of the index for a local scale analysis. The PAWS agrees well with the NWF and WSD in the basin suggesting the
reliability of the index for basin-wise assessments of water stress.

SOURCES OF FUNDING
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

DATA AVAILABILITY STATEMENT

All relevant data are available from an online repository or repositories.

Downloaded from http://iwaponline.com/jh/article-pdf/doi/10.2166/hydro.2023.062/1243609/jh2023062.pdf

by [email protected]
 Uncorrected Proof

Journal of Hydroinformatics Vol 00 No 0, 11

CONFLICT OF INTEREST
The authors declare there is no conflict.

REFERENCES

Alcamo, J. & Henrichs, T. 2002 Critical regions: a model-based estimation of world water resources sensitive to global changes. Aquatic
Sciences 64, 352–362.
Ali, S., Liu, D., Fu, Q., Cheema, M. J. M., Pham, Q. B. & Rahaman, M. et al. 2021 Improving the resolution of GRACE data for spatio-
temporal groundwater storage assessment. Remote Sensing 13 (17), 3513. https://doi.org/10.3390/rs13173513.
Ali, S., Khorrami, B., Jehanzaib, M., Arshad, A., Ajmal, M., Shafeeque, M., Dilawar, A., Basit, I., Zhang, L. L., Tariq, M. A. U. & Niaz, M. A.
2023 Spatial downscaling of GRACE data for improved hydrological droughts in the Indus Basin Irrigation System (IBIS). Remote
Sensing 15 (4), 873. https://doi.org/10.3390/rs15040873.
Arshad, A., Mirchi, A., Samimi, M. & Ahmad, B. 2022 Combining downscaled-GRACE data with SWAT to improve the estimation of
groundwater storage and depletion variations in the irrigated Indus basin. Science of the Total Environment 156044. https://doi.org/10.
1016/j.scitotenv.2022.156044.
Çelekli, A. & Lekesiz, Ö. 2020 Eco-assessment of West Mediterranean basin’s rivers (Turkey) using diatom metrics and multivariate
approaches. Environmental Science and Pollution Research 27 (22), 27796–27806.
Chaves, H. M. L. & Alipaz, S. 2006 An integrated indicator based on basin hydrology, environment, life, and policy: the watershed
sustainability index. Water Resources Management 21, 883–895.
Chen, L., He, Q., Liu, K., Li, J. & Jing, C. 2019 Downscaling of grace-derived groundwater storage based on the random forest model. Remote
Sensing 11 (24), 1–19. https://doi.org/10.3390/rs11242979.
Dore, M. H. 2005 Climate change and changes in global precipitation patterns: what do we know? Environment International 31 (8),
1167–1181.
Falkenmark, M. & Widstrand, C. 1992 Population and water resources: a delicate balance. Population Bulletin 47, 1–36.
Forootan, E., Safari, A., Mostafaie, A., Schumacher, M., Delavar, M. & Awange, J. L. 2017 Large-scale total water storage and water flux
changes over the arid and semiarid parts of the Middle East from GRACE and reanalysis products. Surveys in Geophysics 38, 591–615.
Gao, X. & Giorgi, F. 2008 Increased aridity in the Mediterranean region under greenhouse gas forcing estimated from high resolution
simulations with a regional climate model. Global and Planetary Change 62 (3–4), 195–209. http://dx.doi.org/10.1016/j.gloplacha.2008.
02.002.
Gregory, P. J., Ingram, J. S. & Brklacich, M. 2005 Climate change and food security. Philosophical Transactions of the Royal Society B:
Biological Sciences 360 (1463), 2139–2148.
Hoekstra, A. Y., Chapagain, A. K., Aldaya, M. M. & Mekonnen, M. M. 2011 The Water Footprint Assessment Manual: Setting the Global
Standard. Earthscan, London, DC, USA.
Khorrami, B. & Gunduz, O. 2021a Evaluation of the temporal variations of groundwater storage and its interactions with climatic variables
using GRACE data and hydrological models: a study from Turkey. Hydrological Processes 35 (3), e14076. https://doi.org/10.1002/hyp.
14076.
Khorrami, B. & Gunduz, O. 2021b An enhanced water storage deficit index (EWSDI) for drought detection using GRACE gravity estimates.
Journal of Hydrology 603, 126812. https://doi.org/10.1016/j.jhydrol.2021.126812.
Khorrami, B. & Gündüz, O. 2022 Detection and analysis of drought over Turkey with remote sensing and model-based drought indices.
Geocarto International 1–23. https://doi.org/10.1080/10106049.2022.2066197.
Khorrami, B., Fistikoglu, O. & Gunduz, O. 2022 A systematic assessment of flooding potential in a semi-arid basin using GRACE gravity
estimates and large-scale hydrological modeling. Geocarto International 1–22. https://doi.org/10.1080/10106049.2022.2045365.
Khorrami, B., Ali, S., Sahin, O. G. & Gunduz, O. 2023a Model-Coupled GRACE-Based analysis of hydrological dynamics of drying lake
Urmia and its basin. Hydrological Processes e14893. https://doi.org/10.1002/hyp.14893.
Khorrami, B., Ali, S. & Gündüz, O. 2023b Investigating the local-scale fluctuations of groundwater storage by using downscaled GRACE/
GRACE-FO JPL mascon product based on Machine Learning (ML) algorithm. Water Resources Management. https://doi.org/10.1007/
s11269-023-03509-w.
Kurnaz, L. 2014 Drought in Turkey. _Istanbul Policy Center, Sabancı Universitesi_Istanbul. Available from: https://ipc.sabanciuniv.edu/
Content/Images/CKeditorImages/20200323-16034017.pdf.
Lawrence, P., Meigh, J. & Sullivan, C. 2002 The Water Poverty Index: An International Comparison. Department of Economics, Keele
University, Keele, UK.
Loeser, C., Rui, H., Teng, W. L., Ostrenga, D. M., Wei, J. C., Mcnally, A. L. & … Meyer, D. J. 2020 Famine early warning systems network
(FEWS NET) land data assimilation system (LDAS) and other assimilated hydrological data at NASA GES DISC. In AMeriç an
Meteorological Society Annual Meeting (No. GSFC-E-DAA-TN77212).
Mandea, M., Dehant, V. & Cazenave, A. 2020 GRACE – gravity data for understanding the deep earth’s interior. Remote Sensing 12 (24),
4186.
McCarty, J. P. 2001 Ecological consequences of recent climate change. Conservation Biology 15 (2), 320–331.

Downloaded from http://iwaponline.com/jh/article-pdf/doi/10.2166/hydro.2023.062/1243609/jh2023062.pdf

by [email protected]
 Uncorrected Proof

Journal of Hydroinformatics Vol 00 No 0, 12

McNulty, S., Sun, G., Myers, J. M., Cohen, E. & Caldwell, P. 2010 Robbing peter to pay paul: tradeoffs between ecosystem carbon
sequestration and water yield. In: Proceedings of the Environmental Water Resources Institute Meeting, 23–27 August 2010, Madison,
WI, USA.
McNally, A., Arsenault, K., Kumar, S., Shukla, S., Peterson, P. & Wang, S. et al. 2017 A land data assimilation system for sub-Saharan Africa
food and water security applications. Scientific Data 4, 170012. https://doi.org/10.1038/sdata.2017.12.
Moghim, S. 2020 Assessment of water storage changes using GRACE and GLDAS. Water Resources Management 34 (2), 685–697.
Okay Ahi, G. & Jin, S. 2019 Hydrologic mass changes and their implications in Mediterranean-climate Turkey from GRACE measurements.
Remote Sensing 11 (2), 120. https://doi.org/10.3390/rs11020120.
Özüpekçe, S. 2021 Drought analysis and relationship with water resources of western Mediterranean basins closed area. International
Journal of Geography and Geography Education (IGGE) 43, 317–337.
Paca, V. H. D. M., Espinoza-Dávalos, G. E., Moreira, D. M. & Comair, G. 2020 Variability of trends in precipitation across the Amazon river
basin determined from the CHIRPS precipitation product and from station records. Water 12 (5), 1244. https://doi.org/10.3390/
w12051244.
Rahaman, M. M., Thakur, B., Kalra, A., Li, R. & Maheshwari, P. 2019 Estimating high-resolution groundwater storage from GRACE: a
random forest approach. Environments 6 (6), 63. https://doi.org/10.3390/environments6060063.
Ray, R. L., Sishodia, R. P. & Tefera, G. W. 2022 Evaluation of gridded precipitation data for hydrologic modeling in north-central Texas.
Remote Sensing 14 (16), 3860. https://doi.org/10.3390/rs14163860.
Richey, A. S., Thomas, B. F., Lo, M. H., Reager, J. T., Famiglietti, J. S., Voss, K. & … Rodell, M. 2015 Quantifying renewable groundwater
stress with GRACE. Water Resources Research 51 (7), 5217–5238.
Rockström, J., Falkenmark, M., Karlberg, L., Hoff, H., Rost, S. & Gerten, D. 2009 Future water availability for global food production: the
potential of green water for increasing resilience to global change. Water Resources Research 45.
Rodell, M., Chen, J., Kato, H., Famiglietti, J. S., Nigro, J. & Wilson, C. R. 2007 Estimating groundwater storage changes in the Mississippi River
basin (USA) using GRACE. Hydrogeology Journal 15 (1), 159–166. https://doi.org/10.1007/s10040-006-0103-7.
Rodell, M., Famiglietti, J. S., Wiese, D. N., Reager, J. T., Beaudoing, H. K., Landerer, F. W. & Lo, M. H. 2018 Emerging trends in global
freshwater availability. Nature 557 (7707), 651–659.
Save, H., Bettadpur, S. & Tapley, B. D. 2016 High-resolution CSR GRACE RL05 mascons. Journal of Geophysical Research: Solid Earth 121
(10), 7547–7569. https://doi.org/10.1002/2016JB013007.
Souza, V. D. A., Roberti, D. R., Ruhoff, A. L., Zimmer, T., Adamatti, D. S., Gonçalves, L. G. G. D. & … Moraes, O. L. D. 2019 Evaluation of
MOD16 algorithm over irrigated rice paddy using flux tower measurements in Southern Brazil. Water 11 (9), 1911. https://doi.org/
10.3390/w11091911.
Türkeş, M., Koc, T. & Saris, F. 2009 Spatiotemporal variability of precipitation total series over Turkey. International Journal of Climatology
29 (8), 1056–1074.
Vorosmarty, C. J., Douglas, E. M., Green, P. A. & Revenga, C. 2005 Geospatial indicators of emerging water stress: an application to Africa.
Ambio (R. Sweedish Acad. Sci.) 34, 230–236.
Wang, L., Kaban, M. K., Thomas, M., Chen, C. & Ma, X. 2019 The challenge of spatial resolutions for GRACE-based estimates volume
changes of larger man-made lake: the case of China’s three gorges reservoir in the Yangtze river. Remote Sensing 11 (1), 99. https://doi.
org/10.3390/rs11010099.
Wang, J., Chen, Y., Wang, Z. & Shang, P. 2020 Drought evaluation over Yangtze River basin based on weighted water storage deficit. Journal
of Hydrology 591, 125283. doi: 10.1016/j.jhydrol.2020.125283.
Yang, H., Reichert, P., Abbaspour, K. C. & Zehnder, A. J. B. 2003 A water resources threshold and Its implications for food security.
Environmental Science & Technology 37, 3048–3054.
Yin, W., Zhang, G., Liu, F., Zhang, D., Zhang, X. & Chen, S. 2022 Improving the spatial resolution of GRACE-based groundwater storage
estimates using a machine learning algorithm and hydrological model. Hydrogeology Journal 30 (3), 947–963. https://doi.org/10.1007/
s10040-021-02447-4.

First received 14 March 2023; accepted in revised form 15 June 2023. Available online 27 June 2023

Downloaded from http://iwaponline.com/jh/article-pdf/doi/10.2166/hydro.2023.062/1243609/jh2023062.pdf

by [email protected]