Terrestrial Water Storage Research Articles

Investigation of 2020–2022 extreme floods and droughts in Sichuan Province of China based on joint inversion of GNSS and GRACE/GFO data

The Global Navigation Satellite System (GNSS) surface deformation measurement and Gravity Recovery and Climate Experiment (GRACE)/GRACE Follow-on (GFO) satellite gravimetry provide complementary strengths in estimating terrestrial water storage (TWS) changes, which helps quantify the evolutions of extreme floods and droughts. We implemented a joint inversion framework that only incorporates the contributions of GNSS and GRACE/GFO to recover the TWS changes for monitoring extreme floods and droughts in the Sichuan Province of China (SCP). The performance of TWS changes from the joint inversion is evaluated through closed-loop simulation and independent hydrometeorological data (i.e., precipitation (P), evapotranspiration (ET), and runoff (R)) based on the water balance equation (named P-ET-R) in the spatio-temporal domains. The simulation analysis results show that joint inversion presents better performance in recovering TWS changes than GNSS-only and GRACE/GFO-only solutions in the spatio-temporal scales. The measured joint inversion results reveal more spatial details of TWS changes than GNSS-only and GRACE/GFO-only solutions and present better correlations with GLDAS- and ERA5-derived P-ET-R (0.77 vs. 0.79) than GNSS-only (0.49 vs. 0.54) and GRACE/GFO-only (0.61 vs. 0.68) in the spatial domain. Meanwhile, the joint inversion results effectively characterize the spatio-temporal dynamics of extreme floods and droughts in SCP during 2020–2022, the average TWS surpluses are 33.61 and 32.86 mm in 2020 and 2022 flood events and the average TWS deficits are −36.78 and −63.87 mm in 2021 and 2022 drought events. In addition, the drought propagation time generally presents an increasing trend from southwestern to northeastern SCP. Furthermore, the average and minimum recovery time of drought and flood in the Sichuan basin is longer than those in the western mountains during 2020–2022. The proposed approach provides an effective estimation strategy for recovering reliable TWS changes for the investigation of extreme floods and droughts in regions where GNSS stations are distributed.

Global and regional ocean mass budget closure since 2003

In recent sea level studies, discrepancies have arisen in ocean mass observations obtained from the Gravity Recovery and Climate Experiment and its successor, GRACE Follow-On, with GRACE estimates consistently appearing lower than density-corrected ocean volume observations since 2015. These disparities have raised concerns about potential systematic biases in sea-level observations, with significant implications for our understanding of this essential climate variable. Here, we reconstruct the global and regional ocean mass change through models of ice and water mass changes on land and find that it closely aligns with both GRACE and density-corrected ocean volume observations after implementing recent adjustments to the wet troposphere correction and halosteric sea level. While natural variability in terrestrial water storage is important on interannual timescales, we find that the net increase in ocean mass over 20 years can be almost entirely attributed to ice wastage and human management of water resources.

Detection of extreme hydrological droughts in the poyang lake basin during 2021–2022 using GNSS-derived daily terrestrial water storage anomalies

Poyang Lake is the largest freshwater lake in China, serving as a natural reservoir and playing a paramount role in climate regulation, ecological environment, and water resource management. However, in recent years, Poyang Lake has approached desiccation multiple times, with severe droughts becoming increasingly common. Consequently, precise quantification and analysis of the terrestrial water storage anomalies (TWSA) and drought characteristics of the Poyang Lake basin (PLB) are of profound scientific and practical significance. This paper, for the first time, utilizes data for the period 2021–2022 from 77 newly-established GNSS observation stations in the PLB to precisely determine its vertical crustal displacement, invert daily and monthly TWSA, and investigate extreme hydrological drought. The results reveal the following: 1) The annual amplitude range of vertical surface displacements at GNSS stations in the Poyang Lake basin is from 7 to 14 mm, with the most substantial seasonal vertical displacements occurring during the months of June and July; 2) monthly GNSS-TWSA maintains a commendable consistency with TWSA data obtained from the Gravity Recovery and Climate Experiment (GRACE), the Global Land Data Assimilation System (GLDAS), and precipitation, with correlation coefficients of 0.67, 0.55, and 0.62, respectively; 3) at daily scale, the GNSS-derived Drought Severity Index (GNSS-DSI) accurately recorded the severity and intensity of eight drought events in the PLB during 2021–2022, in particular the period of extensive drought between October 2021 and February 2022, when drought intensity reaching a notable 1.03, which is classified as an extreme and prolonged drought event. Additionally, at local temporal scales, daily GNSS-DSI exhibits heightened sensitivity to drought signals. This study provides novel technological tools and datasets for multi-source satellite-based drought monitoring in the PLB.

Spatial-Temporal Dynamic Evolution of Land Deformation Driven by Hydrological Signals around Chaohu Lake.

The frequent occurrence of extreme climate events has a significant impact on people's lives. Heavy rainfall can lead to an increase of regional Terrestrial Water Storage (TWS), which will cause land subsidence due to the influence of hydrological load. At present, regional TWS is mostly obtained from Gravity Recovery and Climate Experiment (GRACE) data, but the method has limitations for small areas. This paper used water level and flow data as hydrological signals to study the land subsidence caused by heavy rainfall in the Chaohu Lake area of East China (June 2016-August 2016). Pearson's correlation coefficient was used to study the interconnection between water resource changes and Global Navigation Satellites System (GNSS) vertical displacement. Meanwhile, to address the reliability of the research results, combined with the Coefficient of determination method, the research findings were validated by using different institutional models. The results showed that: (1) During heavy rainfall, the vertical displacement caused by atmospheric load was larger than non-tidal oceanic load, and the influence trends of the two were opposite. (2) The rapidly increasing hydrologic load in the Chaohu Lake area resulted in greater subsidence displacement at the closer CORS station (CHCH station) than the more distant CORS station (LALA station). The Pearson correlation coefficients between the vertical displacement and water level were as high as -0.80 and -0.64, respectively. The phenomenon confirmed the elastic deformation principle of disc load. (3) Although there was a systematic bias between the different environmental load deformation models, the deformation trends were generally consistent with the GNSS monitoring results. The average Coefficients of determination between the different models and the GNSS results were 0.63 and 0.77, respectively. The results demonstrated the effectiveness of GNSS in monitoring short-term hydrological load. This study reveals the spatial-temporal evolution of land deformation during heavy rainfall around Chaohu Lake, which is of reference significance for water resource management and infrastructure maintenance in this area.

Is groundwater running out in the Western Cape, South Africa? Evaluating GRACE data to assess groundwater storage during droughts

Study regionThe Western Cape (WC), South Africa. Study focusThe WC has become increasingly dependent on groundwater in recent years due to repeated droughts. A framework to monitor the regional groundwater levels is urgently required to sustainably manage the WC’s water resources, since the region has inconsistent or unavailable monitoring data. Therefore, this study aims to understand how Gravity Recovery and Climate Experiment (GRACE) and Global Land Data Assimilation System (GLDAS) data can be used to monitor groundwater storage variations (ΔGWS) in the WC. In-situ ΔGWS time-series from twelve aquifers in the WC were compared to GRACE and GLDAS data. New hydrological insights for the regionGRACE terrestrial water storage anomalies (ΔTWS) showed moderate positive correlation (r = 0.69) with in-situ ΔGWS from the Adelaide Subgroup Aquifer (ASA), an unconfined aquifer with large areal extent and large ΔGWS. The Table Mountain Group Upper Aquifer Unit (TMG UAU) and Cape Flats Aquifer (CFA) also showed significant positive correlations with GLDAS ΔGWS of 0.83 and 0.73, respectively. Our results suggest that ΔGWS in the ASA can be monitored using GRACE ΔTWS, while GLDAS ΔGWS data can be used to monitor ΔGWS in the unconfined TMG UAU and CFA. GRACE and GLDAS data may be suitable to monitor groundwater availability in other water- and data-scarce regions of Africa.

Comparative evaluation of the dynamics of terrestrial water storage and drought incidences using multiple data sources: Tana sub-basin, Ethiopia

Abstract Evaluating water storage changes and addressing drought challenges in areas like the Tana sub-basin in Ethiopia is difficult due to limited data availability. The aim of this study was to evaluate the dynamics of terrestrial water anomaly and drought incidences by employing multiple data source. The Gravity Recovery and Climate Experiment (GRACE) and Global Land Data Assimilation System (GLDAS) datasets were used to assess the long-term water storage dynamics and drought incidences using the weighted water storage deficit index (WWSDI). WWSDI was used to identify drought periods, which ranged from severe to extreme drought. Despite the overall increase in average annual total water storage anomaly (TWSA) by 0.43 cm/year and a net gain of 50.68 cm equivalent water height from 2003 to 2022, there were instances of terrestrial water storage deficits, particularly in 2005, 2006, and 2009, during historical drought periods. The TWSA exhibited a strong correlation with Lake Tana water storage and precipitation anomalies after adjusting lag times. WWSDI displayed a high correlation with WSDI but a weak correlation with SPI and SPEI. Therefore, utilization of GRACE and GLDAS data is promising for evaluating terrestrial water storage and monitoring drought in data-deficient regions like the Tana sub-basin in Ethiopia.

Winter snow deficit was a harbinger of summer 2022 socio-hydrologic drought in the Po Basin, Italy

Snow in mountainous regions is a key water resource from local to continental scales. However, the link between increasingly frequent snow droughts and socio-hydrologic impacts from headwaters to lowlands is not fully understood. To untangle this relationship, here we combine snow and climate reanalyses with a range of data of terrestrial water storage, streamflow, and emergency water-use restrictions for the archetypal 2022 snow drought in the Po basin (Italy). We find that a persistent high-pressure ridge during winter 2021–2022 translated into a −88% anomaly in peak snow water equivalent, intraseasonal snowmelt, and earlier melt-out dates. Compounded by reduced summer precipitation and importantly increased temperatures, this snow deficit led to the lowest terrestrial water storage on record in summer 2022. Emergency water-use restrictions were concurrent with the peak in snowmelt deficit in early summer, rather than the peak in precipitation deficit in mid-winter. This study highlights the contribution of snowmelt deficit in driving the 2022 socio-hydrologic drought in the Po Basin.

Analysis and partitioning of terrestrial water storage in the Yellow River source region

AbstractIn order to increase the capability to understand and quantify the spatial differences in terrestrial water storage (TWS), and to reflect the unique energy balance processes and soil freeze–thaw mechanisms in the Qinghai‐Tibet Plateau (QTP), this study improved the energy balance processes of the water and energy transfer processes model, including its surface radiation calculations and snowmelt module. By integrating these improvements, a water and energy transfer processes model in Qinghai‐Tibet Plateau (WEP‐QTP) for the Yellow River source region (YRSR) is developed. Using the improved WEP‐QTP model to perform simulations, we assessed the daily changes in snow cover, soil moisture (SM), permafrost (PM), and groundwater storage (GWS) in the YRSR. Our analysis revealed an increase in TWS of 0.24 mm/yr from 1961 to 2020. Snow water equivalent (SWE), SM, PM, and GWS have proportional contributions of 8.33%, 216.67%, −154.17%, and 29.17% to the increased TWS, respectively. SM is the primary component of TWS. Temperature (T), precipitation (P), evapotranspiration (E), and solar radiation (Rs) influence the spatiotemporal variations in TWS, as well as those of its components. The increase in P is the primary cause for the rise in TWS, SWE, and SM, while the increase in T predominantly contributes to the decrease in PM. Furthermore, permafrost degradation and climate‐induced warming and humidification lead to increased infiltration, resulting in elevated GWS.

Impacts of Topography‐Driven Water Redistribution on Terrestrial Water Storage Change in California Through Ecosystem Responses

AbstractLateral subsurface flow plays an essential role in sustaining the terrestrial ecosystem, but it is not explicitly represented in most Earth System Models. In this study, we implemented an explicit lateral saturated flow model into the E3SM land model (ELM). The model explicitly describes lateral flow in the saturated zone by representing, for each model grid, an idealized hillslope consisting of five hydrologically connected soil columns. We conducted three model experiments driven by 0.125° atmospheric forcing data during 1980–2015 over California using models of the default ELM, a modified version of ELM to enhance infiltration, and the model with the lateral saturated flow model. The simulated runoff, evapotranspiration, and terrestrial water storage anomaly (TWSA) from the three simulations were evaluated against available observations, and the model explicitly representing lateral flow performs best. The new model produces greater gridcell‐averaged evapotranspiration especially over the mountainous regions with moderate relief and seasonally dry climates. Most importantly, it improves the modeled seasonal variations, interannual variabilities, and the recent decadal decline of TWSA. Many of these improvements can be attributed to the enhanced ecosystem resilience to droughts as demonstrated by transpiration increases caused by lateral flow. Model sensitivity experiments suggest that subsurface runoff is most sensitive to the ratio between horizontal and vertical saturated hydraulic conductivity, followed by hillslope planforms (convergent, divergent, and uniform), number of columns, and lower boundary conditions. Future work should effectively characterize hillslopes in global models and explore the long‐term influences of lateral water movement on modeled biogeochemical cycle.

Hydrological Assessment of Iraq for the Period 2002-2020 Using GRACE Terrestrial Water Storage Deficit

The research examines the effectiveness of remote sensing for monitoring drought conditions, specifically focusing on the utilization of satellite missions like GRACE to detect alterations in water levels. The study analyzes GRACE data for 10 Jet Propulsion Laboratory (JPL) grid cells alongside precipitation data from 11 meteorological stations. It employs the DrinC program to calculate the standard precipitation index (SPI) at different time intervals. Non- Non-parametric Mann-Kendall tests and Sen's slope statistics are used to determine the trend and magnitude of the GRACE-based anomaly in terrestrial water storage. The statistical findings reveal a significant decrease in Iraq's terrestrial water storage anomaly between 2007 and 2017, indicating drier climatic conditions. However, there is an increase in anomaly between 2002 and 2006, as well as from 2018 to 2020, indicating wetter conditions. The water storage deficit varies among grid cells, each exhibiting distinct characteristics and patterns. The study also highlights that more intense and prolonged droughts tend to occur in the northern and southern regions of Iraq, while the central region experiences more frequent but less severe droughts. In most grid cells, the Mann-Kendall test illustrates a substantial decrease in the water storage deficit index and a significant increase in SPI-06. On the other hand, only one grid cell shows a noticeable precipitation trend. Comparatively, trends in the water storage deficit index and SPI-06 are similar. Overall, the water storage deficit and its index prove valuable in predicting drought severity in locations lacking hydrological measurements.

Future water storage changes over the Mediterranean, Middle East, and North Africa in response to global warming and stratospheric aerosol intervention

Abstract. Water storage plays a profound role in the lives of people across the Middle East and North Africa (MENA) as it is the most water-stressed region worldwide. The lands around the Caspian and Mediterranean seas are simulated to be very sensitive to future climate warming. Available water capacity depends on hydroclimate variables such as temperature and precipitation that will depend on socioeconomic pathways and changes in climate. This work explores changes in both the mean and extreme terrestrial water storage (TWS) under an unmitigated greenhouse gas (GHG) scenario (SSP5-8.5) and stratospheric aerosol intervention (SAI) designed to offset GHG-induced warming above 1.5 ∘C and compares both with historical period simulations. Both mean TWS and extreme TWS are projected to significantly decrease under SSP5-8.5 over the domain, except for the Arabian Peninsula, particularly in the wetter lands around the Caspian and Mediterranean seas. Relative to global warming, SAI partially ameliorates the decreased mean TWS in the wet regions, while it has no significant effect on the increased TWS in drier lands. In the entire domain studied, the mean TWS is larger under SAI than pure GHG forcing, mainly due to the significant cooling and, in turn, a substantial decrease in evapotranspiration under SAI relative to SSP5-8.5. Changes in extreme water storage excursions under global warming are reduced by SAI. Extreme TWS under both future climate scenarios is larger than throughout the historical period across Iran, Iraq, and the Arabian Peninsula, but the response of the more continental eastern North Africa hyper-arid climate is different from the neighboring dry lands. In the latter case, we note a reduction in the mean TWS trend under both GHG and SAI scenarios, with extreme TWS values also showing a decline compared to historical conditions.

Quantifying the 2022 extreme drought in the Yangtze River Basin using GRACE-FO

In 2022, the Yangtze River Basin (YRB) experienced its most extreme drought event since 1961. Conducting a comprehensive assessment of this severe drought is crucial for understanding the entire drought process. GRACE and GRACE-FO gravity satellites have the capability to monitor global mass transport, making them suitable for large-scale drought monitoring. In this study, we investigated the changes in water storage caused by the drought using GRACE/GRACE-FO data and multi-source observations. Our evaluation incorporated meteorological drought indices, including the Standardized Precipitation Index (SPI) and the Standardized Precipitation Evapotranspiration Index (SPEI), to explore the drought propagation process. Additionally, we examined the effects of trends in terrestrial water storage in the YRB, resulting from reservoir impoundment and climatic factors, on the drought index calculation. The results indicate that the extreme drought in the YRB began in August 2022, resulting in a water storage loss of approximately −158 ± 22 km3 compared to July 2022. There was almost no delay in the response of soil moisture drought and hydrological drought to the meteorological drought at a monthly scale. This extreme drought had its most severe impact on the middle and lower reaches of the YRB, especially in the Dongting Lake Basin, where the GRACE drought index value remained below −2.5 until December 2022. This study demonstrates the significant water storage deficit during the extreme drought in the YRB in 2022, providing a comprehensive understanding of this historic drought.

Analysis of gap filling techniques for GRACE/GRACE-FO terrestrial water storage anomalies in Canada

Terrestrial water storage anomaly (TWSA) data derived from the Gravity Recovery and Climate Experiment (GRACE/GRACE-FO) have become indispensable for hydrological monitoring since the first mission launched in 2002. Evaluation of basin dynamics is enabled through TWSA integration with independent hydrological and climate datasets such as precipitation, evapotranspiration, and surface runoff. With over 20 years of observations, long-term statistical analysis reveals water storage trends, provided the 11-month gap between the two missions is filled. This study compares four methods for reconstructing GRACE-like TWSA over Canada, namely: extreme gradient boosting, artificial neural networks, automated machine learning, and projection onto convex sets. The GRACE mascon product released by the Jet Propulsion Laboratory is used over the study region. Nine sets of hydrological and climate parameters are used as predictors for the machine learning models, namely GRACE average seasonal signals, TWSA from the GLDAS Catchment Land Surface Model, precipitation, air temperature, glacier surface mass balance, ocean tides, the North Atlantic Oscillation, the Multivariate ENSO Index, and sea surface temperature. For each of the four methods, 20 % of the GRACE TWSA data (the ‘testing data’) is omitted from the model training, and the model is used to fill the artificial gap. Root mean square error (RMSE) is computed using the difference between testing data and model predictions. Normalized RMSE expresses the RMSE as a proportion of the range of the GRACE TWSA data. The automated machine learning algorithm performs best in terms of mean normalized root mean square error, with 6 % (2.4 cm equivalent water height RMSE), followed by projection onto convex sets, extreme gradient boosting, and artificial neural networks. Inclusion of glacier surface mass balance models derived from the GMAO Modern-Era Retrospective Analysis for Research and Applications Version 2 reanalysis and the Randolph Glacier Inventory improved the automated machine learning average root mean square error by 70 %. Results indicate that filling the gap between missions allows for more comprehensive analysis of terrestrial water storage changes, including basin-by-basin analysis, which is ongoing.

Spatiotemporal distribution of groundwater drought using GRACE-based satellite estimates: a case study of Lower Gangetic Basin, India.

Droughts frequently occurring in India have significant societal, economic, and environmental effects. The lack of direct measurements of groundwater in location and time hinders quantitative methods to analyse the intricate nature of groundwater drought. This work used the datasets derived from the Gravity and Climate Experiment (GRACE and GRACE-FO) and Global Land Data Assimilation System (GLDAS) to extensively analyse Groundwater Storage changes in the Lower Gangetic Basin (LGB) using unique hydrological parameters between the years 2003 and 2022. The analysis highlights that the GRACE-derived terrestrial water storage anomaly in the LGB decreased significantly (-12.12 mm/yr), and the amount of Groundwater Storage Anomaly (GWSA) decreased similarly (-10.80 mm/yr), while in the GRACE-FO period, a positive trend has been noticed in TWSA (33.96 mm/yr) and GWSA (64.8 mm/yr) respectively. A drought indicator called the GRACE-derived groundwater drought index (GGDI) has been computed for the entire LGB region. A traditional drought study viz. Standardised Precipitation Index (SPI) was performed over LGB to justify the results of the GGDI. The results from GGDI study effectively matched the periods of significant drought occurrences with the 12-month SPI time series. From the GGDI, this study examined groundwater drought's spatial distribution, temporal evolution, and trend (Modified Mann Kendall trend) aspects. According to research findings, the LGB experienced three major drought periods between 2009-2010, 2019 (moderate), and 2015-2016 (severe). The study offers reliable quantitative data on the evolution of GRACE-derived groundwater drought, which may add a new perspective to additional drought research in the densely populated study area, which depends majorly on agriculture, livestock and less skilled water-intensive industries such as leather and textile industries in a sub-tropical climate. This paradigm incorporates changes in groundwater resources caused by human activities and climate change, paving the way for measuring progress towards sustainable use and water security.

Evaluating dynamics of land water storage and its response to climate variation in the Deccan Plateau, India

Inter-state river water disputes are frequent over the rain shadow (RS) region of the Western Ghats (WG), where the annual rainfall is much lower (75.2 cm) than that of overall India (118 cm). This study investigated the TWS (Terrestrial Water Storage) dynamics of the entire Deccan plateau (DP) and RS regions, located at the leeward side of the WG mountain range, with the help of Gravity Recovery and Climate Experiment (GRACE) based TWS anomalies (TWSA) data. Here, the latest TRMM (Tropical Rainfall Measuring Mission) rainfall product was evaluated at different time scales against gauge-based measurements to check its applicability as an alternative to gauge-based observed data. Trends in climate variables and TWSA were estimated to understand their variations during 2003–2016 in both regions. The response of GRACE-based TWSA and its constituent components to rainfall was determined using time-lagged correlation analysis. The contribution of constituent variables to TWSA was estimated using the component contribution ratio (CCR) method. It was observed that the TRMM-3B43 data captured the rainfall pattern over the entire DP with great accuracy. Results showed that the increasing trend of TWS in 2003–2009 had reversed in 2010–2016 due to the reduction of climate moistening in both regions. The warming trend of climate is more than twice in the RS region than that found in the DP, leading to a higher decreasing rate of TWS in this region than in DP. For both DP and RS regions, the time lags of groundwater storage anomalies (GWSA), total runoff anomalies (RA), canopy water storage anomalies (CWSA), and soil moisture storage anomalies (SMSA) to rainfall can be arranged as CWSA ≤ RA ≤ SMSA ≤ GWSA. SMSA and GWSA are the main contributors to the TWSA in both areas. However, SWSA also played a vital role in TWS variability, particularly in the RS region.

Evaluation Of Terrestrial Water Storage Products From Remote Sensing, Land Surface Model And Regional Hydrological Model Over Northern European Russia

Water storage is one of the key components of terrestrial water balance, therefore its accurate assessment is necessary for a sufficient description of hydrological processes within river basins. Here we compare terrestrial water storage using calibrated hydrological model ECOMAG forced by gauge observations, uncalibrated INM RAS–MSU land surface model forced by reanalysis and GRACE satellite-based data over Northern Dvina and Pechora River basins. To clearly identify differences between the datasets long-term, seasonal and residual components were derived. Results show a predominance of the seasonal component variability over the region (~64% of the total) by all datasets but INM RAS–MSU shows a substantial percentage of long-term component variability as well (~31%), while GRACE and ECOMAG demonstrate the magnitude around 18%. Moreover, INM RAS–MSU shows lowest magnitude of annual range. ECOMAG and INM RAS–MSU is distinguished by earliest begin of TWS decline in spring, while GRACE demonstrates latest dates. Overall, ECOMAG has shown the lowest magnitude of random error from 9 mm for Northern Dvina basin to 10 mm for Pechora basin, while INM RAS–MSU has shown largest one.

GLOBGM v1.0: a parallel implementation of a 30 arcsec PCR-GLOBWB-MODFLOW global-scale groundwater model

Abstract. We discuss the various performance aspects of parallelizing our transient global-scale groundwater model at 30′′ resolution (30 arcsec; ∼ 1 km at the Equator) on large distributed memory parallel clusters. This model, referred to as GLOBGM, is the successor of our 5′ (5 arcmin; ∼ 10 km at the Equator) PCR-GLOBWB 2 (PCRaster Global Water Balance model) groundwater model, based on MODFLOW having two model layers. The current version of GLOBGM (v1.0) used in this study also has two model layers, is uncalibrated, and uses available 30′′ PCR-GLOBWB data. Increasing the model resolution from 5′ to 30′′ creates challenges, including increased runtime, memory usage, and data storage that exceed the capacity of a single computer. We show that our parallelization tackles these problems with relatively low parallel hardware requirements to meet the needs of users or modelers who do not have exclusive access to hundreds or thousands of nodes within a supercomputer. For our simulation, we use unstructured grids and a prototype version of MODFLOW 6 that we have parallelized using the message-passing interface. We construct independent unstructured grids with a total of 278 million active cells to cancel all redundant sea and land cells, while satisfying all necessary boundary conditions, and distribute them over three continental-scale groundwater models (168 million – Afro–Eurasia; 77 million – the Americas; 16 million – Australia) and one remaining model for the smaller islands (17 million). Each of the four groundwater models is partitioned into multiple non-overlapping submodels that are tightly coupled within the MODFLOW linear solver, where each submodel is uniquely assigned to one processor core, and associated submodel data are written in parallel during the pre-processing, using data tiles. For balancing the parallel workload in advance, we apply the widely used METIS graph partitioner in two ways: it is straightforwardly applied to all (lateral) model grid cells, and it is applied in an area-based manner to HydroBASINS catchments that are assigned to submodels for pre-sorting to a future coupling with surface water. We consider an experiment for simulating the years 1958–2015 with daily time steps and monthly input, including a 20-year spin-up, on the Dutch national supercomputer Snellius. Given that the serial simulation would require ∼ 4.5 months of runtime, we set a hypothetical target of a maximum of 16 h of simulation runtime. We show that 12 nodes (32 cores per node; 384 cores in total) are sufficient to achieve this target, resulting in a speedup of 138 for the largest Afro–Eurasia model when using 7 nodes (224 cores) in parallel. A limited evaluation of the model output using the United States Geological Survey (USGS) National Water Information System (NWIS) head observations for the contiguous United States was conducted. This showed that increasing the resolution from 5′ to 30′′ results in a significant improvement with GLOBGM for the steady-state simulation when compared to the 5′ PCR-GLOBWB groundwater model. However, results for the transient simulation are quite similar, and there is much room for improvement. Monthly and multi-year total terrestrial water storage anomalies derived from the GLOBGM and PCR-GLOBWB models, however, compared favorably with observations from the GRACE satellite. For the next versions of GLOBGM, further improvements require a more detailed (hydro)geological schematization and better information on the locations, depths, and pumping rates of abstraction wells.

Advancing SDGs: Predicting Future Shifts in Saudi Arabia’s Terrestrial Water Storage Using Multi-Step-Ahead Machine Learning Based on GRACE Data

The availability of water is crucial for the growth and sustainability of human development. The effective management of water resources is essential due to their renewable nature and their critical role in ensuring food security and water safety. In this study, the multi-step-ahead modeling approach of the Gravity Recovery and Climate Experiment (GRACE) terrestrial water storage (TWS) was utilized to gain insights into and forecast the fluctuations in water resources within Saudi Arabia. This study was conducted using mascon solutions obtained from the University of Texas Center for Space Research (UT-CSR) over the period of 2007 to 2017. The data were used in the development of artificial intelligence models, namely, an Elman neural network (ENN), a backpropagation neural network (BPNN), and kernel support vector regression (k-SVR). These models were constructed using various input variables, such as t-12, t-24, t-36, t-48, and TWS, with the output variable being the focus. A simple and weighted average ensemble was introduced to improve the accuracy of marginal and weak predictive results. The performance of the models was assessed with the use of several evaluation metrics, including mean absolute error (MAE), root mean square error (RMSE), mean absolute percentage error (MAPE), correlation coefficient (CC), and Nash–Sutcliffe efficiency (NSE). The results of the estimate indicate that k-SVR-M1 (NSE = 0.993, MAE = 0.0346) produced favorable outcomes, whereas ENN-M3 (NSE = 0.6586, MAE = 0.6895) emerged as the second most effective model. The combinations of all other models exhibited accuracies ranging from excellent to marginal, rendering them unreliable for decision-making purposes. Error ensemble methods improved the standalone model and proved merit. The results also serve as an important tool for monitoring changes in global water resources, aiding in drought management, and understanding the Earth’s water cycle.

Water-balance-based evapotranspiration for 56 large river basins: A benchmarking dataset for global terrestrial evapotranspiration modeling

A thorough validation of global terrestrial evapotranspiration (ET) models requires reliable ground-observed ET data. While the open-access eddy-covariance flux measurements are widely used for such a purpose, they have certain drawbacks, and thus, other alternative publicly available reference datasets are urgently needed. Using remote sensing and ground-based observational data, this study provides water-balance-based evapotranspiration (ETwb) estimates for 56 large (>105 km2) river basins of the world over the 1983–2016 period. For each basin, the observed runoff and four different precipitation (Prec) data sources were combined with three types of terrestrial water storage change (dS) estimates, yielding altogether 12 annual ETwb time-series. An optimally merged ETwb time-series was eventually produced using the Bayesian-based three-cornered hat method. The relative uncertainty in the ETwb estimates is less than 10 % in most basins and it stems primarily from the uncertainty in Prec. In summary, this new ETwb dataset has the following advantages: i) The gauge undercatch in Prec was corrected, thus mitigating the well-known general underestimation of ETwb in mid- and high-latitudes; ii) Multiple Prec and dS datasets were combined to account for the uncertainty in the water balance approach, thereby enabling the quantification of the uncertainty in ETwb and its sources, and; iii) The ETwb dataset stretches more than three decades, making it appropriate for evaluating long-term trends in global ET models. This ETwb dataset is publicly available (https://data.tpdc.ac.cn/en/data/e010cd0d-0881-4e7e-9d63-d36992750b04) and may serve as a benchmarking tool to calibrate/validate large-scale ET models for an improved understanding of regional- and/or global-scale ET processes.

Remote Sensing-Based Long-Term Assessment of Water Dynamics and Influencing Factors in Abaya and Chamo Lakes, East African Rift Valley, Ethiopia

Lakes are essential to global freshwater systems, supporting ecosystem services and ecological processes, but they are increasingly impacted by climate change and human activities. This study examined the long-term dynamics of Lakes Abaya and Chamo in the Ethiopian Rift Valley using Landsat images, altimetry-derived lake level data, and a machine learning method within Google Earth Engine, specifically a Multi-Index-based Random Forest (MIRF) classifier. The MIRF classifier achieved high accuracy, ranging from 97.58% to 99.13%, in extracting lake surface water. Substantial fluctuations in lake areas and lake levels were observed: Lake Abaya’s area decreased from 2000 to 2005 at a rate of 6.67 km²/year, then expanded until 2022 at the same increasing rate; Lake Chamo’s area decreased from 2000 to 2010 at a rate of 1.62 km²/year, then expanded until 2022 at a rate of 2.88 km²/year. The correlation analysis between lake areas and environmental factors such as Rainfall (RF), Temperature (TEMP), Normalized Difference Vegetation Index (NDVI), Groundwater Storage (GWS), Terrestrial Water Storage (TWS), and Soil Moisture (SM) revealed important associations. For Lake Abaya, strong correlations were identified with NDVI and TWS, suggesting that vegetation cover and terrestrial water significantly influence its area changes. In contrast, for Lake Chamo, NDVI emerged as the key driver, indicating that vegetation dynamics play a crucial role in the lake’s fluctuations. Furthermore, higher-order polynomial regression models were developed to better capture the complex relationships between lake area and water levels for both lakes. In general, this study integrates remote sensing, machine learning, and cloud computing, offering valuable insights into the lakes’ long-term characteristic and providing critical information for future water resource management strategies.