Downscaling GRACE Remote Sensing Datasets to High-Resolution Groundwater Storage Change Maps of California’s Central Valley

Statistical downscaling of GRACE terrestrial water storage changes based on the Australian Water Outlook model

The coarse spatial resolution of the Gravity Recovery and Climate Experiment (GRACE) dataset has limited its application in local water resource management and accounting. Despite efforts to improve GRACE spatial resolution, achieving high resolution downscaled grids that correspond to local hydrological behaviour and patterns is still limited. To overcome this issue, we propose a novel statistical downscaling approach to improve the spatial resolution of GRACE-terrestrial water storage changes (ΔTWS) using precipitation, evapotranspiration (ET), and runoff data from the Australian Water Outlook. These water budget components drive changes in the GRACE water column in much of the global land area. Here, the GRACE dataset is downscaled from the original resolution of 1.0° × 1.0° to 0.05° × 0.05° over a large hydro-geologic basin in northern Australia (the Cambrian Limestone Aquifer-CLA), capturing sub- grid heterogeneity in ΔTWS of the region. The downscaled results are validated using data from 12 in-situ groundwater monitoring stations and water budget estimates of the CLA's land water storage changes from April 2002 to June 2017. The change in water storage over time (ds/dt) estimated from the water budget model was weakly correlated (r = 0.34) with the downscaled GRACE ΔTWS. The weak relationship was attributed to the possible uncertainties inherent in the ET datasets used in the water budget, particularly during the summer months. Our proposed methodology provides an opportunity to improve freshwater reporting using GRACE and enhances the feasibility of downscaling efforts for other hydrological data to strengthen local-scale applications.

Spatial and temporal downscaling schemes to reconstruct high-resolution GRACE data: A case study in the Tarim River Basin, Northwest China

Climate change and excessive exploitation of water resources exert pressure on groundwater supply and the ecosystem in drylands. Although The Gravity Recovery and Climate Experiment (GRACE) satellites has demonstrated the feasibility of quantifying global groundwater storage variations, monitoring regional-scale groundwater has been challenging due to the coarse resolution of GRACE. Previous GRACE downscaling studies focused on develop new algorithms based on the perspective of pixel spatial correlation to improve resolution, which cannot better capture the temporal evolution of GRACE data effectively. In this study, we employ the semi-supervised variational autoencoder (SSVAER) algorithm and the multi-scale geographically weighted regression (MGWR) model to establish two different downscaling schemes: pixel temporal continuity downscaling and pixel spatial correlation downscaling. These schemes achieve spatial resolution downscaling of GRACE-derived groundwater storage anomalies (GWSA) from 0.5° to 0.1°. Additionally, the applicability of the PCR-GLOBWB model in drylands is verified. Furtherly, the spatiotemporal distribution patterns of GWSA are analyzed. The results show that (1) Both the temporal and spatial downscaling methods produced consistent results, with data correlations ranged from 0.94 to 0.98 observed in over 80 % of the range before and after downscaling; (2) The groundwater storage change rate in the northern Tarim River Basin (TRB) is 25 times greater than the model results, while in other regions, the average deviation is 2.6 times; (3) The two schemes enhance the correlation (0.27) between GWSA and groundwater level anomaly (GWLA) to 0.59 and 0.52, respectively, with a three-month lag in GWSA relative to GWLA. The temporal downscaling approach exhibited higher CC and lower RMSE, outperforming the spatial downscaling approach. The high-resolution results in this study can well complement groundwater level prediction efforts in arid regions and provide quantitative information for local water resource management.

Using Downscaled GRACE Mascon Data to Assess Total Water Storage in Mississippi Alluvial Plain Aquifer

The importance of high-resolution and continuous hydrologic data for monitoring and predicting water levels is crucial for sustainable water management. Monitoring Total Water Storage (TWS) over large areas by using satellite images such as Gravity Recovery and Climate Experiment (GRACE) data with coarse resolution (1°) is acceptable. However, using coarse satellite images for monitoring TWS and changes over a small area is challenging. In this study, we used the Random Forest model (RFM) to spatially downscale the GRACE mascon image of April 2020 from 0.5° to ~5 km. We initially used eight different physical and hydrological parameters in the model and finally used the four most significant of them for the final output. We executed the RFM for Mississippi Alluvial Plain. The validating data R² for each model was 0.88. Large R² and small RMSE and MAE are indicative of a good fit and accurate predictions by RFM. The result of this research aligns with the reported water depletion in the central Mississippi Delta area. Therefore, by using the Random Forest model and appropriate parameters as input of the model, we can downscale the GRACE mascon image to provide a more beneficial result that can be used for activities such as groundwater management at a sub-county-level scale in the Mississippi Delta.

Machine learning based downscaling of GRACE-estimated groundwater in Central Valley, California

California's Central Valley, one of the most agriculturally productive regions, is also one of the most stressed aquifers in the world due to anthropogenic groundwater over-extraction primarily for irrigation. Groundwater depletion is further exacerbated by climate-driven droughts. Gravity Recovery and Climate Experiment (GRACE) satellite gravimetry has demonstrated the feasibility of quantifying global groundwater storage changes at uniform monthly sampling, though at a coarse resolution and is thus impractical for effective water resources management. Here, we employ the Random Forest machine learning algorithm to establish empirical relationships between GRACE-derived groundwater storage and in situ groundwater level variations over the Central Valley during 2002-2016 and achieved spatial downscaling of GRACE-observed groundwater storage changes from a few hundred km to 5 km. Validations of our modeled groundwater level with in situ groundwater level indicate excellent Nash-Sutcliffe Efficiency coefficients ranging from 0.94 to 0.97. In addition, the secular components of modeled groundwater show good agreements with those of vertical displacements observed by GPS, and CryoSat-2 radar altimetry measurements and is perfectly consistent with findings from previous studies. Our estimated groundwater loss is about 30 km³ from 2002 to 2016, which also agrees well with previous studies in Central Valley. We find the maximum groundwater storage loss rates of -5.7 ± 1.2 km³ yr^-1 and -9.8 ± 1.7 km³ yr^-1 occurred during the extended drought periods of January 2007-December 2009, and October 2011-September 2015, respectively while Central Valley also experienced groundwater recharges during prolonged flood episodes. The 5-km resolution Central Valley-wide groundwater storage trends reveal that groundwater depletion occurs mostly in southern San Joaquin Valley collocated with severe land subsidence due to aquifer compaction from excessive groundwater over withdrawal.

Enhancing spatial resolution of GRACE-derived groundwater storage anomalies in Urmia catchment using machine learning downscaling methods

The Urmia lake in north-west Iran has dried up to perilously low levels in the past two decades. In this study, we investigate the drivers behind the decline in lake water level with the help of in-situ and remote sensing data. We use total water storage (TWS) changes from the gravity recovery and climate experiment (GRACE) satellite mission. TWS from GRACE includes all the water storage compartments in a column and is the only remote sensing product that can help in estimating groundwater storage (GWS) changes. The coarse spatial (approx. 300 km) resolution of GRACE does not allow us to identify local changes that may have led to the Urmia lake disaster. In this study, we tackle the poor resolution of the GRACE data by employing three machine learning (ML) methods including random forest (RF), support vector regression (SVR) and multi-layer perceptron (MLP). The methods predict the groundwater storage anomaly (GWSA), derived from GRACE, as a function of hydro-climatic variables such as precipitation, evapotranspiration, land surface temperature (LST) and normalized difference vegetation index (NDVI) on a finer scale of 0.25° × 0.25°. We found that i) The RF model exhibited highest R (0.98), highest NSE (0.96) and lowest RMSE (18.36 mm) values. ii) The RF downscaled data indicated that the exploitation of groundwater resources in the aquifers is the main driver of groundwater storage and changes in the regional ecosystem, which has been corroborated by few other studies as well. The impact of precipitation and evapotranspiration on the GWSA was found to be rather weak, indicating that the anthropogenic derivers had the most significant impact on the GWSA changes. iii) We generally observed a significant negative trend in GWSA, having also significant positive correlations with the well data. However, over regions with dam construction significant negative correlations were found.

A Comparison of Ensemble and Deep Learning Algorithms to Model Groundwater Levels in a Data-Scarce Aquifer of Southern Africa

Machine learning and deep learning have demonstrated usefulness in modelling various groundwater phenomena. However, these techniques require large amounts of data to develop reliable models. In the Southern African Development Community, groundwater datasets are generally poorly developed. Hence, the question arises as to whether machine learning can be a reliable tool to support groundwater management in the data-scarce environments of Southern Africa. This study tests two machine learning algorithms, a gradient-boosted decision tree (GBDT) and a long short-term memory neural network (LSTM-NN), to model groundwater level (GWL) changes in the Shire Valley Alluvial Aquifer. Using data from two boreholes, Ngabu (sample size = 96) and Nsanje (sample size = 45), we model two predictive scenarios: (I) predicting the change in the current month’s groundwater level, and (II) predicting the change in the following month’s groundwater level. For the Ngabu borehole, GBDT achieved R2 scores of 0.19 and 0.14, while LSTM achieved R2 scores of 0.30 and 0.30, in experiments I and II, respectively. For the Nsanje borehole, GBDT achieved R2 of −0.04 and −0.21, while LSTM achieved R2 scores of 0.03 and −0.15, in experiments I and II, respectively. The results illustrate that LSTM performs better than the GBDT model, especially regarding slightly greater time series and extreme GWL changes. However, closer inspection reveals that where datasets are relatively small (e.g., Nsanje), the GBDT model may be more efficient, considering the cost required to tune, train, and test the LSTM model. Assessing the full spectrum of results, we concluded that these small sample sizes might not be sufficient to develop generalised and reliable machine learning models.

Improving the Spatial Resolution of GRACE-Derived Terrestrial Water Storage Changes in Small Areas Using the Machine Learning Spatial Downscaling Method

Gravity Recovery and Climate Experiment (GRACE) satellites can effectively monitor terrestrial water storage (TWS) changes in large-scale areas. However, due to the coarse resolution of GRACE products, there is still a large number of deficiencies that need to be considered when investigating TWS changes in small-scale areas. Hence, it is necessary to downscale the GRACE products with a coarse resolution. First, in order to solve this problem, the present study employs modeling windows of different sizes (Window Size, WS) combined with multiple machine learning algorithms to develop a new machine learning spatial downscaling method (MLSDM) in the spatial dimension. Second, The MLSDM is used to improve the spatial resolution of GRACE observations from 0.5° to 0.25°, which is applied to Guantao County. The present study has verified the downscaling accuracy of the model developed through the combination of WS3, WS5, WS7, and WS9 and jointed with Random Forest (RF), Extra Tree Regressor (ETR), Adaptive Boosting Regressor (ABR), and Gradient Boosting Regressor (GBR) algorithms. The analysis shows that the accuracy of each combined model is improved after adding the residuals to the high-resolution downscaled results. In each modeling window, the accuracy of RF is better than that of ETR, ABR, and GBR. Additionally, compared to the changes in the TWS time series that are derived by the model before and after downscaling, the results indicate that the downscaling accuracy of WS5 is slightly more superior compared to that of WS3, WS7, and WS9. Third, the spatial resolution of the GRACE data was increased from 0.5° to 0.05° by integrating the WS5 and RF algorithm. The results are as follows: (1) The TWS (GWS) changes before and after downscaling are consistent, decreasing at −20.86 mm/yr and −21.79 mm/yr (−14.53 mm/yr and −15.46 mm/yr), respectively, and the Nash–Sutcliffe efficiency coefficient (NSE) and correlation coefficient (CC) values of both are above 0.99 (0.98). (2) The CC between the 80% deep groundwater well data and the downscaled GWS changes are above 0.70. Overall, the MLSDM can not only effectively improve the spatial resolution of GRACE products but also can preserve the spatial distribution of the original signal, which can provide a reference scheme for research focusing on the downscaling of GRACE products.

Improving the Resolution of GRACE Data for Spatio-Temporal Groundwater Storage Assessment

Groundwater has a significant contribution to water storage and is considered to be one of the sources for agricultural irrigation; industrial; and domestic water use. The Gravity Recovery and Climate Experiment (GRACE) satellite provides a unique opportunity to evaluate terrestrial water storage (TWS) and groundwater storage (GWS) at a large spatial scale. However; the coarse resolution of GRACE limits its ability to investigate the water storage change at a small scale. It is; therefore; needed to improve the resolution of GRACE data at a spatial scale applicable for regional-level studies. In this study; a machine-learning-based downscaling random forest model (RFM) and artificial neural network (ANN) model were developed to downscale GRACE data (TWS and GWS) from 1° to a higher resolution (0.25°). The spatial maps of downscaled TWS and GWS were generated over the Indus basin irrigation system (IBIS). Variations in TWS of GRACE in combination with geospatial variables; including digital elevation model (DEM), slope; aspect; and hydrological variables; including soil moisture; evapotranspiration; rainfall; surface runoff; canopy water; and temperature; were used. The geospatial and hydrological variables could potentially contribute to; or correlate with; GRACE TWS. The RFM outperformed the ANN model and results show Pearson correlation coefficient (R) (0.97), root mean square error (RMSE) (11.83 mm), mean absolute error (MAE) (7.71 mm), and Nash–Sutcliffe efficiency (NSE) (0.94) while comparing with the training dataset from 2003 to 2016. These results indicate the suitability of RFM to downscale GRACE data at a regional scale. The downscaled GWS data were analyzed; and we observed that the region has lost GWS of about −9.54 ± 1.27 km3 at the rate of −0.68 ± 0.09 km3/year from 2003 to 2016. The validation results showed that R between downscaled GWS and observational wells GWS are 0.67 and 0.77 at seasonal and annual scales with a confidence level of 95%, respectively. It can; therefore; be concluded that the RFM has the potential to downscale GRACE data at a spatial scale suitable to predict GWS at regional scales.

Downscaling GRACE total water storage change using partial least squares regression

The Gravity Recovery And Climate Experiment (GRACE) satellite mission recorded temporal variations in the Earth’s gravity field, which are then converted to Total Water Storage Change (TWSC) fields representing an anomaly in the water mass stored in all three physical states, on and below the surface of the Earth. GRACE provided a first global observational record of water mass redistribution at spatial scales greater than 63000 km². This limits their usability in regional hydrological applications. In this study, we implement a statistical downscaling approach that assimilates 0.5° × 0.5° water storage fields from the WaterGAP hydrology model (WGHM), precipitation fields from 3 models, evapotranspiration and runoff from 2 models, with GRACE data to obtain TWSC at a 0.5° × 0.5° grid. The downscaled product exploits dominant common statistical modes between all the hydrological datasets to improve the spatial resolution of GRACE. We also provide open access to scripts that researchers can use to produce downscaled TWSC fields with input observations and models of their own choice.

Measurement(s)	Gravity
Technology Type(s)	gravity field theory • computational modeling technique
Factor Type(s)	geographic location • temporal interval
Sample Characteristic - Environment	water body
Sample Characteristic - Location	global

Machine-accessible metadata file describing the reported data: 10.6084/m9.figshare.13503114

Exploring groundwater and soil water storage changes across the CONUS at 12.5 km resolution by a Bayesian integration of GRACE data into W3RA

Climate variability and change along with anthropogenic water use have affected the (re)distribution of water storage and fluxes across the Contiguous United States (CONUS). Available hydrological models, however, do not represent recent changes in the water cycle. Therefore, in this study, a novel Bayesian Markov Chain Monte Carlo-based Data Assimilation (MCMC-DA) approach is formulated to integrate Terrestrial Water Storage changes (TWSC) from the Gravity Recovery and Climate Experiment (GRACE) satellite mission into the W3RA water balance model. The benefit of this integration is its dynamic solution that uses GRACE TWSC to update W3RA's individual water storage estimates while rigorously accounting for uncertainties. It also down-scales GRACE data and provides groundwater and soil water storage changes at ~12.5 km resolution across the CONUS covering 2003-2017. Independent validations are performed against in-situ groundwater data (from USGS) and Climate Change Initiative (CCI) soil moisture products from the European Space Agency (ESA). Our results indicate that MCMC-DA introduces trends, which exist in GRACE TWSC, mostly to the groundwater storage and to a lesser extent to the soil water storage. Higher similarity is found between groundwater estimation of MCMC-DA and those of USGS in the southeastern CONUS. We also show a stronger linear trend in MCMC-DA soil water storage across the CONUS, compared to W3RA (changing from ±0.5 mm/yr to ±2 mm/yr), which is closer to independent estimates from the ESA CCI. MCMC-DA also improves the estimation of soil water storage in regions with high forest intensity, where ESA CCI and hydrological models have difficulties in capturing the soil-vegetation-atmosphere continuum. The representation of El Niño Southern Oscillation (ENSO)-related variability in groundwater and soil water storage are found to be considerably improved after integrating GRACE TWSC with W3RA. This new hybrid approach shows promise for understanding the links between climate and the water balance over broad regions.

Downscaling Groundwater Storage Data in China to a 1-km Resolution Using Machine Learning Methods

High-resolution and continuous hydrological products have tremendous importance for the prediction of water-related trends and enhancing the capability for sustainable water resources management under climate change and human impacts. In this study, we used the random forest (RF) and extreme gradient boosting (XGBoost) methods to downscale groundwater storage (GWS) from 1° (~110 km) to 1 km by downscaling Gravity Recovery and Climate Experiment (GRACE) and Global Land Data Assimilation System (GLDAS) data from 1° (~110 km) and 0.25° (~25 km) respectively, to 1 km for China. Three evaluation metrics were employed for the testing dataset for 2004−2016: The R2 ranged from 0.77−0.89 for XGBoost (0.74−0.86 for RF), the correlation coefficient (CC) ranged from 0.88−0.94 for XGBoost (0.88−0.93 for RF) and the root-mean-square error (RMSE) ranged from 0.37−2.3 for XGBoost (0.4−2.53 for RF). The R2 of the XGBoost models for GLDAS was 0.64−0.82 (0.63−0.82 for RF), the CC was 0.80−0.91 (0.80−0.90 for RF) and the RMSE was 0.63−1.75 (0.63−1.77 for RF). The downscaled GWS derived from GRACE and GLDAS were validated using in situ measurements by comparing the time series variations and the downscaled products maintained the accuracy of the original data. The interannual changes within 9 river basins between pre- and post-downscaling were consistent, emphasizing the reliability of the downscaled products. Ultimately, annual downscaled TWS, GLDAS and GWS products were provided from 2004 to 2016, providing a solid data foundation for studying local GWS changes, conducting finer-scale hydrological studies and adapting water resources management and policy formulation to local condition.

Improving the Resolution and Accuracy of Groundwater Level Anomalies Using the Machine Learning-Based Fusion Model in the North China Plain

The launch of GRACE satellites has provided a new avenue for studying the terrestrial water storage anomalies (TWSA) with unprecedented accuracy. However, the coarse spatial resolution greatly limits its application in hydrology researches on local scales. To overcome this limitation, this study develops a machine learning-based fusion model to obtain high-resolution (0.25°) groundwater level anomalies (GWLA) by integrating GRACE observations in the North China Plain. Specifically, the fusion model consists of three modules, namely the downscaling module, the data fusion module, and the prediction module, respectively. In terms of the downscaling module, the GRACE-Noah model outperforms traditional data-driven models (multiple linear regression and gradient boosting decision tree (GBDT)) with the correlation coefficient (CC) values from 0.24 to 0.78. With respect to the data fusion module, the groundwater level from 12 monitoring wells is incorporated with climate variables (precipitation, runoff, and evapotranspiration) using the GBDT algorithm, achieving satisfactory performance (mean values: CC: 0.97, RMSE: 1.10 m, and MAE: 0.87 m). By merging the downscaled TWSA and fused groundwater level based on the GBDT algorithm, the prediction module can predict the water level in specified pixels. The predicted groundwater level is validated against 6 in-situ groundwater level data sets in the study area. Compare to the downscaling module, there is a significant improvement in terms of CC metrics, on average, from 0.43 to 0.71. This study provides a feasible and accurate fusion model for downscaling GRACE observations and predicting groundwater level with improved accuracy.

Statistical Applications to Downscale GRACE-Derived Terrestrial Water Storage Data and to Fill Temporal Gaps

The Gravity Recovery and Climate Experiment (GRACE) has been successfully used to monitor variations in terrestrial water storage (GRACETWS) and groundwater storage (GRACEGWS) across the globe, yet such applications are hindered on local scales by the limited spatial resolution of GRACE data. Using the Lower Peninsula of Michigan as a test site, we developed optimum procedures to downscale GRACE Release-06 monthly mascon solutions. A four-fold exercise was conducted. Cluster analysis was performed to identify the optimum number and distribution of clusters (areas) of contiguous pixels of similar geophysical signals (GRACETWS time series); three clusters were identified (cluster 1: 13,700 km2; cluster 2: 59,200 km2; cluster 3: 33,100 km2; Step I). Variables (total precipitation, normalized difference vegetation index (NDVI), snow cover, streamflow, Lake Michigan level, Lake Huron level, land surface temperature, soil moisture, air temperature, and evapotranspiration (ET)), which could potentially contribute to, or correlate with, GRACETWS over the test site were identified, and the dataset was randomly partitioned into training (80%) and testing (20%) datasets (Step II). Multivariate regression, artificial neural network, and extreme gradient boosting techniques were applied on the training dataset for each of the identified clusters to extract relationships between the identified hydro-climatic variables and GRACETWS solutions on a coarser scale (13,700–33,100 km2), and were used to estimate GRACETWS at a spatial resolution matching that of the fine-scale (0.125° × 0.125° or 120 km2) inputs. The statistical models were evaluated by comparing the observed and modeled GRACETWS values using the R-squared, the Nash–Sutcliffe model efficiency coefficient (NSE), and the normalized root-mean-square error (NRMSE; Step III). Lastly, temporal variations in GRACEGWS were extracted using outputs of land surface models and those of the optimum downscaling methodology (downscaled GRACETWS) (Step IV). Findings demonstrate that (1) consideration should be given to the cluster-based extreme gradient boosting technique in downscaling GRACETWS for local applications given their apparent enhanced performance (average value: R-squared: 0.86; NRMSE 0.37; NSE 0.86) over the multivariate regression (R-squared: 0.74; NRMSE 0.56; NSE 0.64) and artificial neural network (R-squared: 0.76; NRMSE 0.5; NSE 0.37) methods; and (2) identifying local hydrologic variables and the optimum downscaling approach for individual clusters is critical to implementing this method. The adopted method could potentially be used for groundwater management purposes on local scales in the study area and in similar settings elsewhere.

Downscaling of GRACE-Derived Groundwater Storage Based on the Random Forest Model

Groundwater is an important part of water storage and one of the important sources of agricultural irrigation, urban living, and industrial water use. The recent launch of Gravity Recovery and Climate Experiment (GRACE) Satellite has provided a new way for studying large-scale water storage. The application of GRACE in local water resources has been greatly limited because of the coarse spatial resolution, and low temporal resolution. Therefore, it is of great significance to improve the spatial resolution of groundwater storage for regional water management. Based on the method of random forest (RF), this study combined six hydrological variables, including precipitation, evapotranspiration, runoff, soil moisture, snow water equivalent, and canopy water to conduct downscaling study, aiming at downscaling the resolution of the total water storage and groundwater storage from 1° (110 km) and to 0.25° (approximately 25 km). The results showed that, from the perspective of long time series, the prediction results of the RF model are ideal in the whole research area and the observations wells area. From the perspective of space, the detailed changes of water storage could be captured in greater detail after downscaling. The verification results show that, on the monthly scale and annual scale, the correlation between the downscaling results and the observation wells is 0.78 and 0.94, respectively, and they both reach the confidence level of 0.01. Therefore, the RF downscaling model has great potential for predicting groundwater storage.

Spatial Downscaling of GRACE TWSA Data to Identify Spatiotemporal Groundwater Level Trends in the Upper Floridan Aquifer, Georgia, USA

Accurate assessments of groundwater resources in major aquifers across the globe are crucial for sustainable management of freshwater reservoirs. Observations from the Gravity Recovery and Climate Experiment (GRACE) satellite have become invaluable as a means to identify regions groundwater change. While there is a large body of research that focuses on downscaling coarse (1°) GRACE products, few studies have attempted to spatially downscale GRACE to produce fine resolution (5 km) maps that are more useful to resource managers. This study trained a boosted regression tree model to statistically downscale GRACE total water storage anomaly to monthly 5 km groundwater level anomaly maps in the karstic upper Floridan aquifer (UFA) using multiple hydrologic datasets. Evaluation of spatial predictions with existing groundwater wells indicated satisfactory performance (R = 0.79, NSE = 0.61). Results demonstrate that groundwater levels were stable between 2002–2016 but varied seasonally. The data also highlights areas where groundwater pumping is exacerbating UFA water-level declines. While results demonstrate the applicability of machine learning based methods for spatial downscaling of GRACE data, future studies should account for preferential flowpaths (i.e., conduits, lineaments) in karstic systems.

An Improved GPS-Inferred Seasonal Terrestrial Water Storage Using Terrain-Corrected Vertical Crustal Displacements Constrained by GRACE

Based on a geophysical model for elastic loading, the application potential of Global Positioning System (GPS) vertical crustal displacements for inverting terrestrial water storage has been demonstrated using the Tikhonov regularization and the Helmert variance component estimation since 2014. However, the GPS-inferred terrestrial water storage has larger resulting amplitudes than those inferred from satellite gravimetry (i.e., Gravity Recovery and Climate Experiment (GRACE)) and those simulated from hydrological models (e.g., Global Land Data Assimilation System (GLDAS)). We speculate that the enlarged amplitudes should be partly due to irregularly distributed GPS stations and the neglect of the terrain effect. Within southwest China, covering part of southeastern Tibet as a study region, a novel GPS-inferred terrestrial water storage approach is proposed via terrain-corrected GPS and supplementary vertical crustal displacements inferred from GRACE, serving as "virtual GPS stations" for constraining the inversion. Compared to the Tikhonov regularization and Helmert variance component estimation, we employ Akaike’s Bayesian Information Criterion as an inverse method to prove the effectiveness of our solution. Our results indicate that the combined application of the terrain-corrected GPS vertical crustal displacements and supplementary GRACE spatial data constraints improves the inversion accuracy of the GPS-inferred terrestrial water storage from the Helmert variance component estimation, Tikhonov regularization, and Akaike’s Bayesian Information Criterion, by 55%, 33%, and 41%, respectively, when compared to that of the GLDAS-modeled terrestrial water storage. The solution inverted with Akaike’s Bayesian Information Criterion exhibits more stability regardless of the constraint conditions, when compared to those of other inferred solutions. The best Akaike’s Bayesian Information Criterion inverted solution agrees well with the GLDAS-modeled one, with a root-mean-square error (RMSE) of 3.75 cm, equivalent to a 15.6% relative error, when compared to 39.4% obtained in previous studies. The remaining discrepancy might be due to the difference between GPS and GRACE in sensing different surface water storage components, the remaining effect of the water storage changes in rivers and reservoirs, and the internal error in the geophysical model for elastic loading.

Estimating High-Resolution Groundwater Storage from GRACE: A Random Forest Approach

Gravity Recovery and Climate Experiment (GRACE) data have become a widely used global dataset for evaluating the variability in groundwater storage for the different major aquifers. Moreover, the application of GRACE has been constrained to the local scale due to lower spatial resolution. The current study proposes Random Forest (RF), a recently developed unsupervised machine learning method, to downscale a GRACE-derived groundwater storage anomaly (GWSA) from 1° × 1° to 0.25° × 0.25° in the Northern High Plains aquifer. The RF algorithm integrated GRACE to other satellite-based geospatial and hydro-climatological variables, obtained from the Noah land surface model, to generate a high-resolution GWSA map for the period 2009 to 2016. This RF approach replicates local groundwater variability (the combined effect of climatic and human impacts) with acceptable Pearson correlation (0.58 ~ 0.84), percentage bias (−14.67 ~ 2.85), root mean square error (15.53 ~ 46.69 mm), and Nash-Sutcliffe efficiency (0.58 ~ 0.84). This developed RF model has significant potential to generate finer scale GWSA maps for managing groundwater at both local and regional scales, especially for areas with sparse groundwater monitoring wells.

Downscaling GRACE TWSA Data into High-Resolution Groundwater Level Anomaly Using Machine Learning-Based Models in a Glacial Aquifer System

With continued threat from climate change and human impacts, high-resolution and continuous hydrologic data accessibility has a paramount importance for predicting trends and availability of water resources. This study presents a novel machine learning (ML)-based downscaling algorithm that produces a high spatial resolution groundwater level anomaly (GWLA) from the Gravity Recovery and Climate Experiment (GRACE) data by utilizing the relationship between Terrestrial Water Storage Anomaly (TWSA) from GRACE and other land surface and hydro-climatic variables (e.g., vegetation coverage, land surface temperature, precipitation, streamflow, and in-situ groundwater level data). The predicted downscaled GWLA data were tested using monthly in-situ groundwater level observations. Of the 32 groundwater monitoring wells available in the study site, 21 wells were used to develop the ML-based downscaling model, while the remaining 11 wells were used to assess the performance of the ML-based downscaling model. The test results showed that the model satisfactorily reproduces the spatial and temporal variation of the GWLA in the area, with acceptable correlation coefficient and Nash-Sutcliffe Efficiency values of ~0.76 and ~0.45, respectively. GRACE TWSA was the most influential predictor variable in the models, followed by stream discharge and soil moisture storage. Though model limitations and uncertainty could exist due to high spatial heterogeneity of the geologic materials and omission of human impact (e.g., abstraction), the significance of the result is undeniable, particularly in areas where in-situ well measurements are sparse.

A Multi-Sourced Data Retrodiction of Remotely Sensed Terrestrial Water Storage Changes for West Africa

Remotely sensed terrestrial water storage changes (TWSC) from the past Gravity Recovery and Climate Experiment (GRACE) mission cover a relatively short period (≈15 years). This short span presents challenges for long-term studies (e.g., drought assessment) in data-poor regions like West Africa (WA). Thus, we developed a Nonlinear Autoregressive model with eXogenous input (NARX) neural network to backcast GRACE-derived TWSC series to 1979 over WA. We trained the network to simulate TWSC based on its relationship with rainfall, evaporation, surface temperature, net-precipitation, soil moisture, and climate indices. The reconstructed TWSC series, upon validation, indicate high skill performance with a root-mean-square error (RMSE) of 11.83 mm/month and coefficient correlation of 0.89. The validation was performed considering only 15% of the available TWSC data not used to train the network. More so, we used the total water content changes (TWCC) synthesized from Noah driven global land data assimilation system in a simulation under the same condition as the GRACE data. The results based on this simulation show the feasibility of the NARX networks in hindcasting TWCC with RMSE of 8.06 mm/month and correlation coefficient of 0.88. The NARX network proved robust to adequately reconstruct GRACE-derived TWSC estimates back to 1979.

Estimating Groundwater Abstractions at the Aquifer Scale Using GRACE Observations

Groundwater monitoring requires costly in situ networks, which are difficult to maintain over long time periods, especially in countries facing economic recession such as Greece. Our work aims at providing a methodology to estimate groundwater abstractions at the aquifer scale using publicly available remotely sensed data from the NASA’s Gravity Recovery and Climate Experiment (GRACE) together with publicly available meteorological observations that serve as input variables to an Artificial Neural Network (ANN) method. The methodology was demonstrated in an alluvial aquifer in NE Greece for a 10-year period (2005–2014), where irrigation agriculture poses a serious threat to both groundwater resources and their dependent ecosystems. To generalize the developed model, an ensemble of 100 ANNs was created by the initial weight randomization approach and output was computed by averaging the output of each individual model. Scaled Root Mean Square Error and Nash–Sutcliffe coefficient were used to test the model efficiency. Both of these performance metrics indicated that monthly groundwater abstractions can be estimated efficiently and that the developed methodology offers an inexpensive substitute for in situ groundwater monitoring when in situ networks are not available or cannot operate properly.