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

Abstract

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.

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

“…However, the highest resolution of predicted results is mainly determined by the resolution of climate variables (Module #1) and water storage estimates (Module #3). The water balance variables are the most widely used in previous studies [ 31 , 37 ], provided at the maximum resolution of 0.25°. Similarly, the water storage components are obtained from the GLDAS-Noah model, which provides simulated outputs at the resolutions of 1° and 0.25°.…”

“…Similarly, some tree-based machine learning algorithms (e.g., random forest (RF) and gradient boosting decision tree (GBDT)) become popular in regression tasks with the advantages of simplicity and effectiveness. The RF algorithm has been utilized to downscale GRACE observations and obtained satisfactory results in some areas [ 37 , 38 ]. As a kind of ensemble machine learning algorithm, GBDT performs well in constructing non-linear regression models, which is often employed to forecast ET [ 39 ] and urban flood [ 40 , 41 ], but rarely in GWLA.…”

“…The North China Plain (NCP), which is the political, economic, and cultural center of China [ 43 ], has been suffering from water shortage and over-exploitation of groundwater for a long period [ 44 , 45 , 46 ]. Some downscaling researches have been conducted in the area, aiming to provide high-resolution water storage estimates [ 33 , 37 , 47 ]. Water resources managers are often more concerned with information about water levels at specified locations, while few studies are conducted with respect to this aspect.…”

Self‐Validating Deep Learning for Recovering Terrestrial Water Storage From Gravity and Altimetry Measurements

“…Downscaling approaches try to recover or predict TWS components, for example, ground water level and storage (Chen et al, 2019; Seyoum et al, 2019; Seo & Lee, 2019) or watersheds (Ahmed et al, 2019), from GRACE observations and auxiliary data products, including precipitation, land surface temperatures, and vegetation cover. In related work, machine learning is also used for the removal of correlated errors in GRACE data (Piretzidis et al, 2018) and for reconstructing missing TWS observations between GRACE and GRACE‐Follow On (Sun, Long, et al, 2020).…”

A Self‐Calibration Variance‐Component Model for Spatial Downscaling of GRACE Observations Using Land Surface Model Outputs

“…Seyoum et al (2019) tried to downscale GRACE-derived Total Water Storage Anomaly (GRACE TWSA) into high-resolution groundwater level anomaly using machine learning-based models in a glacial aquifer system. Chen et al (2019) evaluated downscaling of GRACE-derived GWS based on the random forest model. Sahour et al (2019) developed optimum procedures to downscale GRACE Release-06 monthly mascon solutions based on statistical applications.…”