Groundwater is one of the most important freshwater resources in arid and semi-arid regions. Especially in the Yellow River Basin, groundwater has long been subject to intensive exploitation, highlighting the urgent need for a high-resolution groundwater monitoring and management system. In this study, based on the GRACE gravity satellite observations and related hydrological components, we obtained the groundwater storage data in the Yellow River Basin with a resolution of 0.25°. To improve the spatial resolution, a downscaling framework integrating Partial Least Squares Regression (PLSR) for factor selection and the Extreme Gradient Boosting (XGBoost) model was developed, resulting in a monthly groundwater storage dataset in the Yellow River Basin with a spatial resolution of 0.1° from 2003 to 2023. Meanwhile, the accuracy of this dataset was systematically validated from multiple dimensions. The results show that this dataset exhibits a high level of accuracy in both model performance and validation against observational data, effectively improving the spatial resolution of groundwater storage data and enabling a more precise characterization of ground dynamics in the Yellow River Basin. This dataset overcomes the limitation of the 0.25° spatial resolution of the original GRACE gravity satellite data. While retaining its advantage in large-scale monitoring, it enables higher-accuracy dynamic analysis of groundwater. This dataset provides important data support for groundwater resource monitoring and management in the Yellow River Basin, and is expected to help promote the scientific management and sustainable utilization of groundwater resources in the region.
XIE et al. (Sun,) studied this question.