No abstract

Terrestrial water storage (TWS), which includes ice and snow, lakes, rivers, soil water and groundwater, is a key parameter to close the water balance at global river basins (e.g., Liu et al., 2016). Moreover, the changes in TWS (∆S) and its components present considerable spatiotemporal heterogeneity at the global scale (Frappart & Ramillien, 2018). The traditional methods of estimating ∆S, including in situ observations and hydrological modeling, have their own limitations in different aspects. First, in situ observations are limited to the point scale and it is difficult to obtain ΔS observations in terrain with complicated topography (A. Wang & Zeng, 2012). Second, uncertainties in models and the lack of observation data in some areas will lead to large variations in the accuracy of model simulation results. Although microwave remote sensing could retrieve the TWS components, it is restricted to obtaining information about changes in soil moisture content only at depths up to 5 cm from the surface. The Gravity Recovery and Climate Experiment (GRACE) twin satellite mission, jointly implemented by the National Aeronautics and Space Administration (NASA), and German Aerospace Center, is primarily aimed at obtaining temporal variations in the Earth's gravity field with unprecedented accuracy (e.g., Save et al., 2016). GRACE not only effectively avoids the shortcomings of traditional methods, but also provides a new tool for directly monitoring large-scale ΔS. However, GRACE has its own limitations. In addition to the coarse spatial and temporal resolution of the raw data, it is difficult to trace