“…The other group is to establish the relationship between different driving factors and TWSA through mathematical or statistical methods to isolate the effects of different factors. This type of method can be divided into the following three subgroups according to previous studies: (a) treating various influencing factors as independent variables, including temperature, precipitation, groundwater level, urbanization level, etc., and counting the contribution share of various factors in the change of water storage by different statistical methods such as multiple linear regression, regression subset selection approach, dominance analysis, gray relation analysis, and correlation analysis (H. Chen et al., 2020; Cui et al., 2022; H. J. Deng & Chen, 2017; L. Deng et al., 2022; T. F. Feng et al., 2022; K. Liu et al., 2021; Thomas & Famiglietti, 2019); (b) establishing the relationship between annual precipitation anomaly and annual TWSC through linear regression to represent climate‐driven changes in TWS (Yi et al., 2016); (c) reconstructing climate‐driven TWSA by a statistical model which relates precipitation and temperature to TWSA (Humphrey & Gudmundsson, 2019; B. Liu et al., 2021; Zhong et al., 2019). We summarize above separation methods in Figure A1 to make it easier to understand.…”