Monitoring the spatiotemporal terrestrial water storage changes in the Yarlung Zangbo River Basin by applying the P-LSA and EOF methods to GRACE data.

The Yarlung Zangbo River Basin is a regulator of water vapor changes in China and even Asia. To avoid the shortcomings of traditional water resource monitoring methods, this study used Gravity Recovery and Climate Experiment (GRACE) data to monitor the terrestrial water storage anomaly (TWSA) in this river from 2002 to 2015 with the help of the polynomial-least squares approach (P-LSA) and the empirical orthogonal function (EOF). The obtained TWSA was compared with hydrometeorological data from several sources to discuss the applicability, uniqueness and response relationship. The results showed that (1) the combination of P-LSA and EOF had strong applicability to explore the TWSA in the study area, with R = 0.75 and 0.80, respectively, and could indirectly reflect dry and wet conditions in southwestern China. (2) The TWSA revealed significant cyclical and seasonal fluctuations of approximately 12 months and increased from upstream to downstream and from north to south, which was discussed for the first time in the research area. (3) The EOF method can effectively identify the TWSA principal component and structure (EOF1 contribution = 91.08%) by removing noise and redundancy, which is beneficial for revealing the laws essential for TWSA changes. (4) The TWSA in the studied watershed was unique (i.e., the clearest periodic changes with the best fitting effect (R = 0.90); peak, low and peak-low difference values that were 1.82, 1.19 and 1.52 times larger than those of the 8 other rivers; and the largest downward trend of 4.13 mm·a). (5) Rainfall was the decisive factor influencing the TWSA, with correlation coefficients (R) >0.60. This study enhances our overall understanding of the TWSA in this plateau watershed and provides a scientific basis for optimal water resource management.