Abstract: Abstract: The Qinghai-Tibet Plateau has been greatly sensitive to global climate change in recent years, particularly the change in water storage. However, the differences in multi-source data and calculation have caused great uncertainty in the accurate estimation of terrestrial water storage. Taking the Yarlung Zangbo River Basin located in the southeast of the Qinghai-Tibet Plateau as the study area, this study aims to investigate the spatio-temporal variation characteristics of terrestrial water storage change (TWSC). The performance evaluation of TWSC was conducted using the GRACE (Gravity Recovery and Climate Experiment) data from 2003 to 2017, combining with ERA5 (the fifth-generation reanalysis product of the European Centre for Medium-Range Weather Forecasts) data and GLDAS (Global Land Data Assimilation System) data. Four sets of TWSC were obtained using ERA5 data and GLDAS data from 2003 to 2017, where the terrestrial water balance (PER) and the summation (SS) were used to estimate TWSC, compared with TWSC derived from GRACE. It was found that TWSC estimated by the SS was more consistent than the PER in both temporal and spatial patterns with TWSC derived from GRACE, whereas, TWSC estimated by the SS using GLDAS was most consistent with the results from GRACE over the Yarlung Zangbo River Basin. A sensitivity analysis was carried out using GLDAS and the SS, in order to further investigate the influence of different components on the estimation of TWSC. It was found that SS (GLDAS) performed best for TWSC estimation over the Yarlung Zangbo River Basin, when only soil moisture and snow water equivalent were taken into consideration, with the basin average TWSC correlation coefficient of 0.53. Therefore, the SS calculated from the soil moisture and snow water equivalent was utilized to further explore the long-term temporal and spatial evolution of TWSC from 1948 to 2017 over the Yarlung Zangbo River Basin. TWSC showed a significantly increasing trend from 1948 to 2017. The wavelet analysis indicated that there was an abrupt change in TWSC around 2002 during 1948-2017. There was a significant increase before 2002 (slope=0.024 mm/month, P<0.01), and a significant decrease (slope=?0.397 mm/month, P<0.01) after 2002. The change trend of soil water content and snow water equivalent before and after 2002 is consistent with TWSC. Additional attribution analysis was performed on the abrupt change in TWSC before and after 2002. The contributions of changes in soil moisture and snow water equivalent to TWSC were 61% and 39% during 1948-2002, respectively, whereas, a much more significant domination of soil moisture was identified during 2003-2017. The contributions of changes in soil moisture and snow water equivalent to TWSC were 99% and 1%, respectively. Compared with the change in snow water equivalent, the change in soil moisture dominated the long-term variation of TWSC during 1948-2017. In terms of spatial distribution, TWSC showed a large spatial heterogeneity, mainly in the middle reaches with a high intensity of human activities and the Yarlung Zangbo River Basin with intensive glaciers. The finding can provide reliable data support to explore the spatio-temporal evolution of terrestrial water storage in data-scarce alpine regions.