Abstract: Hydrological drought is jointly affected by climate change and human activities. Revealing the evolution characteristics of hydrological drought and its driving factors in the changing environment will contribute to improving the capacities of drought control and drought resistance. Based on the long time series of meteorological and hydrological data of 1960-2020 from the Upper Yangtze River Basin (UYRB), the present study first generates naturalized runoff time series using multi-model ensemble simulation method, and uses the reconstructed natural runoff to calculate the traditional standardized runoff index (SRIr) for characterizing the hydrological drought under natural conditions. Then, we simulate the time-dependent standardized runoff index (SRIt) using the generalized additive model for location, scale and shape (GAMLSS) with time as the covariate to represent the hydrological drought under non-stationary environment. Finally, the impacts of climate change and human activities on hydrological drought are distinguished quantitatively by comparative analysis of SRIr and SRIt series. The results show the overall trend of the hydrological drought evolution in this basin under natural conditions is intensified and has been further aggravated by human activities. The dominant factors of its hydrological drought evolution present obvious temporal and spatial differences: on the annual scale, climate change is the dominant factor in the basins of the Jinsha River, Tuo River and Wu River, while human activities are dominant in the Min River, Jialing River, and the whole upper Yangtze basin. Dominant factors of hydrological drought evolution on the seasonal scale are not completely consistent with those on the annual scale.

Key words: hydrological drought, climate change, human activities, ensemble simulation, non-stationary, upper Yangtze River basin