Abstract: Most previous studies on drought assessment were limited to static risks, but to practical drought prevention and drought resistance, it is more significant to reveal the dynamic evolution characteristics of drought multi-attribute risks under changing environments. This study calculates the joint return periods of moderate and severe drought events in the Wei River basin using Copula functions, and examines the dynamic characteristics of drought multi-attribute risks based on a sliding window of 31-year time interval. The driving force of drought dynamic evolution is explored using a cross wavelet transform. Results indicate that (1) the drought risk of the Jing River is the highest in the whole basin, and its static joint return period of "and" type is 11.4 and 21.6 years under two drought scenarios, respectively; (2) the drought multi-attribute risk in the basin has become generally higher since 1980s; (3) the drought risk in the basin is closely related to sunspots. The correlation between sunspots and drought index series was subjected to significant phase changes in the 1980s, which could be one of the major reasons for the increased drought risk.

Key words: drought, multi-attribute, Copula functions, joint return period, dynamic characteristic, Wei River basin