Abstract: Drought is a disruptive natural hazard that profoundly impacts agriculture and ecosystems, it is important to investigate and assess spatiotemporal characteristics of drought events. Based on the daily soil moisture datasets with a spatial resolution of 50 km × 50 km simulated by the Variable Infiltration Capacity(VIC) land surface macroscale hydrology model, the soil moisture anomaly percentage index(SMAPI) was calculated and used to reconstruct historical drought events and investigation of drought spatiotemporal characteristics between 1953 and 2013 in the upper reaches of the Yangtze River basin．Some important results are as follows:(1) Monthly discharge variable processes at Yichang section simulated by the VIC model were relatively consistent with observed processes during 1980-2000, the results showed that the VIC model built in the study area was feasible and could be used to simulate soil moisture. In addition to good performance in reconstructing historical drought events in the upper reaches of the Yangtze River, SMAPI can also describe drought occurrence, development and severity of largescale droughts in time and space. SMAPI has clear physical meanings and the drought events identified by SMAPI can be used to analyze the spatiotemporal characteristics of historical drought events.(2) A total of 12 regional drought events were identified in the upper reaches of the Yangtze River basin over the past 61 years, the first ranking severe drought event was from March 4, 1969 to July 5, 1969(124 days) and areal extent was 51.3%. On average, the drought duration, areal extent and drought intensity were 90 days(about 3 months), 48.8% and -11.1% respectively in the upper reaches of the Yangtze River basin from 1953 to 2013.(3) The relationship between drought area and drought intensity for different drought durations were analyzed. The results showed that drought intensity strengthen as drought area increases for the same drought duration. The research results can provide some theoretical values for the comprehensively understanding spatialtemporal features of historical drought events and a valuable reference for regional drought prevention and resistance in the upper reaches of the Yangtze River basin.
Key words：VIC model; soil moisture anomaly percentage index; spatiotemporal characteristics; the upper reaches of the Yangtze River basin