Identification and characteristics of regional rainstorm events in China based on a dual-threshold method

Abstract

Regional extreme precipitation can cause severe flooding with major environmental and social impacts. While most existing studies either ignored their spatiotemporally contiguous or abnormally break down the events with multiple rainstorm cores, rapid movement, or transient interruptions into multiple isolated fragments. Here, we propose a novel approach for identifying regional rainstorm events (RREs), which integrates dual-threshold with a three-dimensional (longitude, latitude, and time) image connectivity perspective. There were 1163 RREs identified across mainland China from 1991 to 2022 based on the CN05.1 dataset. The spatial, temporal, and movement patterns of RREs were analyzed. Results showed that 71.2 % of RREs lasted only one day, with intensity mostly around 60–70 mm/day. The cumulative magnitude of RREs was most positively correlated with its affected area (r = 0.95). The occurrence frequency of RREs had a significant upward trend during 1991–2022, approximately increased 2.7 events per decade, while the severity slightly declined. Spatially, the highest frequencies of RREs were observed in southeastern China. Based on the movement direction and distance, RREs were classified into four patterns: Eastward Moving, Southwestward Moving, Localized, and Northward Moving. Northward Moving RREs exhibited the largest affected area, the greatest cumulative magnitude, and the longest movement distance, followed by Eastward Moving RREs. Localized RREs had the shortest duration, the smallest affected area and movement distance, but the highest intensity (averaging 92.6 mm/day). Our findings provide new insights for understanding the characteristics of regional extreme precipitation and their driving mechanisms, which are valuable for forecasting and adapting to the resultant disasters.