Probability of rainstorm and flood disasters due to extreme precipitation in Fen River Basin, China

Abstract

Analysis of the probability of extreme precipitation events leading to rainstorm and flood disasters can aid in disaster prevention policy development. Using daily precipitation data from 16 meteorological stations from 1960 to 2019, we calculated eight extreme precipitation indices to analyze the spatio-temporal characteristics of extreme precipitation in the Fen River Basin (FRB) through ensemble empirical mode decomposition and Kriging interpolation. Extreme precipitation events and disasters were defined based on a combination of the antecedent precipitation index (API) and extreme precipitation on the event day and classified; extreme precipitation and the API were ranked from small to large and classified into dry, wet, and moderate (mod) precipitation periods, respectively, yielding nine extreme precipitation event categories. The probability of disasters caused by different types of extreme precipitation events was calculated using a binomial distribution. The results are as follows: (1) between 1960 and 2019, except for extreme precipitation period length, which continuously increased, the extreme precipitation indices changed from a downward to an upward trend since the 1980s. All extreme precipitation indices showed similar interannual variation over short periods and different interdecadal variation over long periods. (2) The extreme precipitation indices showed latitudinal and zonal spatial divergence patterns, but different spatial characteristics were observed around the 1980s. (3) More than 70% of extreme precipitation events in the midstream and downstream fell into four categories: "dry-dry," "dry-mod," "mod-dry," and "mod-mod." (4) A single category VII (VIII) extreme precipitation event in the midstream (downstream) had a maximum probability of causing disaster of 14%. When more than four extreme precipitation events occurred in a year, the probability of one disaster was the highest and that of four or more disasters was < 0.1%. The probability of rainstorm and flood disasters increased gradually with increasing frequency of annual extreme precipitation events.

Keywords: Antecedent precipitation index; Binomial distribution; Extreme precipitation; Probability; Rainstorm and flood disaster.