Spatial patterns of typhoon rainfall and associated flood characteristics over a mountainous watershed of a tropical island

The spatial distribution of rainfall in a watershed has an important impact on flood warning and forecasting. In this study, we used the K-means++ clustering method to classify the spatial characteristics of typhoon rainfall events in the Songtao reservoir watershed in the central region of Hainan Island, based on 63 typhoon rainfall event data from 1990 to 2020. Then, the relationship between the synoptic fields, spatial patterns of rainfall, and flood characteristics was explained based on tropical cyclone best-track datasets, atmospheric reanalysis data, and inflow flood series. The results showed that there are three types of cumulative rainfall patterns in the Songtao Reservoir watershed. The heavy rainfall region of pattern 1 is concentrated east of the watershed, whereas the heavy rainfall regions of pattern 2 are mainly concentrated southwest and east of the watersheds of many mountains. The heavy rainfall region of pattern 3, with a southwest-northeast orientation, is located on the west side of the watershed. The spatial coefficient of variation and index of rainfall location clearly distinguished among these three patterns. The composite analysis of typhoon tracks, wind fields, and moisture flux fields can be used to interpret the formation mechanism of these spatial patterns of rainfall events in the watershed. The spatial patterns of cumulative rainfall for each type were dominated by one or two types of hourly spatial rainfall patterns. Flood events resulting from cumulative rainfall patterns 1 and 2 have similar flood behavior metrics. Large floods in the watershed show an obvious seasonal variation, and the corresponding rainfall events mostly show the spatial characteristics of the cumulative rainfall pattern 3. These results can enhance the understanding of the linkage between typhoons, mountainous watershed rainfall, and flooding processes.

Automated summary

The spatial distribution of rainfall in a watershed has a significant impact on flood warning and forecasting. This study classified the spatial characteristics of typhoon rainfall events in the Songtao reservoir watershed using the K-means++ clustering method. The results showed three types of cumulative rainfall patterns in the watershed, with different flood characteristics. The composite analysis of typhoon tracks, wind fields, and moisture flux fields helped to explain the formation mechanism of these spatial patterns of rainfall events.