Abstract
With the increasing frequency and intensity of extreme precipitation events globally, it is imperative to study the characteristics of extreme precipitation for disaster prevention and mitigation. Traditional statistical analyses to study extreme precipitation often tend to separate temporal and spatial features, and spatial features are often studied based on complex networks, with few studies conducted from a combined temporal and spatial perspective. In this study, we aimed to analyze the spatiotemporal evolution characteristics of extreme precipitation based on visible graph network and state transition networks at different percentile thresholds, and to analyze and compare with the topology, network type, anomalous year, trend of the change stage, key modes, and the future evolution trend. The results indicated that the extreme precipitation networks exhibited high clustering coefficients and short average path lengths, identifying the key year around 1990, and the community was divided into 4-6 stages with an overall increasing trend. The Hangzhou-Wuhan dual-core driving mode, and displayed a consistent trend in the future. This study innovatively reveals the spatiotemporal evolution characteristics of extreme precipitation from the perspective of complex networks, which provides a reference point and basis for preventing, mitigating and predicting extreme precipitation.