Nonlinear spatiotemporal and evolution characteristics of extreme precipitation during the Mei-yu period based on a state transition network

Abstract

The frequent occurrence of extreme precipitation events notably affects the middle and lower reaches of the Yangtze River (MLYR) during the Mei-yu period. Based on meteorological station data from the China Meteorological Administration collected from 1970 to 2019, we employed nonlinear time series analysis and the state transition networks (STN) to identify the spatiotemporal and evolution characteristics of 10 extreme precipitation indices (EPIs). Although previous studies have largely focused on investigating the spatiotemporal characteristics, the suitability of STN for extreme precipitation has not been extensively explored. This study focused on not only the spatiotemporal characteristics of extreme precipitation but also its evolution using the STN. The results showed that the periodic oscillations in extreme precipitation ranged from 2 to 5 years and an abrupt change occurred primarily around 1990. Extreme precipitation displayed similar network parameters, yet different network structures and transition probabilities. The evolution of extreme precipitation exhibited from decreasing to increasing from 1970 to 2019, with an increasing trend after 1990. Similar network structures tended to exhibit higher Pearson correlations. Spatially, high EPIs values were mainly concentrated in the central region of the MLYR. This study could provide new methods and perspectives for studying the characteristics of extreme precipitation.

Keywords

• Extreme precipitation index
• Mei-yu period
• state transition network
• nonlinear time series analysis