Analysis of Extreme Precipitation under Nonstationary Conditions in the Yangtze River Basin

As global warming intensifies, both the frequency and magnitude of extreme precipitation events have undergone significant changes, leading to a widespread manifestation of nonstationary characteristics in extreme precipitation sequences. Regarding the current research status in the Yangtze River Basin (YRB), there is a deficiency in comprehensive investigations into the connection between extreme precipitation and climate change. This paper aims to identify the primary climate factors affecting extreme precipitation in the YRB and use them for dynamic assessment of future rainstorm risks. It thoroughly examines the correlation between 11 climate factors and extreme rainfall across different temporal scales to identify potential climate drivers for nonstationary models. Subsequently, ten nonstationary generalized extreme value (GEV) models were constructed, and the optimal covariates for 190 stations were determined within a Bayesian framework. Finally, a complete forecasting process was established, and the extreme rainfall in the YRB under nonstationary conditions was assessed. The results indicate that, on average, 144 stations exhibited nonstationary characteristics across various event scales, accounting for an average of 75.8%. The extreme rainfall in the YRB is influenced by multiple factors, with no single factor demonstrating absolute dominance across different time scales. Neglecting the nonstationarity of extreme precipitation will inevitably lead to misconceptions about extreme rainfall. Furthermore, the spatial distribution of the maximum scenarios of extreme rainfall under nonstationary conditions is significantly different from that under stationary conditions.