Non-stationary fitting of extreme precipitation in the upper reaches of the Yangtze River and analysis of its disaster causing effects

Abstract

[Objective] In recent years, extreme precipitation events occurring in changing environments have become increasingly frequent, and the assumption of stationary in traditional frequency analysis has been gradually questioned. Therefore, it is necessary to reasonably analyze the non-stationary characteristics of extreme precipitation in the basin, which will help to dynamically assess the risk of rainstorm and improve the disaster prevention and mitigation capacity of the basin. [Methods] Based on the GAMLSS model, a time-varying distribution model with time as the explanatory variable is established for the extreme precipitation frequency index and intensity index of 82 meteorological stations in the upper reaches of the Yangtze River Basin. Then, the Mann Kendall method is used to analyze the trend of the extreme precipitation frequency index and intensity index series, and the disaster causing effect of extreme precipitation in the upper reaches of the Yangtze River Basin is evaluated.[Results] The result indicate that:(1) the extreme precipitation index of most stations in the upper reaches of the Yangtze River basin exhibits non-stationary characteristics. Among them, more than 70% of stations have non-stationary characteristics for moderate rain days R10 and total rainfall PRCPTOT, over 40% have non-stationary characteristics for heavy rain days R25,heavy rainfall R95P, and extreme daily rainfall RX1day, and about 13% have non-stationary characteristics for extremely heavy rainfall R99p;(2) The proportion of stations with significantly increasing distribution parameter sequences of extreme precipitation frequency and intensity index in the GAMLSS optimal model is 20/82 and 36/82, respectively. [Conclusion] The frequency of extreme precipitation in the middle and upper reaches of the Yibin to Yichang section will increase in the future; the intensity of extreme precipitation in the central region of the upper reaches of the Yangtze River will increase in the future; the frequency and intensity of extreme precipitation in the border areas below Shigu in the Jinsha River and the middle and lower reaches of the Mintuo River Basin will increase in the future. Compared to simple trend analysis, time-varying distribution models can diagnose the non-stationary characteristics of sequences in more detail.

Keywords

the upper reaches of the Yangtze River / extreme precipitation index / non-stationary / disaster causing effect / floods / temporal and spatial changes / rainfall / climate change