Renmin Zhujiang (Jan 2024)
Impact of Urbanization Process on Extreme Precipitation Events Based on Non-Stationary GEV Model
Abstract
In recent years, the increase in extreme precipitation events has significantly impeded urban development. To explore the relationship between urbanization and extreme precipitation events, Shenzhen City was investigated in this study. Urbanization indicators were integrated into the generalized extreme value (GEV) distribution model by using four parameter extension methods. After two screenings, the optimal model was obtained, and the non-stationary precipitation intensity-duration-frequency (IDF) curve was obtained. The non-stationarity was quantitatively analyzed. The results show that: ① Non-stationary GEV models incorporating urbanization indicators are superior to the stationary models. The applicability of the stationary GEV models is reduced. ② The stationary IDF curves underestimate the precipitation intensity of all extreme precipitation events by approximately 6.3% and overestimate the non-exceedance probability. ③ Urbanization significantly impacts short-duration precipitation events, with a relatively smaller effect on long-duration precipitation events. As the return period increases, the influence of urbanization on precipitation intensifies; for each additional year of the return period, its impact on precipitation intensity increases by approximately 0.13 mm/h. Exploring non-stationarity under the influence of urbanization can provide a scientific basis for urban responses to extreme precipitation events. This is of great significance for urban flood control.
