Copula-Based Risk Analysis of Agricultural Water Shortage under Natural Precipitation Conditions in the Guanzhong Plain, a Drought-Prone Region of China
Lu Zhao,Ningbo Cui,Jing Guan,Pu Du,Yaling Zhang,Shouzheng Jiang
DOI: https://doi.org/10.1061/(ASCE)HE.1943-5584.0002084
2021-03-24
Abstract:Accurate risk forecasting of agricultural water shortages has important implications for the prevention and reduction of disasters in regional agricultural production. Encountering effective precipitation ( Pe ) and crop water requirements ( ETc ) can determine the regional agricultural water shortage risk under natural precipitation supply conditions. The Guanzhong Plain (GP) of China, divided into the Eastern Guanzhong Plain (EGP) and the Western Guanzhong Plain (WGP), is sensitive to water shortage due to the local semihumid and drought-prone climate. Based on the daily meteorological data at six representative weather stations in the GP from 1962 to 2016, Pe and ETc (marked as Pey , Pew , and Pes , and ETcy , ETcw , and ETcs over a hydrological year, winter wheat growth period, and summer maize growth period, respectively) were computed, and copula functions were employed to model the joint distribution of Pe and ETc . The results showed that the mean values of Pey , Pew , and Pes (533, 245, and 287 mm) in the EGP were smaller than those of Pey , Pew , and Pes (599, 274, and 325 mm) in the WGP. However, the mean values of ETcy , ETcw , and ETcs (997, 494, and 502 mm) in the EGP were greater than those of ETcy , ETcw , and ETcs (895, 417, and 477 mm) in the WGP. The Frank copula was identified as the most suitable model for the joint modeling of Pe and ETc series. According to the joint probability distribution, the asynchronous encounter probability of Pe and ETc was around three or fourfold of the synchronous encounter probability in the GP. The occurrence probabilities of severe and extreme agricultural water shortage under natural precipitation supply condition were 0.24–0.27 and 0.04–0.06 in the EGP, while the corresponding values were 0.20–0.23 and 0.03–0.04 in the WGP. This study proved that natural precipitation could not meet crop water demand, and the joint encounter probability of low effective precipitation and high crop water demand was high, so irrigation and optimum agricultural water resource allocation are thus necessary for agricultural production in the semihumid and drought-prone region.