Drought and flood risk assessment for rainfed agriculture based on Copula-Bayesian conditional probabilities

Abstract

Rainfed agriculture ensures global food security. However, periodic droughts and floods endanger rainfed crop productivity. This work aims to identify drought and flood risks in rainfed agriculture using conditional probabilities and to reduce food losses caused by drought and floods. The standardized precipitation evapotranspiration index is produced from 1996 to 2020 using meteorological data from the three northeastern provinces and is used to establish four time-scale drought indicators. The Pearson correlation coefficient is then used to identify drought indicators related to rainfed crop production. The Copula-Bayesian conditional probabilities were used to calculate the probability of yield reduction of rainfed crops when drought indicators indicate varied drought and flood conditions. The results show that: (1) Droughts from May to July and floods from January to March had the most significant impact on rainfed crops in the three northeastern provinces. (2) The cities of Huludao, Changchun, Baishan, Baicheng, Yanbian, Hegang, and Jiamusi face a more significant danger of drought in rainfed crops. In contrast, Yichun and Suihua face a higher risk of flooding in rainfed crops. (3) When drought monitoring indicators decreased by one unit, rainfed crop output declined the most in Fuxin city; when the drought indicator increased by one unit, rainfed crop yield decreased the most in Liaoyuan city. The study’s findings can provide quantitative direction for food security work in China’s three northeastern regions.