Estimation of water required to recover from agricultural drought: Perspective from regression and probabilistic analysis methods

Agriculture is one of the sectors most directly and severely affected by droughts. Therefore, information regarding the water required (WR) to recover from agricultural drought and the probability of agricultural drought recovery under different precipitation scenarios is of great importance for drought risk management. In this study, we developed a quantitative evaluation method to estimate the WR of agricultural drought event recovery and a probabilistic drought recovery framework based on the copula function to calculate the proba-bilities of recovery from different grades of agricultural drought under given precipitation scenarios. The developed method and framework were demonstrated in the Yangtze River Basin (YRB) using the improved soil moisture anomaly percentage index (ISMAPI). The average duration of agricultural drought recovery in the YRB was 28 d. The difference in soil moisture persistence was the main factor affecting the spatial pattern of the grid mean drought recovery characteristics. For individual drought events, precipitation was the most important factor affecting drought recovery, followed by drought development intensity. There is a clear nonlinear rela-tionship between the PCP value (percentage of cumulative precipitation during drought recovery to its corre-sponding climatological mean) and the drought recovery duration. In most grids, the three-parameter power function can be used to characterize this relationship and further estimate the WR to recover from an agricultural drought event under different expected recovery durations. We estimated the probabilities of extreme and severe agricultural drought recovery under different precipitation scenarios when the expected recovery durations were 10, 20, 30, 40, 50, and 60 d based on the developed probabilistic drought recovery framework. For example, when the cumulative precipitation is no less than its climatological mean within 30 d, the probability of recovering from extreme agricultural drought is 0.59. These findings are important for decision-making in water and drought management.

Automated summary

Agriculture, as one of the sectors most affected by droughts, requires information about recovery water and probabilities under different precipitation scenarios for drought risk management. This study developed a method and framework to estimate agricultural drought event recovery and probabilities based on copula function. The method and framework were applied in the Yangtze River Basin using improved soil moisture anomaly index. Findings include an average recovery duration of 28 days, with soil moisture persistence and precipitation as main factors affecting recovery. The study also estimated recovery probabilities under different scenarios for decision-making in water and drought management.