Spatiotemporal response of agricultural drought to meteorological drought in the upper Hanjiang River Basin from three-dimensional perspective

Abstract

Understanding the relationships among different types of droughts is critical for effective drought early warning systems and water resource management. While much attention has recently been given to how agricultural drought (AD) is influenced by meteorological drought (MD), the spatial continuity of this relationship has often been overlooked. In this study, we introduce a set of criteria for identifying drought that takes into account its spatiotemporal continuity. We also establish a framework for quantifying the uncertainty associated with drought response using Copula functions in conjunction with a Bayesian network probabilistic model. We apply this framework to comprehensively assess the likelihood of MD leading to AD under various drought conditions (e.g., duration, area, severity) in the upper Hanjiang River Basin (UHJRB) from 1963 to 2014. Our findings indicate that by incorporating spatiotemporal continuity criteria, drought events can be more effectively identified. Specifically, 78 % of AD events followed MD events, with an average response time of 2.4 months. Generally, the probability of AD occurrence increases as the corresponding MD characteristics (such as duration, area, and severity) increase. For instance, when MD duration exceeds 6 months, the affected area surpasses 60 % of the basin, or the severity reaches 5.0 × 10⁵ km²·months, the probability of AD occurrence exceeds 80 %. The outcomes of this research can serve as a valuable reference for drought response efforts in the UHJRB. Furthermore, the proposed research framework can be adapted for use in other regions to enhance our understanding of MD and its impacts.