Drought monitoring and agricultural drought loss risk assessment based on multisource information fusion

Determining the loss mechanism of drought is crucial for the prevention of and adaptation to agricultural drought. This paper proposes a theoretical framework of agricultural drought loss analysis including drought monitoring, factor identification, and risk assessment based on multisource information fusion. First, drought events are monitored by combining the drought severity index and run theory, which help to extract the characteristic variables of drought frequency, duration, and intensity. Second, the agricultural drought loss rate is calculated by the “trend-fluctuation” decomposition model. Finally, the multisource information of remote sensing, meteorology, hydrology, and socioeconomic data are fused, and the improved grey incidence model with the B-mode based on panel data is proposed to identify key factors and evaluate the risk of agricultural drought loss. The results exemplified by China’s Henan Province show that the frequency of drought in 2001–2018 is inversely related to the average duration and intensity of drought to some extent. The agricultural drought losses in the central and northern areas are higher, while those in the southern areas are relatively low. The water production coefficient as well as surface water resources, precipitation, and groundwater resources are the main factors that affect agricultural drought losses. This study provides theoretical support for analyzing drought formation mechanisms and preventing and controlling drought risk.