The diagnosis about spatio-temporal characteristics and driving factors of flash drought and its prediction over typical humid and semi-arid basins in China

Abstract

Flash drought is a rapid-onset drought, which has greatly threatened the agricultural production and economic development. However, the unclear development mechanism of flash droughts brings challenges for its monitoring, forecasting and mitigation. This study investigates the spatio-temporal characteristics, driving factors and the prediction of flash drought over typical humid and semi-arid basins. The main objectives and findings are as follows: (1) The patterns of flash drought are compared under different climate types. The results show that flash drought is more serious in the chosen humid basin than that in the semiarid basin, with more events, longer duration, larger frequency of occurrence (FOC). (2) The development mechanisms of flash drought are explored by analyzing the anomalies of seven meteorological variables in the evolution of flash drought. The results indicate that the main driving factors are the negative anomalies of precipitation and the positive anomalies of temperature, which usually occur at two pentads before the onset of flash drought. (3) The prediction of soil moisture as a key variable in flash droughts developing process is conducted using support vector machine (SVM), with meteorological variables and a remote sensing soil moisture, namely Soil Moisture Active and Passive (SMAP), as inputs. In this study, about 65%-70% of flash droughts can be captured by the prediction. However, some events are missed, and false alarms also exist in most stations in both basins. This study can provide some references for monitoring and early warning of flash drought, which is important to reduce the losses and risks in agriculture production.