Abstract: Flash droughts are extreme events characterized by a short duration, strong intensity, and rapid onset. Owing to their unusually rapid occurrence, the use of current drought monitoring tools to capture them is still challenging. In this study, using ERA-Interim soil moisture reanalysis products from the European Center for Medium-Range Weather Forecast, an approach that considers the rate of decrease in soil moisture content during the onset stage of flash droughts was proposed for the identification of the flash droughts that occurred during the 1979-2018 period. Further, the anomalies associated with meteorological variables during the drought onset stage of these events were analyzed, and the feasibility of using precedent meteorological driving forces to realize flash drought simulation based on a multivariate linear regression model was investigated. The results obtained showed that the rate of soil moisture intensification during the onset state of flash droughts showed significant spatial discrepancy. Specifically, during this stage, soil moisture content declined much faster in the south of the Yangtze River than in Northwestern China. Additionally, the mean, peak, and variation ranges of the anomalies of the different meteorological variables exhibited more significant spatial discrepancies during this stage. Particularly, the peak values showed the most significant variation, and were half the standard deviation higher than the slowly-evolving drought. Furthermore, by integrating the anomalies associated with multiple meteorological variables during the early and onset stages of the flash drought events, the regression model provided satisfying simulations of the soil moisture intensification rate, suggesting that it is possible to simulate flash droughts from a meteorological perspective.