Classifying Flash Flood Disasters From Disaster‐Prone Environments to Support Mitigation Measures

Abstract Spatiotemporal heterogeneities in climatic, physiographic, and socio‐economic environments cause complex and varied formation mechanisms in flash flood disasters. However, previous studies were usually conducted at event or catchment scale in specific environments. Investigation on disaster formation mechanisms in climatic, physiographic, and socio‐economic environments with different combinations and quantities at large scale is not available, which further affects the decision‐making of mitigation measures. Our study develops a type‐based analytical framework of flash flood disasters and their causes from disaster‐prone environments using ten‐fold multivariate analysis including cluster analysis, analysis of similarities, and ordination analysis. Application of this framework to environment factors and losses of 37,332 disaster events across China revealed three disaster‐prone environment types, contributing 55.5% ± 0.3%, 55.9% ± 0.3%, and 50.9% ± 0.2% to variations in disaster attributes, respectively. The events with low disaster intensities (24.6%) in undeveloped northwestern China were governed by short rainfall, low retention capacity, and low prevention investments, and their mitigation focused on afforestation and construction of rainfall and flash flood monitoring systems. Those with high disaster intensities (38.5%) in developed and disturbed central and southeastern China were interpreted by frequent intense rainfall and good flood prevention infrastructures, and their mitigation prioritized development of flash flood forecasting warning models, and grain for green, etc. Those with intermediate disaster intensities (36.9%) in undeveloped southwestern and central China were shaped by frequent short intense rainfall and steep rivers, and their mitigation required satellites or radars in alpine regions, multi‐disaster prevention technology development, and dam construction.