Understanding flood events from the compound perspective

Compound flood events result from multiple physical processes across spatial and temporal scales. Analyzing floods from a single or univariate perspective can underestimate their complexity and dynamics. Conversely, understanding flood mechanisms from a compound perspective facilitates the development of effective adaptation strategies. Here, we made a process-based analysis using a hybrid model, that integrates a hydrological model (namely HBV) with a machine learning model for hydrodynamics, to assess flood events from a compound perspective. This hybrid approach balances model interpretability with computational efficiency. This study aims to quantify the contribution of compound factors to flood events, selected from the Global Flood Database. The findings may serve as a reference for analyzing other types of compound events.