Transferability of machine learning-based modeling frameworks across flood events for hindcasting maximum river flood depths in coastal watersheds

Abstract. Despite applications of machine learning (ML) models for predicting floods, their transferability for out-of-sample data has not been explored. This paper developed an ML-based model for hindcasting maximum flood depths during major events in coastal watersheds and evaluated its transferability across other events (out-of-sample). The model considered spatial distribution of influential factors, which explain underlying physical processes, to hindcast maximum river flood depths. Our model evaluation in a HUC6 watershed in Northeastern US showed that the model satisfactorily hindcasted maximum flood depths at 116 stream gauges during a major flood event, Hurricane Ida (R² of 0.92). The pre-trained, validated model was successfully transferred to three other major flood events, Hurricanes Isaias, Sandy, and Irene (R² > 0.71). Our results showed that ML-based models can be transferable for hindcasting maximum river flood depths across events when informed by the spatial distribution of pertinent features and underlying physical processes in coastal watersheds.