Understanding Compound Flood Drivers through a Hybrid Modeling Approach along Southeastern U.S. Coastal River Systems

Compound flooding, which results from the interaction of multiple flood drivers such as high river flows and storm surge events, can have a profound impact on flood risk preparation. Mitigating future flood risk in populated areas along coastal rivers depends on identifying the processes driving such events as well as any potential patterns in flood occurrence. While observational records may provide data to estimate the forcing conditions that drive flooding at a specific location, stream gage stations are often spatially and temporally limited. Applying modeling frameworks to data-driven studies can help specify conditions likely to cause flooding for both minor and major flood events, allowing for a more complete understanding of how joint ocean and river events combine to drive compound flooding that is lacking in exclusively statistical analyses. The goal of this study is to identify how observational and modeled data can be used to characterize river and ocean processes that cause flooding along coastal rivers. First, we identify flood events at United States Geological Survey stream gage stations moving from the coast upriver and specify whether conditions of upstream river discharge and downstream coastal water level combine to drive flooding. Next, we employ a hybrid statistical/numerical modeling framework which stochastically simulates a large sample of upstream and downstream boundary conditions and then uses the Hydraulic Engineering Centers River Analysis System to model along-river water levels in a computationally efficient manner. Together, the observational record analysis and hybrid modeling allow for a comparison of conditions likely to result in flood events to aid our understanding of how observational records can be used for compound flood evaluation. We apply these methods to the Savannah River, separating Georgia and South Carolina, and the Suwannee River in northern Florida. We identify comparisons in the characteristics of statistical and modeling approaches, extreme and non-extreme flood tendencies, and local differences between flood drivers along the Suwannee and the Savannah rivers. This project contributes to future efforts for incorporating compound flooding into graduated coastal hazard risk assessments with the aim of ultimately assisting flood-prone communities.