Advancements in large floodplain inundation mapping: Integrated computational modeling and remote sensing techniques

• HEC-RAS 2D and REM provided consistent estimates of floodplain inundation. • Agreement between models improved at higher flows, with differences <6%. • REM showed slightly lower error metrics than HEC-RAS. • Findings highlight the sensitivity of floodplain connectivity to the flow regime of the Lower Apalachicola River. The devastating impacts of flooding can be mitigated by strengthening the integration between hydrodynamic modeling and remote sensing, thereby improving floodplain management and the accuracy of inundation mapping. However, the relationship between river discharge and floodplain response remains insufficiently quantified in many large river systems, limiting reliable estimation of inundation extent, flood volumes, and floodplain connectivity. This study evaluates the hydraulic behavior of the Lower Apalachicola River and its associated floodplain within the Wewahitchka-Sumatra reaches using a two-dimensional HEC-RAS modeling framework. A high-resolution LiDAR-Sonar integrated digital elevation model (DEM) and terrain surfaces generated using the Relative Elevation Model (REM) approach were implemented to simulate floodplain inundation depth and extent under representative low-flow (175.3 m³/s) and high-flow (4360.8 m³/s) conditions observed during the 2015-2016 flood event. The REM centerline method was applied to smooth channel elevations, resulting in systematically deeper channel representation relative to the LiDAR-Sonar DEM. Model performance was evaluated using inundation extents derived from Landsat-8, MODIS, and Sentinel-1 satellite imagery. Results indicate that differences in inundation extent between DEM-based and REM-based simulations increase with flow magnitude, with areal differences of 9.93% and 4.02% under low- and high-flow conditions, respectively. DEM-based HEC-RAS simulations consistently produced larger flood extents than REM-based simulations, primarily due to differences in the representation of main-channel elevation. Maximum flood-water depth differences of 0.78 m and 0.76 m were observed for DEM and REM simulations under low- and high-flow conditions, respectively. Overall, the findings confirm the robustness of the HEC-RAS 2D framework for simulating floodplain inundation in large, low-gradient river systems and demonstrate the value of integrating high-resolution topographic data with complementary modeling approaches. The proposed framework offers a transferable methodology for floodplain management, restoration assessment, and improved understanding of river-floodplain connectivity under varying hydrologic conditions. Highlights Lidar-Sonar combined DEM and REM provided consistent estimates of floodplain inundation. Agreement between models improved at higher flows, with differences <6%. REM showed slightly lower error metrics than DEM. Findings highlight the sensitivity of floodplain connectivity to the flow regime of the Lower Apalachicola River.