Dynamic Flood Risk Assessment for Transportation Infrastructure in Fresno, CA

Urban roadway flooding in flat metropolitan environments often shows up as small areas of standing water rather than large, flowing floods. In low-gradient urban environments, subtle micro-topographic features—such as pavement crowns, curbs, and localized depressions—often control where surface water accumulates following rainfall. Accurately identifying these localized ponding areas is important for roadway flood screening and infrastructure management. This study develops a LiDAR-based depression mapping framework to identify roadway segments susceptible to surface ponding in flat urban environments. High-resolution airborne LiDAR datasets were used to generate detailed elevation maps (down to less than a foot in resolution) from ground-classified point clouds (collection of data points). Surface depressions were identified using a hydrologically filled DEM approach, where depression depth was computed as the difference between the filled and original terrain surfaces. Depressions exceeding a depth threshold were segmented into polygon features, and ponding storage volumes were estimated using their depth and surface area. To focus the analysis on transportation infrastructure, ponding features intersecting buffered roadway corridors were used. Flood risk potential was then evaluated by integrating ponding severity with proximity to stormwater drainage inlets to generate a composite roadway flood risk score. The methodology was applied to two independent LiDAR datasets covering the same urban study area. Results show that dataset characteristics significantly influence the number and hydraulic magnitude of detected roadway depressions. One dataset identified many smaller, shallow areas where water could collect, while the other identified fewer but deeper and larger pools. Despite these differences, both datasets revealed consistent spatial patterns of roadway ponding potential. The results demonstrate that LiDAR-derived depression analysis provides an efficient screening tool for identifying flood-prone roadway segments in flat urban terrain. By integrating terrain morphology with drainage infrastructure proximity, the proposed framework