Heuristic Approach to Urban Sewershed Delineation for Urban Flood Modeling
Description
The escalating risk of urban pluvial flooding, exacerbated by increasing heavy rainfall and urbanization globally, calls for the development of reliable and streamlined urban flood modeling techniques. The necessity for optimal sewershed delineation in semi-distributed hydrological models has become crucial in urban hydrology. Conventional delineation methods, mainly based on natural topography, fall short in urban environment because they do not consider the critical role of human-made drainage systems on stormwater flow. Despite local governments ordinarily pre-defining sewersheds for their municipalities, limited access to this geographical information or its low spatial resolution has become an obstacle to urban flood research. To address this, the study introduces a heuristic approach to urban sewershed delineation by identifying and compartmentalizing the target sewershed into smaller sub-sewersheds, solely utilizing publicly accessible information such as trunk sewer layouts, road networks, and land use/land cover. The stormwater model using heuristic sewershed delineation achieved satisfactory or better performance (0.60<NSE<0.91, 0.31<RSR<0.63, -8.8%<PBIAS<13.9%) for calibration and validation against observed sewer flow depth at multiple locations in Seoul's Gangnam district, South Korea. Findings demonstrate that a heuristic approach to urban sewershed delineation can achieve as high model accuracy as pre-defined sewershed models by the government. By doing so, heuristic sewershed delineation minimizes dependency on confidential urban drainage network data, while still achieving reliable results, to strike the balance between model performance and parsimony. The proposed heuristic approach can leverage local expertise collectively and contribute to best stormwater management practice and urban flood risk reduction.