450 - The Importance of High Resolution Local Datasets for Improved Rain on Grid Two Dimensional Modeling Results in Urbanized Watersheds: A Case Study in Pittsburgh Pennsylvania

The 19 square-mile urbanized Saw Mill Run watershed near Pittsburgh, Pennsylvania is prone to hazardous flash flooding conditions. According to publicly available records, from 2010 to 2020 there were 49 days of individual flash flooding occurrences resulting in widespread flood damages and emergency management services. To characterize the flooding behavior, a full watershed two-dimensional rain on grid flood model in HEC-RAS 6.2 was developed using publicly available high resolution datasets including USGS LiDAR elevation data and local impervious landuse data. The results of the model were validated using an historical observed 100-year rain event that caused widespread flooding in the watershed. To achieve improved model validation results, one square kilometer gage adjusted radar rainfall data was used for the rain event. The validated model was then simulated for a range of return frequency flood events under both present day and future climate scenarios to determine the limits of flood inundation under both existing and future climate conditions. This presentation will provide the audience insights and lessons learned into improving the results of two dimensional rain on grid watershed flood models in urbanized settings. Comparisons in model results between the high resolution local datasets and more commonly used courser datasets such as the USGS National Land Cover Database will be provided.