Comparison of 1D, 2D, and 1D/2D Approaches for Flood Inundation Modeling in an Urban Watershed

The widespread devastation caused by Hurricane Harvey in August 2017 to Houston, TX has accentuated the region's flood vulnerability. In fact, several watersheds within the region has incurred substantial flood damages from multiple storms in recent years. Alarmingly, many of the damaged properties in these watersheds are located outside of the regulatory floodplains, which indicates that the current 1-D steady approach of mapping floodplains is often inadequate to represent actual flood risks for watersheds in this region. This condition suggests that new flood inundation modeling / mapping approaches might be necessary in order to effectively plan for future development practices, flood regulations, stormwater management strategies, and flood mitigation efforts.

This study compares three flood inundation modeling approaches using the US Army Corps of Engineers' hydraulic model, HEC-RAS (Hydrologic Engineering Center - River Analysis System): 1D-unsteady, 2D-unsteady, and a coupled 1D/2D unsteady model. The study is focused on the Meyerland area, a residential community within Brays Bayou watershed in Houston, TX, which had suffered massive flood damages from Harvey and other severe storms in the last three years (i.e., Memorial Day Storm 2015 and Tax Day Flood 2016). The performances of the three approaches are evaluated by comparing modeled inundation to available observed data (e.g., stream gage records and known high water marks) for Hurricane Harvey. Additionally, inundation extent and potential flood damage (i.e., number of parcels within floodplains) are also simulated for the 100-yr and 500-yr storms, and compared against FEMA's 100-yr and 500-yr floodplains at corresponding areas of interest.