Development of Operational Flood Prediction System for Coastal Urban Areas with WRF-Hydro: A Case Study in Tybee Island, Georgia

The frequency and severity of flood inundation in coastal watersheds have been increasing due to accelerating sea-level rises and intensifying rainfalls caused by climate change. As a result, nuisance floods have become common for coastal communities already at higher risk of flooding. The emerging threats urgently need an adaptation plan for climate change in multiple sectors, including transportation, stormwater management, and emergency planning. Numerical models have enabled floodplain delineations in a regional scale, mostly focusing on extreme flood events such as hurricanes. However, there is still a lack of numerical models to rapidly inform local nuisance floods in advance. Moreover, no medium exists for coastal communities to tailor such models based on their experiences in the past and present flood events.

To narrow the gap, our study presents a web-based dashboard for operational flood predictions that focuses on a coastal urban area, Tybee Island in Georgia, U.S., as a pilot case study. We use a hyper-local scale numerical modeling framework, which consists of the land surface model, shallow subsurface model, and overland flow model from a distributed hydrologic model (WRF-Hydro) and 1-D hydraulic model for stormwater drainages (SWMM), to provide the real-time and 3-day forecast results on the flood inundation depth, extent, and timing. In addition, the platform includes real-time images and videos such as traffic and beach cams to allow for rapid comparisons with actual flood inundation. Our study also highlights the potential benefits and costs of developing an operational product that can leverage public engagement. Our web-based dashboard is expected to serve as an interfacing channel for coastal communities to integrate end users' knowledge and observations into the model prediction system as well as an informative basis for adaptation planning. This engagement is essential for the numerical flood models to be consistently managed and improved to meet the local stakeholders' demands through calibration and updates.