iFLOOD: Multiscale–Multitemporal Total Water and Nearshore Waves Guidance System for the Chesapeake Bay and the National Capital Region

Flooding is one of the most costly and deadly natural hazard impacting the world today and forecasting flood hazards is important to support efficient emergency response and management. Recent hurricane seasons resulted in losses of lives and major infrastructure damages from combined urban, river and coastal flooding events, demonstrating the importance of integrated total water predictions in coastal and tidal areas. The Chesapeake Bay and the tidal areas around the Potomac River, including Washington DC, are highly vulnerable to multi-flood hazards. The surge- and tide-driven coastal inundation from the Bay, urban runoff from inland precipitation, and high streamflows from the Potomac and Anacostia Rivers result in increasing nuisance flooding and potentially greater flood levels within the National Capital Region. This presentation will describe the iFLOOD flood guidance system (http://iflood.vse.gmu.edu/), which is an experimental and educational platform that resembles aspects of the current operational total water guidance systems in the Chesapeake Bay. The iFLOOD incorporates multi-scale physical processes for total water predictions, including large scale oceanic processes, off-shore and near shore waves, estuarine hydrodynamics, coastal processes and riverine flows. In a constantly evolving framework, it combines a multi-model framework (ADCIRC, SWAN, WaveWatchIII, NAM, X-Beach, WRF-Hydro, NWM and HEC-RAS2D) that provides several flood forecasting parameters at a range of spatial and temporal scales. iFLOOD produces total water and near-shore waves forecasts in a short range mode for 3.5 days initialized 4 times a day, and weekly total water forecasts extending up to four weeks using the SubX ensemble forcing. iFLOOD operational total water guidance has been operational for about 2 years and is analyzed through a partnership with the Baltimore/Washington and Wakefield National Weather Service Forecast Offices, alongside the operational CBOFS, ESTOFS and ETSS forecast systems. Comparison of model results to observation stations demonstrate that the iFLOOD predicted water levels and significant wave heights with a Root Mean Square Error (RMSE) of 0.118 m and 0.55 m during the period of 6 months (Jan’19-Jun’19). A real time bias correcting algorithm resulted in water level predictions with the least RMSE of 0.115 m when compared to other operational forecast systems (CBOFS, ESTOFS and ETSS) for a period of 6 months (Jan’19-Jun’19) considering 15 stations inside and around Chesapeake Bay. A web-based user interface provides operational forecasters and the public a user-friendly access to total water guidance and real-time model performance evaluation.