Calibration and Verification of a Hydrometeorological System for Urban Flash Flood Forecasting

Integrated hydro-meteorological systems that can accurately forecast flash flood events at urban scales have become requisite tools to support emergency management and decision making for business operations. Components of such complex systems are typically developed, tested and validated as stand-alone applications, with varying spatial and temporal resolution and parameterization schemes configured for different purposes. When these components are linked together in a customized application, it is necessary to calibrate the entire integrated system to ensure adequate interpretation of the flash flood forecast for business operations. We present a numerical urban flood forecasting system based on coupling between a mesoscale numerical weather prediction model with data assimilation capabilities (WRF-ARW) and a novel hydrological model employing unstructured two-dimensional grids in a conservative approach to simulate the surface flow. Within an ensemble framework the forecast uncertainty of the flood extent and timing is estimated. Our integrated hydro-meteorological system is a part of a decision support system for flash flood forecasting and management in urban areas, which uses high-resolution infrastructure data available for major metropolitan areas. A combination of data sources from local surface weather stations, hydrographs, radar, and satellite imagery is used for calibration and validation. In this presentation, we will discuss the results based on retrospective analysis and assessment of flash flood events, while the integrated system is used operationally to provide forecasts and flood risk analysis to local area urban businesses.