NEXRAD Flood Warning System and Floodplain Library for Houston, TX

Abstract

Early flood warning systems protect both infrastructure and human life by issuing accurate flood warnings to public and private entities, enabling emergency personnel to employ strategic procedures. One such system, the RiceU/TMC Flood Alert System (FAS) has proved effective at predicting flood levels along Brays Bayou for 30 major storm events since 1997, including T.S. Allison in 2001. The current flood alert version, FAS2, couples available radar data (NEXRAD) with real-time hydrologic models to predict peak flood levels with 2 to 3 hours of lead time, and issues warnings via the internet to emergency personnel. The FAS2 is also currently being expanded to other areas of Texas. The future installation of a distributed high-resolution CASA radar-sensing network (NetRAD) in Houston will increase the city's ability to forecast severe events and to direct emergency response at a street scale. Existing hydraulic models developed by the T.S. Allison Recovery Program (TSARP) for Brays, White Oak and Buffalo Bayous that drain downtown Houston and the Texas Medical Center were revised and merged to analyze the coastal storm surge impacts to the inland drainage areas by incorporating hurricane induced storm surge data in Galveston Bay. The combination of storm surge data, accurate and current hydraulic models, and NEXRAD radar within a GIS framework demonstrated severe inundations in non storm surge zones including major evacuation routes from Galveston. The next generation flood warning system should incorporate storm surge information to better predict the impacts of severe storms on inland watersheds. A unique aspect of this flood warning system is that a predictive floodplain library has been developed for the White Oak watershed in Houston; the library analyzes rainfall intensities and patterns to quantify water surface elevations and delineate floodplains under various spatial and temporal conditions. The development of the floodplain map library for Brays Bayou is currently on-going. FAS2 employs an algorithm that allows the library to dynamically link real-time rainfall observations to appropriate floodplain maps that have been pre-delineated using more than 100 combinations of rainfall variations, allowing the emergency personnel to quickly determine likely inundations and begin flood preparations with as much lead time as possible.