REAL TIME FLOOD FORECASTING WITH PCSWMM, NEXRAD AND SWMM5

摘要

Reliable flood forecasting systems have been emphasized in the last few decades. Responsibility for flood management and flood warning in Ontario has been assigned to the Ontario conservation authorities and agencies within their local areas. In an effort to improve the accuracy and lead time for flood forecasting, a remote-sensing real-time flood forecasting and web-based decision support system has been developed. The real-time system consists of four main components: 1. real-time radar rainfall acquisition, processing and forecasting, 2. continuous SWMM modeling, 3. flood inundation and flood vulnerable asset analysis, and 4. real-time web-based decision support. Basic utilities include (a) PCSWMM Real-Time, (b) remote-sensing radar, (c) predictive weather modeling, (d) real-time rain gauging, (e) US EPA SWMM5, (f) HTML5 and (g) Google Maps. Near-future water surface elevations are forecast throughout the watersheds, and auto-reported in real-time to flood duty staff, who make final, informed decisions. Next Generation Radar (NEXRAD) level Ⅲ data is examined, and this paper describes how radar is incorporated in hydraulic models to estimate water surface elevations. Three Level Ⅲ products of NEXRAD radar, and three different processing bias methods for incorporating rain gage networks were evaluated, and the best-performing combination recommended.