Two-Dimensional Modeling of Flood Events in Denver, Colorado

Advances in radar-rainfall estimation and two-dimensional physically-based runoff modeling offer tools to possibly mitigate the effects of flash flooding, reducing the potential for loss of life and property damage in urban catchments. The ability to model extreme hydrologic events in detail has been demonstrated using the CASC2D model developed by Julien et al. (1995) on numerous rural and urban watersheds (Ogden et. al., 2000). If models of this type are to be used for flash-flood warning in urban environments, it is imperative that input handling and display of hydrologic predictions make use of visualization tools and web technologies to facilitate the dissemination of information and interpretation of results in near real-time. The objective of this study is to determine how much detail in urban topography and drainage characteristics is needed for accurate simulations using data from urban flood events. Radar rainfall estimates, after rigorous quality control, are input to the hydrologic model to simulate flood events in Denver, Colorado. Comparison of simulated and observed discharges demonstrates the role of detailed distributed modeling in improving hydrologic predictions in urban settings. Data preparation and output display are facilitated using advanced server technology so the system can be implemented within a real-time environment.