Inland flood impacts in urbanized coastal areas affected by tropical storms can be severe as witnessed by Tropical Storm Allison and many other historic events. Site specific hydrologic prediction is made possible by remotely sensed precipitation by radar and telemetered gauges (multisensor), and high-resolution geospatial data within a distributed model context. Radar systems provide high-resolution precipitation measurements that are useful for urban hydrologic applications. An engineering testbed for advanced quantitative precipitation estimates using radar is under development for urban watersheds in the Houston area as shown in Figure 1. Radar measurement characteristics have important consequences on the predictability of urban watersheds. Operational system simulation experiments provide guidance on design of quantitative precipitation estimates for distributed hydrologic modeling. The results of such experiments help assess prediction accuracy and predictability limits in support of continuous real-time flood alert system operation.

Figure 1 Planned urban hydrologic engineering testbed for testing quantitative precipitation estimates from radar networks. Operational system design depends on the ability of radar to provide accurate measurements of rainfall over small drainage areas that are characteristic of urban drainage systems. The subject of this paper is the examination of short-term hydrologic prediction accuracy for critical urban drainage facilities using radar rainfall and physics-based distributed hydrological modeling. Using radar QPE derived from the existing WSR-88D (KHGX) as input to a physics-based hydrologic model of Brays Bayou (260 km2) provides a baseline for comparison and guides design of future radar networks. This paper will present a validation experiment for a small convective event using a model calibrated during the named tropical storms Allison and Frances, which created significant inland flood losses. Short-term hydrologic prediction accuracy for these storms are being evaluated for a real-time continuous radar-based flood alert system that serves the Texas Medical Center located in Houston Texas.