The Effect of Urbanization-Induced Rainfall Variability on Hydrologic Response: Linking Mesoscale Meteorological Model Output to an Urban Watershed Model

Abstract

This paper reports the preliminary results of a study of the relative effect of urban-induced precipitation variability on the hydrological response of a large urban watershed. The study involves (1) developing a mesoscale meteorological model centered on the Houston metropolitan area to simulate urban effects on meteorological fields, particularly precipitation, (2) developing a stormwater runoff model for a large urban watershed in the Houston metropolitan area, and (3) using the simulated rainfall fields as input to the watershed model. Meteorological simulations were performed to elicit the urban effect on a case study precipitation event and the simulated rainfall fields are being used as input to the stormwater runoff model to study the differences in runoff response caused by the urban rainfall modification. This paper describes the development and verification of the models and outlines the concept of the use of the simulated rainfall fields as input to the runoff model. Results of the linkage and analysis of urban rainfall modification effect on runoff response will be presented at the conference. Background There have been several recent efforts to link mesoscale meteorological models with hydrologic models. The objectives have typically been to provide improved representation of the spatial variability of precipitation for input to hydrologic models to improve the prediction of flood magnitudes and extent and to increase the forecasting lead-time. Thielen and Creutin (1997) conducted a runoff analysis using a coupled meteorological-hydrologic model where the simulated rainfall fields served as input to the hydrologic model. In their study, TRAPPES Radiosonde data were used as input to the Clark Module meteorological model. The Clark Module uses two way interactive grid nesting, which provides the ability to use multiple resolution grid domains. Horizontal grids consisted of 9, 3, 1 and 0.5 km resolutions. The vertical resolutions of the first three grids were 250 meters and 125 meters for the finest horizontal resolution. The urban watershed area studied is located near Paris and has an approximate area of 116 km 2 . Land surface and hydraulic systems for the study area were analyzed using the CAREDAS (Calcul des Reseaux d’Assainissement) model. The CAREDAS modeling application uses two modules to estimate urban stormwater runoff. The PLUTON module uses a lumped approach to calculate rainfall-runoff transformations for subbasins. PLUTON