Numerical modeling is a powerful tool for studying the effects of stormwater management in urban catchments, but numerical modeling requires high quality meteorological input data and a well-calibrated hydrological model. The main hypothesis of this research is that consideration of spatial variability in meteorological data and careful selection of calibration method and parameters will result in good performance of an urban hydrological model. The 20.63 km2, 30.3% impervious West Creek watershed in the Cleveland metropolitan area is modeled in the Personal Computer Storm Water Management Model (PCSWMM), based on an existing model provided by the Northeast Ohio Regional Sewer District (NEORSD). 5-minute rainfall data from 4 rain gauges was also obtained from NEORSD. All other meteorological parameters are from the Cleveland airport, after comparison with a station within West Creek watershed. Missing data were filled using linear regression. Identification of sensitive parameters for calibration was done with the help of existing literature and the PCSWMM SRTC tool. Snowpack parameters were manually calibrated based on available data from the Cleveland airport. Uncertainty-based automatic calibration for streamflow using Differential Evolution Markov Chain Algorithm (DREAM) is underway. The percentage of missing data of all meteorological variables varies from 0.21% – 0.61%, a substantial improvement over the single station previously used. Manual calibration of snowpack and addition of an aquifer leads to a Nash-Sutcliffe efficiency of 0.764 for 2017-2018, which will be further improved by using the DREAM algorithm. The final, well-calibrated model will be used to simulate different stormwater management scenarios.