Urban flooding is caused due to poor drainage design, extreme weather, and excessive rain. Such flooding severely affects the road infrastructure. While there are a number of hydrologic software (e.g., TR-55, HydroCAD, TR-20, HEC-RAS, StreamStats, L-THIA, SWMM, WMOST, MAST, HY-8) available to examine extent of urban flooding, the softwares primarily require walking through a series of manual steps and address each study area individually preventing a collective view of an urban area in an efficient manner for hydrologic analysis. Furthermore, the softwares have no ability to recommend optimal culver pipe sizes to minimize flooding. In this paper, we develop a non-linear optimization formulation to minimize urban flooding using underdrain pipe size as a decision variable. We propose a solution algorithm in an integrated GIS and Python environment. Monte Carlo Simulation is used to simulate rainfall intensity by using empirical data on extreme weather from the National Oceanic and Atmospheric Administration. An example using the storm-drain system for the Baltimore County is performed. The results show that the model is effective in identifying storm-drain deficiencies and correcting them by choosing appropriate storm-drain inlet types to minimize flooding. The proposed method eliminates the need to examine each study area manually using existing hydrologic tools. Future works may include expanding the methodology for large datasets. They may also include a more sophisticated modeling approach for estimating rainfall intensity based on extreme weather patterns.