Through the Indian subcontinent, there is a major risk of flash floods during the monsoon season. The Indian government tasked the Indian Institute of Science, Banglore and Indian Urban Data Exchange to develop a live prediction algorithm to tackle this problem. I worked on this project as a summer intern and tackled the project using hydrodynamic and hydrological equations. I developed a Python wrapper for the EPA's Storm Water Management Model which can be dynamically used.  

My primary objective was to use Google Earth engine data which includes area and line data that demarcates water bodies in any given location. I planned to use hydrodynamic and hydrological analysis to use the data to find run-off for the given rainfall using the EPA's Stormwater Water Management Model. However, the SWMM software cannot be used dynamically, to develop an application that can serve the Indian government in real-time. This is why I developed a Python wrapper around the SWMM solver which can take in the area and line data from the Google Earth engine and calculate run-offs in real-time. The blue lines in the displayed graph show the run-off from the city of Banglore.