An AI-Driven Framework for Urban Heat Island (UHI) Analysis and Mitigation Simulation in Bengaluru

The growth rate of Bengaluru has been exacerbated and the resulting UHI effect, were raising temperatures in the city, and, potentially, he was right. more energy use, and community health impacts related to that energy use. While the magnitude of this effect has been measured in. some studies yet lack of high-resolution. tools for simulation that the planner can use in real city envi- ronments. In this paper, we propose a full end-to-end ML- based framework to analyze and forecast the UHI of the city of Bengaluru. Here, we integrated several heterogeneous data including Land Surface Temperature (LST), Normalized Difference Vegetation Index (NDVI) as well as Land Use Land Cover (LULC) derived from satellite images with urban and historical climate data. Every set of data is consolidated into the entire city 100m by 100m grid of the whole city. A Random Forest This data set is exploited to build a regression model to estimate LST with. a large determination coefficient (R 2 =0.85). The major novelty One of the outcomes of this work is a Streamlit-based interactive simulation tool. that enables the policy maker to evaluate the policy from the point of view of the what-if-scenario (e.g., planting more green cover) and get instant visual feedback on. the cooling envelope was so forecast. This tool constitutes the missing one the link between UHI research and urban planning with data.