Machine Learning Framework for Urban Green Infrastructure Site Suitability in Mumbai

Abstract

This paper offers a new machine learning framework to analyse site suitability of Urban Green Infrastructure (UGI) in the Mumbai area. Mumbai is one of the megacities of India struggling with environmental stress and limited green space. Through incorporation of spatial data which is GIS-based, the proposed framework integrates machine learning methods, compositing layers like NDVI, land use, PM 2.5, slope, and population density in order to identify the best locations for green infrastructure. The stepwise, flexible solution allows making decisions based on the data without dependence on conventional heuristics. Designed for scalability and interpretability, it supports urban planners in prioritising impactful greening interventions, advancing Mumbai’s sustainability and resilience goals while offering a transferable model for other high-density urban regions.