Multi‐Index Interaction of Remote Sensing and Machine Learning Framework for Urban Heat Island Risk Mapping

ABSTRACT Urban Heat Islands (UHI) represent one of the most significant environmental challenges facing rapidly growing cities worldwide. This study presents an innovative integrated framework combining remote sensing, machine learning, and multi‐index interaction modeling to assess UHI risk patterns. Multi‐temporal Landsat satellite data from 2018–2024 were processed to derive seven key indices: Normalized Difference Vegetation Index (NDVI), Soil Adjusted Vegetation Index (SAVI), Modified Normalized Difference Water Index (MNDWI), Normalized Difference Water Index (NDWI), Normalized Difference Built‐up Index (NDBI), Albedo, and Land Surface Temperature (LST). A supervised Support Vector Machine (SVM) classification was employed for Land Use Land Cover (LULC) mapping, achieving overall accuracies exceeding 85%. The novel contribution of this research lies in the development of a Risk Interaction Model that synthesizes multiple indices into four integrated risk classes: Vegetated & Cool, Bare & Hot, Water‐Buffered, and Urban Heat Hotspot zones. Results reveal distinct UHI patterns in the study area, with mean LST variations ranging from 18.72°C to 33.14°C over the study period. The framework successfully identified critical UHI hotspots concentrated in built‐up areas and provided actionable insights for urban climate adaptation strategies.