Urban Heat and Existing Policy in Small and Medium Sized Cities in South Asia: How Land Use Land Cover Change is Affecting the Heat Island Effect?

Abstract

Rapid urbanization impacts the urban environment in many ways. Urban heat Island is one of the most important impacts created by concrete surfaces and anthropogenic activities, leading to a rural-urban temperature difference. Climate change, coupled with these land use changes, makes people vulnerable to heat in cities. South Asia, home to 7 million urban dwellers, faces extreme heatwaves which increase significantly in both intensity and frequency. This study investigates Land Use Land Cover (LULC) change in nine cities across five South Asian countries over 20 years, from 2003 to 2023, and examines the correlation between Land Surface Temperature (LST) and land types—urban, vegetation, and water. The Random Forest Machine Learning model on the Google Earth Engine (GEE) platform is used for LULC classification, with 70-120 training points for each land type for the years 2003 and 2023. The study finds that built-up areas increased in all the cities over 20 years, while vegetation, water, and other land types decreased, except in Ahmedabad, where greenery increased due to planned vegetation. The overall accuracy of the LULC change detection was high (>96%) and the validation accuracy was moderate, supporting the reliability of the LULC change detection results. The study also found that the mean annual LST has increased in all nine cities, with an average increase of 3.2°C. The Pearson R correlation analysis shows that built-up areas have the highest influence on LST in all cities. Vegetation has a medium influence in most cities, and water has little or no influence due to the limited presence of vegetation and water. This study also investigated the contents of HAPs in Ahmedabad and Bhubaneswar and compared them with the international framework suggested by UNDRR. Policy recommendations suggest risk and vulnerability assessments and long-term land use planning to reduce heat vulnerability. Additionally, a GIS-based Heat Vulnerability Index (HVI) is constructed for the Chattogram City Corporation to show the spatial disparity of vulnerable populations - identifying 5 highly vulnerable, 14 moderately vulnerable, and 21 low vulnerable wards.