A Geographically Weighted Regression analysis of the underlying factors of Urban Heat Island phenomenon in Ghaziabad district of Uttar Pradesh

Abstract

Abstract Urbanization is important for societal development. However, it is accompanied by environmental concerns, such as a rise in the temperature of the local area in comparison to that of its surroundings; this phenomenon is known as an Urban Heat Island (UHI). The purpose of this study is to model UHI in the Ghaziabad district using Temperature of the Land's Surface (LST) for the Years 2011 and 2012, estimated from Landsat 7 and 8 respectively, using ArcGIS and SAGA GIS software. The results show that the mean LST increased by 3.11°C over the course of a decade. The Urban Thermal Field Variance Index (UTFVI) was used to map and characterise regions that had high UHI intensities. The areas with high UHI intensity were typically residential areas whereas the low intensity was found associated with cropland. In the Ordinary Least Squares Regression (OLS) regression model, Surface Albedo (SA), Sky View Factor (SVF), Solar Radiation (SR), and Normalized Difference Vegetation Index (NDVI) accounted for 58% of LST variability in 2011 and 66% in 2021. Because LST is nonstationary, a Geographically Weighted Regression (GWR), was utilised to demonstrate that the residuals are normally and randomly distributed.