Suitability of Satellite Data for Urbanization Study: A Comparative Analysis

The availability of free satellite data and open-source Geographic Information System platforms have significantly contributed to the progress in Land use Land cover change studies. The urban sprawl of the cities is a great challenge for policymakers and urban planners. The current study conducted in Dharamshala city of Himachal Pradesh, India, illustrates the variability in classifying areas, specifically important from an urbanization point of view, using multiple satellite sources. The three satellite images Sentinel 2, Landsat 8 OLI (Operational Land Imager), and Landsat 7 ETM⁺ (Enhanced Thematic Mapper) were used in the study, and an integrated approach comprising unsupervised classification along with use of spectral vegetation indices, Enhanced Vegetation Index (EVI), Normalised Difference Built-up Index (NDBI) and Modified Normalised Difference Water Index (MNDWI) were used to produce four land cover types, viz. Protected Areas, Agricultural Areas, Built-up areas, and water bodies. The overall accuracy for Sentinel 2, Landsat 8 OLI, and Landsat 7 ETM + was 83.40%, 80.06%, and 74.32% respectively, while Kappa hat was 0.72, 0.65, and 0.53 respectively. Based on meticulous scrutiny of accuracy metrics employed in the investigation, it is inferred that the utilization of the Sentinel-2 satellite dataset has demonstrated superior performance, thus meriting recommendations for urbanization studies within the Western Himalayan region. This assertion is substantiated by the rational deduction that the comprehensive spectral and spatial capabilities inherent to the Sentinel-2 dataset provide unparalleled insight into the intricate dynamics of urbanization, thereby facilitating more precise and insightful analyses in this geographically complex and environmentally sensitive area.