Urban Growth Estimation of Moradabad City Using Sentinel-2 Images

Accurate monitoring and assessment are necessary due to the enormous challenges posed by rapid urbanization to sustainable development and resource management. This study utilizes Sentinel-2 satellite images of Moradabad City, located in Uttar Pradesh, India, to analyze urban growth through a machine learning approach. Sentinel-2 imagery is widely used for land cover monitoring, and images from 2017 and 2023 were selected to assess the rapid expansion of urbanization. Various techniques, including SVM, ANN and minimum distance (MD), were employed for land surface classification. Among these, SVM produced the most reliable results for estimating urban areas, based on both the Kappa coefficient (Kc) and overall accuracy (O.A.). The SVM estimates that in 2017 about 8.308% of the region was classified as an urban area, and from 2023 that number had increased to about 11.64%, which could be a serious concern for Moradabad city's residents. The motivation behind this study to explore the potential applications of machine learning methods for small-scale urban dynamics monitoring and analysis through satellite image analysis. The research advances knowledge of urbanization processes and offers insightful advice for sustainable urban growth in quickly expanding places.