ABSTRACT
Guwahati city, considered to be the most important city in north-east India, underwent rapid urbanization due to its geographical connectivity and geo-strategic importance. Therefore, executing a comprehensive assessment and understanding the historical land-use/land-cover (LULC) changes, along with forecasting future possibilities, are imperative for ensuring sustainable development. To track the historical dynamics of LULC in Guwahati city, the Random Forest (RF) algorithm was adopted to classify the Landsat images of 2000, 2014, and 2022. Those images were further validated using kappa statistics and achieved an overall accuracy of 0.92, 0.94, 0.97, respectively. After inclusion of natural and anthropogenic spatial drivers’ urban growth, the LULC map for the years 2030 and 2050 was projected by incorporating multi-layer perceptron-artificial neural network and Markov chain approaches. The results reveal that in comparison to 2022, the simulated model of 2050 anticipated that urban areas will expand by 51.88%, but at the same time, there will be a considerable decline in the amount of vegetation cover (23.94%), forest land (10.17%), and agricultural areas (24.15%). The outcome of the study may be vital in ascertaining valuable information in forecasting the foreseeable trend of changes in LULC, as well as for developing alternative land-use strategies and giving guidance to urban planning officials.
