Prediction of land cover changes in an Urban City of Bangladesh using artificial neural network-based cellular automata

Abstract Savar, a newly developed suburb of Dhaka, is rapidly urbanizing due to various socioeconomic and environmental factors. This study was conducted to evaluate temporal and spatial changes in Land Use and Land Cover (LULC) for the years 1980, 2000, and 2020 and predict future LULC changes. Supervised classification algorithms and cellular automata model based on Artificial Neural Networks (ANN) were used to prepare LULC maps and future simulations. The methodology was designed to overcome limitations in traditional land use and land cover change modeling, including low accuracy, computational inefficiency, and limited adaptability to complex spatial patterns. The study revealed that the rate of built-up area increased significantly over 40 years while barren land and agricultural land decreased drastically. Future LULC simulation results illustrated that the built-up area would increase by 95.07 km 2 (33.29%) in 2040. The model's prediction of the growth of built-up areas by 2040 demonstrated a significant rise in urban coverage with an accuracy rate of 41.14%. Therefore, this study will help us to understand the present and future urban land dynamics along with the trend of temporal and spatial LULC changes that assist planners, policymakers, and stakeholders in sustainable urban planning techniques and urban land management.