Impact of uncontrolled industrialization on environmental parameters in the meghna industrial economic zone using machine learning

Uncontrolled and rapid industrialization poses significant threats to environmental sustainability in developing countries, particularly within newly established economic zones. Addressing the limited predictive assessments of rapidly expanding industrial zones in Bangladesh, this study integrates historical environmental change analysis with machine learning–based projections to establish linkages between future land-use transitions and their associated ecological and thermal stresses. The environmental impacts of industrial expansion in the Meghna Industrial Economic Zone (MIEZ), Bangladesh, were assessed through changes in land use and land cover (LULC), normalized difference vegetation index (NDVI), normalized difference water index (NDWI), land surface temperature (LST), and urban heat island (UHI) intensity, together with predictions of future conditions under continued growth. Landsat 8–9 images (2018–2025) were used to map LULC, NDVI, NDWI, LST, and UHI, while a Cellular Automata-Artificial Neural Network (CA-ANN) model was used to predict future conditions (2029 and 2033). Results show that built-up land increased nearly fourfold from 2018 to 2025, while vegetation and surface water declined. At the same time, warmer areas expanded, with LST above 22 °C covering more than 33% of the zone by 2025. Future projections suggest continued growth of built-up land, further loss of vegetation and moisture, and stronger UHI effects by 2033. These findings highlight rising environmental degradation in MIEZ and the need for planned industrial development, stronger environmental control, and green infrastructure.