Monitoring and prediction of land use land cover change of Chittagong Metropolitan City by CA-ANN model

Anthropogenic activities have significantly changed global land use land cover (LULC), especially in areas with extreme population expansion and climate change. Remote sensing and the geographic information system are popular methods for keeping track of LULC changes. This research analyzed historical changes and predicted future patterns of LULC in Chittagong Metropolitan City, Bangladesh. LULC maps were created using the semi-automated classification plug-in in QGIS, which was used to classify the satellite images from 1990 to 2020. Eight distinct LULC groups, including (i) agricultural land, (ii) fallow land, (iii) trees outside of forest, (iv) hill vegetation, (v) mangrove, (vi) built-up, (vii) pond/lake, and (viii) river, were used to classify the images. The generated LULC maps showed the changes in the area from 1990 to 2020 in several classifications, showing increases in built-up, pond/lake, and river waterbodies of 278.5, 407.69, and 10.07%, respectively. However, a significant decline was observed in agricultural (61.64%), hill vegetation (43.26%), mangrove forest (30.7%), fallow (22.02%), and Tree Outside of Forest (TOF) land (8.81%). The LULC changes between 2020 and 2035 were predicted using the cellular automata-artificial neural network (CA-ANN) model. The prediction maps from 2020-2035 illustrated increasing trends of built-up land (+ 15.65%) and decreasing trends of all other LULC types, predominantly, agricultural (-3.53%), fallow (-3.13%), trees outside of forest (-6.53%), and pond/lake (-1.44%). The findings of this research may be helpful to design future strategies for sustainable landscape management and help decision-makers in government make better choices for the environment and the ecosystem.