Modelling the Effects of Urban Growth on Land Surface Temperature and Carbon Emissions Through Geospatial and Machine Learning Techniques in Gazipur, Bangladesh

ABSTRACT The rapid expansion of urban areas and their significant effects on carbon emissions and the urban heat landscape have become major research subjects in Bangladesh. The primary aim of this study was to analyse the changes in land use and land cover (LULC) and land surface temperature (LST) from 2003 to 2023 and project them to 2043 for Gazipur, Bangladesh. Additionally, spatiotemporal variations in carbon emissions during summer and winter and their relationships with the LST were explored. A support vector machine (SVM) was used to evaluate changes in LULC. Furthermore, Cellular Automata‐Artificial Neural Network (CA‐ANN) models were utilised to investigate the future dynamics of LST and LULC. The findings of the study include urban expansion from 3% in 2003 to 18% in 2023, and its projected value indicates an expansion of 315 km 2 in 2043 from 52 km 2 in 2003. Therefore, the projected LST results indicate a notable increase of 10°C for the winter season in 2043. The correlation between LST and carbon emissions showed strong R 2 values for both summer and winter. Specifically, the R 2 values were 0.93 and 0.97 for summer and 0.91 and 0.94 for winter in 2013 and 2019, respectively. This investigation has the potential to offer novel insights into prospective urban development, effective management of thermal environments and strategies for mitigating carbon emissions.