Abstract
<title>Abstract</title> <p>Rapid urbanization is one of the primary drivers of land use/land cover (LULC) change in metropolitan regions, often resulting in the loss of agricultural lands and increasing environmental pressures. This study investigates the spatiotemporal dynamics of LULC change and future urban expansion in the Tabriz Metropolitan Area (TMA), Iran, between 1972 and 2021 using multi-temporal Landsat imagery and a data-driven modeling framework. Landsat images from 1972, 1995, 2016, and 2021 were processed within the Google Earth Engine (GEE) platform and classified using the Support Vector Machine (SVM) algorithm. A Cellular Automata–Markov (CA-Markov) model was subsequently employed to simulate future LULC patterns and project urban growth to 2030. Classification results achieved overall accuracies ranging from 80.65% to 81.76%, with Kappa coefficients between 0.757 and 0.770. Validation of the CA-Markov model yielded a Kappa coefficient of 0.796, indicating strong predictive performance. The results revealed substantial urban expansion, with settlement areas increasing from 25.01 km² in 1972 to 293.19 km² in 2021. Simultaneously, significant transformations occurred among cultivated lands, rangelands, orchards, and bare lands, reflecting the combined influence of population growth, urban development, and metropolitan boundary expansion. Transition analysis further indicated increasing conversion of cultivated lands to settlement areas during the most recent period. The 2030 projection suggests continued urban expansion, particularly along the eastern and western development corridors, accompanied by further pressure on agricultural lands. These findings highlight the need for sustainable land-use policies and demonstrate the value of integrating remote sensing, machine learning, and predictive modeling for metropolitan planning and management.</p>