Drivers of Urban Growth: Cellular Automata–Markov–Analytic Hierarchy Process Modeling of Land Use Change in Amman City, Jordan

Over the past two decades, rapid urban growth significantly altered land-use patterns in Amman, raising critical concerns regarding sustainability and food security. This study utilized an integrated Cellular Automata–Markov (CA–Markov) model, in combination with the Analytical Hierarchy Process (AHP), to simulate land-use and land-cover (LULC) changes and project future scenarios for 2031 and 2040. The CA–Markov model quantified temporal land-use transitions and simulated spatial growth patterns, while AHP served as a multi-criteria decision-making tool to determine the relative influence of key driving factors on urban growth. Landsat imagery from 2004, 2013, and 2022 was classified into three main categories: built-up areas, agricultural land, and barren land. The simulation framework incorporated key driving factors, including GDP per capita, population density, road accessibility, elevation, and slope. Model validation against actual 2022 LULC data yielded a high accuracy of 91.4% and a Kappa index of 0.89, demonstrating the reliability of the predictive framework. The results projected that built-up areas would increase from 257.35 km² (32.3%) in 2022 to 309.18 km² (38.9%) in 2031 and 349.17 km² (43.9%) by 2040, accompanied by a consistent decline in both agricultural and barren lands. Spatial analysis revealed that districts with higher population density, intense economic activity, and superior road accessibility were particularly susceptible to rapid urbanization. These findings highlighted the urgent need for proactive urban planning policies to protect agricultural land and manage growing infrastructure demands. While the CA–Markov model effectively replicated historical patterns, its reliance on past trends limited its capacity to anticipate sudden policy shifts or environmental shocks. Future research should prioritize integrating higher-resolution datasets, such as QuickBird imagery and detailed cadastral or infrastructure data, to improve the spatial accuracy of LULC simulations. In addition, the development of policy-driven and scenario-based models should incorporate urban growth boundaries, agricultural land protection policies, and transportation expansion plans. This would enable more realistic forecasting of land-use dynamics and provide stronger decision-support tools for resilient and sustainable urban development.