Dynamic suitability-weighted CA-Markov model for projecting urban growth and thermal impacts: a case study of Abuja

Traditional urban growth models often decouple land-use change from its climatic consequences, creating planning blind spots. This study introduces a globally transferable Dynamic Suitability-Weighted CA-Markov (DSW-CA-Markov) framework that, for the first time, integrates Land Surface Temperature (LST) trends as dynamic suitability factors within Cellular Automata transition rules, enabling bidirectional urban-thermal feedback simulation. We develop and validate this framework using multi-temporal Landsat data (2010, 2015, 2020) from Abuja, Nigeria, then project integrated urban-thermal patterns to 2030. Our primary innovation is a dynamic feedback mechanism where pixel-level LST change rates are embedded as evolving suitability factors within CA transition rules, moving beyond static suitability mapping or post-hoc thermal correlation. Results reveal a 157.29% built-up increase (2010-2020) with LST rises of 3.6 °C, and projected continued expansion with amplified UHI effects. The DSW-CA-Markov framework demonstrates superior capability in simulating coupled urban-thermal dynamics compared to conventional CA-Markov approaches (Kappa improvement: 0.08; thermal : 0.73 vs. 0.65). This study provides both a novel methodological template for climate-responsive urban modelling and crucial insights for sustainable planning in fast-growing cities globally.