Cellular Automata–Artificial Neural Network Modelling for Spatio-Temporal Land Use/Land Cover Dynamics

Context and background: Understanding land use and land cover (LULC) dynamics is critical for sustainable management in regions undergoing rapid socio-economic and environmental change. Goal and objectives : This study analyzed decadal LULC shifts in Mazowe District, Zimbabwe, from 2014 to 2024, employing multi-temporal Landsat 8 and 9 imagery, Maximum Likelihood Classification (MLC), and a Cellular Automata–Artificial Neural Network (CA–ANN) modelling approach. Methodology:  Classification accuracy was high, with overall accuracies above 94% and Kappa coefficients over 0.91, confirming model robustness. Results :  The results revealed substantial forest cover loss (−33.03%), a marked expansion of plantation agriculture (+158.78%), and variable trends in agricultural and urban areas, while bare land remained the dominant land use (>58%). Transition matrices highlighted high stability in bare land and water bodies, in contrast to weak persistence in agriculture and forests, which frequently converted to degraded surfaces. Key biophysical factors influencing land conversion included elevation, slope, and proximity to rivers. Model validation against 2024 observations yielded substantial agreement (κ = 0.67), underscoring the reliability of CA–ANN for projecting spatial-temporal land changes. Business-as-usual forecasts to 2034 predict bare land expanding beyond 71% and forest declining below 9%, with urban areas initially growing before contracting. These trends underscore escalating ecological degradation, reduced agricultural productivity, and hydrological stress risks. The findings provide vital insights for sustainable land management planning, emphasizing reforestation, soil, and water conservation, and strengthening the institutional frameworks necessary to balance agricultural development with ecological resilience. This study contributes valuable scientific evidence to inform land management in Mazowe and similarly affected sub-Saharan African landscapes, fostering sustainable futures for livelihoods and ecosystems.