Application of the SLEUTH model for land use prediction: Simulation of urban sprawl in the city of Meknes

Prospective modelling provides decision-makers with valuable insights by exploring possible futures and supporting territorial planning strategies. This study applies the SLEUTH cellular automata model to simulate urban sprawl in the city of Meknes by 2040. The model uses historical land use/land cover data (1990-2020), the road network, elevation and hillshade layers, as well as urban and exclusion maps. The methodology consists of two stages: preparation of input data and model execution. The first stage involved generating land use maps through supervised classification using the Support Vector Machine (SVM) algorithm. The second stage comprised the calibration process to identify the optimal values of the five key growth coefficients for prediction. The simulation results reveal a clear dynamic of urban expansion, concentrated along major road axes and in continuity with existing urban centres. High values of the road gravity (94) and slope (49) coefficients confirm the strong influence of road infrastructure and terrain on urban growth, while northern peripheral zones remain less developed due to rugged topography and limited road networks.