Socioeconomic and Environmental Scenario-based Modeling for Prediction of Future Urban Growth in Riyadh, Saudi Arabia

Keywords: Prediction, urban growth, scenario, modeling, socioeconomic, environment , cellular automata, Markov chain, MLP neural network, Saudi Arabia
Abstract Type: Paper Abstract

Abstract

The growth of urban areas is a dynamic phenomenon resulting from an increasing population. Growth in urban areas contributes significantly to many global land use and cover changes. This research investigates the patterns and trends of future urban growth in Riyadh, Saudi Arabia, between 2030 and 2050, as there is a lack of studies incorporating socioeconomic and environmental variables to predict urban growth in Riyadh. Our study includes two models— cellular automata and Markov chain (the CA-Markov) and multi-layer perceptron (MLP neural network). We developed three scenarios—business as usual (BAU), rapid economic growth (REG), and integrated environmental sustainability (IES)—and analyzed them using both approaches. The validation process was performed using the Kappa Index of Agreement (KIA) to spatially and statistically assess the simulation maps. The results show that the overall Kappa statistics scores were moderate. The Kappa standard for the CA-Markov chain model for 2019 was 0.67, 0.68, and 0.68 for BAU, REG, and IES, respectively. The MLP neural network model was 0.75, 0.76, and 0.76 for BAU, REG, and IES. Both models showed that Riyadh’s actual and predicted urban growth had increased significantly. Comparing the 2019 map to the 2019 prediction map, the MLP neural network model indicated a realistic increase in urban areas, as the difference between the actual and predicted growth was only 94.02 km2 compared to 320.61km2 in the CA-Markov model.