Urban Expansion Prediction in Karbala City, Iraq, Integrating GeoAI and an ANN–CA Model

Abstract

Predicting urban expansion is crucial for rational urban planning and sustainable resource management. This study presents an integrated GeoAI (Geospatial Artificial Intelligence) approach to simulate future urban growth patterns in Karbala City, Iraq. A hybrid ANN-CA model, combining Artificial Neural Networks (ANNs) and Cellular Automata (CA), was developed and calibrated within MATLAB using Landsat imagery (2015-2020). The simulation framework was further informed by principles of the Tietenberg model to account for resource utilization, population dynamics, and developmental pressures, ensuring a sustainability-oriented analysis of land consumption. Validation results, predicting the urban layout for 2025, show a substantial increase in urban area from 13% in 2015 to 24%, indicating intense urban expansion. This study demonstrates the ANN-CA model as a valuable GeoAI tool for urban geographers and planners. The findings provide critical insights for policymakers in Karbala to guide future urban expansion toward more orderly and sustainable development, aligning with the analytical perspectives of the Tietenberg model.