INTEGRATION OF GENETIC ALGORITHM WITH SLEUTH FOR URBAN GROWTH MODELING IN AL-SULAYMANIYAH CITY, NORTHERN IRAQ

Abstract

This work introduces a novel fusion of genetic algorithm (GA) optimization with the SLEUTH cellular automaton model to simulate and forecast urban growth dynamics in Al-Sulaymaniyah City, Kurdistan Region, Iraq. To overcome the computational constraints of conventional SLEUTH calibration, we devised a GA-based optimization framework that attained a 3-22x acceleration while preserving good accuracy. The GA optimization effectively calibrated the SLEUTH growth coefficients, achieving an Optimal SLEUTH Metric of 0.718±0.16 and a Lee-Sallee index of 0.900±0.05. Model validation revealed a continuous increase in accuracy from 83.45% in 2004 to 91.27% in 2024, signifying an improved ability to capture modern urban dynamics. Projections for 2034 anticipate a total urban area of 878.20 km², indicating a 511.1% increase from the 1995 baseline. Spatial analysis indicated a significant northwestern directional bias (549% increase) and a diminishing connection with road proximity over time (r = -0.188 to -0.073), implying a transition from infrastructure-dependent to multi-factor-driven urban expansion. The model delineated specific growth phases: a period of rapid expansion (1995-2010) averaging 23.84 km²/year, succeeded by a reduction to 5.79 km²/year (2020-2034), signifying impending geographical saturation. The SLEUTH-GA framework offers urban planners quick and precise tools for simulating intricate urban systems and developing evidence-based sustainable policies in swiftly expanding Middle Eastern cities.