Geospatial mapping and prediction of informal settlement growth in Addis Ababa using GIS and logistic regression model

Abstract Informal settlements (ISs), characterized by unauthorized occupation and construction without legal recognition, represent a critical urban challenge in rapidly developing cities like Addis Ababa, Ethiopia, where 23.8% of the urban population resides in ISs, driving substantial socio-economic and environmental losses. This study integrates Geographic Information Systems (GIS) and Logistic Regression (LR) modeling to map historical IS growth (2002–2022) using SPOT 5, 6 and Sentinel-2 imagery and project future expansion to 2042. Findings reveal a 224% increase in IS area from 2,260.5 ha (2002) to 7,329.5 ha (2022), primarily converting farmland and open spaces. The validated LR model (Kappa = 0.73, Pseudo-R 2 = 0.32) identified proximity to existing ISs (strongest positive influence; OR = 0.45, p < 0.001), distance to roads (negative influence; OR = 0.74, p < 0.01), and low population density (positive influence; OR = 1.65, p < 0.05) as statistically significant drivers. Projections under current trends indicate IS expansion to 11,483.5 ha (95% CI 10,850 –12,117 ha) by 2032 and 18,557.9 ha (95% CI) by 2042, with uncertainty quantified through sensitivity analysis and assuming static driver conditions. Model robustness also confirmed via spatial validation and advanced error metrics (quantity/allocation disagreement 10%). The study underscores regulatory inefficacy and urges spatially targeted interventions particularly protecting vulnerable public lands in high-probability peri-urban peripheries providing an evidence base for adaptive urban policy reform in Addis Ababa.