Landslide susceptibility mapping using GIS-based logistic regression model in Sekondi- Takoradi Metropolitan Area of Ghana

• Published in: Discover applied sciences Vol. 7; no. 9
• Main Authors: Ankah, Mawuko Luke Yaw, Frimpong, Reuben Akwasi, Odum, Ernest Kojo, Meten, Matebie, Klu, Albert Kafui
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 29.08.2025
• Subjects: Applied and Technical Physics; Chemistry/Food Science; Earth Sciences; Engineering; Environment; Materials Science; GIS; Landslides; Hazard maps; Logistic regression model
• ISSN: 3004-9261
• DOI: 10.1007/s42452-025-07147-2

Landslides pose a significant threat to human life and economic development worldwide. In Sekondi-Takoradi, a twin city in the Western Region of Ghana, a detailed landslide hazard assessment was conducted in the metropolis. This study aimed to identify the key triggers of landslides and develop an accurate landslide susceptibility map for the study area, including the metropolis. Logistic Regression, a statistically based model, was employed to determine the likelihood of landslide occurrence based on key geo-environmental factors. These factors, ranked in order of their causative influence, include land-use-land-cover, NDVI, soil type, aspect, slope angle, rainfall, curvature, proximity to faults, elevation, TWI, and lithology. The landslide susceptibility map was created by integrating raster maps of these factors, classifying the metropolis into five susceptibility zones: very low (12.0%), low (13.3%), moderate (17.7%), high (19.7%), and very high (37.8%). Highly populated areas at risk include Kojokrom, Mpentsem, Bakaekyir, Kweikuma, Fijai, Kansawrodo, Essikado, Ngyiresia, Essipong, Osofokrom, and Takoradi towns. Field observations and historical landslide data confirmed that most landslides occurred in areas identified as highly susceptible by the model. The predictive performance of the model was validated using the Receiver Operator Characteristic (ROC) curve, yielding an Area Under the Curve (AUC) value of 0.74, indicating good model performance. The findings of this research are expected to contribute to urban planning and disaster risk reduction strategies in the study area and contribute meaningfully to achieving the Sustainable Development Goal (SDG 11).