Landslide susceptibility mapping using frequency ratio, analytic hierarchy process, logistic regression, and artificial neural network methods at the Inje area, Korea

• Published in: Environmental earth sciences Vol. 68; no. 5; pp. 1443 - 1464
• Main Authors: Park, Soyoung, Choi, Chuluong, Kim, Byungwoo, Kim, Jinsoo
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer-Verlag 01.03.2013; Springer; Springer Nature B.V
• Subjects: Biogeosciences; Correlation analysis; Earth and Environmental Science; Earth Sciences; Earth, ocean, space; Engineering and environment geology. Geothermics; Environmental Science and Engineering; Exact sciences and technology; Geochemistry; Geology; Hydrology/Water Resources; Landslides; Landslides & mudslides; Mapping; Natural hazards: prediction, damages, etc; Neural networks; Original Article; Regression analysis; Terrestrial Pollution; South Korea; Far East; Asia; Korea; Korea, Rep; Analytic hierarchy process (AHP); Artificial neural network (ANN); Logistic regression (LR); Frequency ratio (FR); Landslide susceptibility index (LSI); damage; models; maps; neural networks; accuracy; frequency; correlation; landslides
• ISSN: 1866-6280; 1866-6299
• DOI: 10.1007/s12665-012-1842-5

Every year, the Republic of Korea experiences numerous landslides, resulting in property damage and casualties. This study compared the abilities of frequency ratio (FR), analytic hierarchy process (AHP), logistic regression (LR), and artificial neural network (ANN) models to produce landslide susceptibility index (LSI) maps for use in predicting possible landslide occurrence and limiting damage. The areas under the relative operating characteristic (ROC) curves for the FR, AHP, LR, and ANN LSI maps were 0.794, 0.789, 0.794, and 0.806, respectively. Thus, the LSI maps developed by all the models had similar accuracy. A cross-tabulation analysis of landslide occurrence against non-occurrence areas showed generally similar overall accuracies of 65.27, 64.35, 65.51, and 68.47 % for the FR, AHP, LR, and ANN models, respectively. A correlation analysis between the models demonstrated that the LR and ANN models had the highest correlation (0.829), whereas the FR and AHP models had the lowest correlation (0.619).