An integrated artificial neural network model for the landslide susceptibility assessment of Osado Island, Japan

• Published in: Natural hazards (Dordrecht) 2015-09, Vol.78 (3), p.1749-1776
• Main Authors: Dou, Jie, Yamagishi, Hiromitsu, Pourghasemi, Hamid Reza, Yunus, Ali P., Song, Xuan, Xu, Yueren, Zhu, Zhongfan
• Format: Article
• Language: English
• Subjects: Artificial neural networks; Back propagation; Civil Engineering; Density; Disasters; Drainage density; Earth and Environmental Science; Earth Sciences; Emergency preparedness; Environmental Management; Geographic information systems; Geophysics/Geodesy; Geotechnical Engineering & Applied Earth Sciences; Hydrogeology; Islands; Landslides; Landslides & mudslides; Learning theory; Lithology; Mapping; Mathematical models; Natural Hazards; Neural networks; Original Paper; Remote sensing; Slope stability
• ISSN: 0921-030X; 1573-0840
• DOI: 10.1007/s11069-015-1799-2

The objective of this study was to select the maximum number of correlated factors with landslide occurrence for slope-instability mapping and assess landslide susceptibility on Osado Island, Niigata Prefecture, Central Japan, integrating two techniques, namely certainty factor (CF) and artificial neural network (ANN), in a geographic information system (GIS) environment. The landslide inventory data of the National Research Institute for Earth Science and Disaster Prevention (NIED) and a 10-m digital elevation model (DEM) from the Geographical Survey of Institute, Japan, were analyzed. Our study identified fourteen possible landslide-conditioning factors. Considering the spatial autocorrelation and factor redundancy, we applied the CF approach to optimize these set of variables. We hypothesize that if the thematic factor layers of the CF values are positive, it implies that these conditioning factors have a correlation with the landslide occurrence. Therefore, based on this assumption and because of their positive CF values, six conditioning factors including slope angle (0.04), slope aspect (0.02), drainage density network (0.34), distance to the geologic boundaries (0.37), distance to fault (0.35), and lithology (0.31) have been selected as landslide-conditioning factors for further analysis. We partitioned the data into two groups: 70 % (520 landslide locations) for model training and the remaining 30 % (220 landslide locations) for validation. Then, a common ANN model, namely the back-propagation neural network (BPNN), was employed to produce the landslide susceptibility maps. The receiver operating characteristics including the area under the curve (AUC) were used to assess the model accuracy. The validation results indicate that the values of the AUC at optimized and non-optimized BPNN were 0.82 and 0.73, respectively. Hence, it is concluded that the optimized factor model can provide superior accuracy in the prediction of landslide susceptibility in the study area. In this context, we propose a method to select the factors with landslide occurrence. This work is fundamental for further study of the landslide susceptibility evaluation and prediction.