http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Qiqing Wang,Wenping Li,Maolin Xing,Yanli Wu,Yabing Pei,Dongdong Yang,Hanying Bai 한국지질과학협의회 2016 Geosciences Journal Vol.20 No.5
The aim of this study was to apply and to verify the use of artificial neural network (ANN) and weight of evidence (WoE) models to landslide susceptibility mapping in the Gongliu county, China, using a geographic information system (GIS). For this aim, in this study, a landslide inventory map was prepared using earlier reports and aerial photographs as well as by carrying out field surveys. A total of 163 landslides (70% out of 233 detected landslides) were randomly selected for model training, and the remaining 70 landslides (30%) were used for the model validation. Then, a total number of twelve landslide conditioning factors, such as slope angle, slope aspect, general curvature, plan curvature, profile curvature, altitude, distance to rivers, distance to roads, lithology, rainfall, normalized difference vegetation index (NDVI), and sediment transport index (STI), were used in the analysis. Landslide hazardous areas were analyzed and mapped using the landslide-occurrence factors by ANN and WoE models. Finally the output maps were validated using the area under the curve (AUC) method. The validation results showed that the ANN model with a success rate of 82.51% and predictive accuracy of 77.31% performs better than WoE (success rate, 79.82%; predictive accuracy, 74.59%) model. Overall, both models showed almost similar results. Therefore, the two landslide susceptibility maps obtained were successful and can be useful for preliminary general land use planning and hazard mitigation purpose.