GIS based frequency ratio and index of entropy models to landslide susceptibility mapping (Daguan, China)

Abstract

The main goal of this study is to produce landslide susceptibility maps by frequency ratio (FR) and index of entropy (IoE) models based on geographic information system (GIS) for the Daguan county of Yunnan province, China. At first, landslide locations were identified by earlier reports and aerial photographs as well as by carrying out field surveys, and a total of 194 landside locations were constructed from various sources. Then the landslide inventory map was randomly split into a training dataset 70 % (136 landslide locations) for training the FR and IoE models and the remaining 30 % (58 landslides locations) were used for validation purposes. Fifteen landslide conditioning factors were constructed in GIS. The susceptibility maps produced using FR and IoE models were divided into five different susceptibility classes such as very low, low, moderate, high, and very high. Finally, the validation of landslide susceptibility maps was carried out based on training dataset and the validation data using the area under the curve (AUC) method. The success rate curve showed that the area under the curve for FR and IoE models were 0.8191 and 0.8109 accuracy, respectively. Similarly, the validation result showed that the prediction accuracy of the two models was 81.75 % for FR model and 81.44 % for IoE model, respectively. The maps produced by both models exhibits better satisfactory properties. The landslide susceptibility maps may be useful for planning purposes for environmental protection and natural hazard mitigation studies in the near future.