Journal
GEOJOURNAL
Volume 86, Issue 4, Pages 1797-1816Publisher
SPRINGER
DOI: 10.1007/s10708-020-10162-y
Keywords
Landslide susceptibility; AHP; Fuzzy; ROC curve; Gilan province
Categories
Ask authors/readers for more resources
This study utilized AHP and fuzzy methods to map landslide susceptibility in Gilan province, finding that 29.53% and 36.53% of the total area were considered high and very high-risk areas, respectively. The AUC values for AHP and fuzzy maps were 92.4 and 91.9, indicating high prediction accuracy.
Landslides are natural destructive phenomena that can cause great damage to property and life loss. One of the fundamental proceedings to reduce the possible damage is identifying landslide-prone areas through different knowledge and data driven methods. Since Gilan province has a high potential for landslides occurrence, thus, the present study goes through to map landslide susceptibility. To accomplish this, two methods of AHP and fuzzy were used and then ROC/AUC curves have been preferred to evaluate the susceptibility map's performance. In this study, seven input layers of landslide causative factors including slope, lithology, land-use, rainfall, distance to fault, distance to road and distance to river were considered. The landslide susceptibility map derived from AHP method was obtained after assigning the weights to different layers, which were all based on the expert's judgments. According to this map, 29.53% of total-area, were considered as high and very high-risk areas. In the fuzzy method, three scenarios were adopted to combine the layers. Among them, the third scenario showed the best result. The output map of this scenario has devoted 36.53% of total area to high and very high-risk areas. Prediction accuracy of these maps showed the values of AUC equal to 92.4 and 91.9 for AHP and fuzzy maps, respectively. Both of these maps, mainly, introduced some parts of central, south and southeast areas of Gilan as landslide-prone areas.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available