Modeling landslide susceptibility of a mountain forests using Adaptive Neuro-Fuzzy Inference System (ANFIS) for forest road planning

• Main Authors: Ismaeil Ghajar, Akbar Najafi
• Format: Article
• Language: fas
• Published: Research Institute of Forests and Rangelands of Iran 2014-11-01
• Series: تحقیقات جنگل و صنوبر ایران
• Subjects: Landslide susceptibility; Neuro-fuzzy; model; ANFIS; forest road
• Online Access: http://ijfpr.areeo.ac.ir/article_12435_a259a3ae3fdb8f4748d44116279ee358.pdf
• ISSN: 1735-0883; 2383-1146

This study presents landslide susceptibility (LS) prediction model using the Adaptive Neuro Fuzzy Inference System (ANFIS) and Geographic Information System (GIS) which incorporates the physiographic information. Such models are is useful for forest road planning. To this aim, a set of factors including the terrain slope, aspect, geology formation, curvature, distance to rivers, and distance to faults at occurred landslide points were integrated into the ANFIS model. The modeling using a subtractive clustering method returned a coefficient of determination (R2) of 0.73 and a root mean square error (RMSE) of 0.27 for the best model. The sensitivity analysis indicated the distance to the rivers, geology formation, terrain slope, curvature, distance to the faults, and aspect as the most effective factors on the landslide occurrence. Furthermore, an evaluation of existing roads on simulated LS map showed that the majority of the currently existing roads are located on “medium” and “high” LS classes.