Article Details
GEOGRAPHICAL INFORMATION SYSTEMS BASED USING LOGISTIC REGRESSION LANDSLIDE SUSCEPTIBILITY ASSESSMENT OF THE ÇUBUK-KALECİK (ANKARA) BETWEEN ŞABANÖZÜ (ÇANKIRI) REGION
Indexed In
Volume 155 / 2017Authors:
Hasan ELMACI, Senem TEKİN, Nail ÜNSALKeywords:
Çankırı,Geographic Information System (GIS),Landslide Susceptibility,Logistic regression.Abstract:
Landslides frequently occur particularly in certain parts in our country as well as in the world. They sometimes turn into natural disaster and cause socio-economic losses in significant magnitude. In this study, the landslide susceptibility analysis was performed in a total area of 2360 km2 at 1/100,000 scale in map sheet of Çankırı-H30. According to the inventory map, 4,79% of the study area is formed of landslide areas and the movement type of available landslides are generally observed in the form of rotational slip. According to the database of the Turkish Landslide Inventory Map prepared by MTA, there are 876 landslides, which cover an area of 122 km2, in the study area. A total of twelve independent variables for landslide conditioning factors were taken into consideration in susceptibility assessments being as; lithological map, land-use classification, digital elevation model, hill slope, profile, planar and tangential slope curvatures, roughness index, slope/aspect ratio, stream power index and topographic wetness index. The landslide susceptibility assessment was carried out using multivariate logistic regression method which is regarded as one of the statistical methods. As mapping unit; 25 x 25 m pixels were used for the statistical assessment. The obtained probability values of the landslide susceptibility maps were assessed in five different grades ranging from very low to very high. The performance evaluation of the susceptibility maps were performed by using the Receiver Operating Characteristic (ROC) and prediction success rate curves. The area under ROC curves were found to be in 0.794. Highly and very highly susceptible areas correspond to 28.06% of the entire area in which 78% of the landslides occurred. This accomplished, high performance landslide susceptibility map will make significant contributions to medium scale planning strategies.
DOI: 10.19111/bulletinofmre.306692
