Date

2002

Author

Süzen, Mehmet Lütfi

Metadata

Show full item record

Item Usage Stats

186
views

239
downloads

Cite This

The purpose of this thesis is to generate and evaluate a concise system of data driven landslide hazard assessment procedure. The Asarsuyu catchment area and Bolu Mountain highway pass are selected for this purpose In this study, three data domains, which are remote sensing products, geological maps and topographical maps are used. The landslides of four different historical periods are interpreted using aerial photographs. Land cover is extracted from satellite images. Thirteen sets of parameter maps are produced from geological and topographical maps and from remote sensing products. In order to store the information of these parameter maps in a concise thematic database a 25x25 meter grid is overlaid to the area. Two different sets of points are defined. First one representing the properties of slided masses and the other set representing the conditions contributing to the sliding phenomena. The information falling on these points are stored in five separate thematic landslide attribute databases, as two main databases are Seed Cells and Slided Mass Databases, with accessory three other as, Polystats, Fuzzystats and photo-characteristics databases. Following the creation of the databases, theinformation stored are evaluated and preliminary landslide generating decision rules are extracted. In conjunction with these preliminary results a hazard assessment procedure, starting from the least detailed and simple ranging up to most complex statistical analyses are initiated. Finally a relative accuracy assessment procedure is carried out using the hazard maps produced. It is found that the most accurate, reliable and realistic results concerning the landslide hazard assessment of Asarsuyu catchment are obtained through logistic regression analyses. The results of landslide hazard assessment of Asarsuyu catchment shows that: the northern slopes of the Asarsuyu catchment is classified as very low hazard due to the presence of low population, undisturbed dense forest land cover, being very distant to E-5 highway and to the major active fault and the favorable lithological conditions. The southeastern slopes especially the Bolu Mountain Highway pass are definitely on very high hazard class due to the removal of lateral supports by E-5 highway cut slopes, close location to active faults, high disturbance of the land cover, high traffic activity along the highway resulting in extra vibration, and the presence of flyschoidal units

Subject Keywords

Landslide hazard analysis, Geographic information systems, Remote sensing, Asarsuyu catchment, Bolu mountain, Landslide hazard assessment, Logical regression

URI

https://hdl.handle.net/11511/12430

Collections

Graduate School of Natural and Applied Sciences, Thesis