Original paper

An analytic multicriteria hierarchical approach to assess landslide vulnerability. Case study: Cornu village, Subcarpathian Prahova Valley/Romania

Armaş, Iuliana

Abstract

Three different types of methodological approaches to landslide vulnerability can be separated in the geomorphologic literature: qualitative, semi-quantitative and quantitative. The present study aims to test a semi-quantitative method of landslide vulnerability assessment in relation to population and road distribution, using an integrated multi-criteria method. This approach is focused on a case study from the Prahova Sub-Carpathians, respectively from Cornu village. The test area was selected based on a set of several criteria, including: small-surface crowding of clear-cut morphological differentiations between the riser and the tread of the major Sub-Carpathian terrace (the second terrace), proneness of the slopes to landslides and increase of the antropic impact during the last decades. The reduced surface of the testing area represented a pre-condition that allowed checking the level of sensitivity of the applied methods through objective field surveys. The spatial analytic hierarchy methodology used in this research involved the creation of two decision trees: environmental vulnerability (the hierarchical structure being built by the landslide susceptibility factor, the assessment of the building environment between 1970 and 2007 and land use) and exposure (estimated based on a Distance analysis). The Spatial Multi Criteria Evaluation (SMCE) module implemented in ILWIS software was used for combining and weighing the different vulnerability factors. The susceptibility factor was calculated using the bivariate landslide susceptibility index (LSI). For this purpose, several thematic layers - including landslide inventory, lithology, slope gradient and aspect, land cover etc. - were prepared. The interactions between environmental items and landslide distribution were tested and the importance of individual criteria for landslide occurrence was defined. The results allowed an objective assessment of the advantages and disadvantages implied by the applied quantitative and semi-quantitative methods. As a general conclusion, the semi-quantitative multi-criteria analytic hierarchy analysis is most sensitive to the input data. The major pitfall of this method is represented by the fact that the outcome depends significantly on the expertise of the researcher and the accuracy of the primary data.

Keywords

subcarpathianslandslides susceptibilitylandslide susceptibility indexanalytic hierarchy processvulnerability assessment