16th International Multidisciplinary Scientific GeoConference SGEM2016, Informatics, Geoinformatics and Remote Sensing, 2016
In Moldavian Plateau (Eastern Romania) landslides affect large areas, causing the exclusion of va... more In Moldavian Plateau (Eastern Romania) landslides affect large areas, causing the exclusion of valuable land from economic exploitation. Generally, landslide risk assessment assumes the analysis and the coupling of landslide susceptibility, landslide hazard, and landslide vulnerability. Regarding the latter there is little research in the literature focused on its measurement. Recent development of statistical tools and of geographic information systems can facilitate fast and accurate computations, including mapping of this component. This study aims to develop a simple model for mapping physical vulnerability to landslides at local level, based on some spatial data indicators. The model takes into account the relationship between the susceptibility to a particular hazard and the degree of losses suffered by the elements at risk, more precisely, it is constructed by combining the susceptibility to landslides with a weighted kernel density of the elements at risk. Both parameters can be assessed with sufficient accuracy by applying different quantitative methods in GIS environment: 1) the logistic regression approach was used to assess the landslide susceptibility by taking into account several quantitative and qualitative landslide contributing factors (slope angle, distance to drainage network, mean annual precipitations, land use etc.); 2) the Gaussian function was used to assess the probability density of elements at risk (buildings, roads, electric network). A suggestive map providing a satisfactory image of landslide vulnerability is obtained by combining these two parameters. Their consideration might prove useful in landslide vulnerability assessment especially in areas with little or no data on landslide hazard and on the elements at risk.
Uploads
Papers by Adrian Grozavu