Evaluation of Methods on Mass Movement Hazard in Iran, Using Geographic Information Systems, First Phase: Taleghan basin with semi - humid

After reviewing literature and methodologies for landslide hazard zonation, an attempt was made for preparation of required data layers in geographic Information Systems (GIS) environment. Tow important and main data layers so called landslide inventory map and causative factor maps have to be prepared for this research. Therefore the former one was prepared through aerial photo interpretation, field survey and making use of geological map at scale I: 100,000. The parameter maps, were prepared by making use of existing thematic maps and satellite imageries. All of the prepared data layers were transferred into digital format, for further use in a GIS. Moreover, beside the graphical database an attribute database also was created for data storage, management and further analysis. Based on data availability, eight models for landslide hazard zonation were applied and evaluated in Taleghan basin in Iran. In order to have a real and impartial evaluation of applied methods, an index called Landslide index was introduced and used. After comparing landslide index in different predicted hazard classes, the following results were achieved; * Bivariate; statistical analysis is relatively good in zoning of landslide hazard in the study area. * Knowledge based modeling is excellent for predicting the low and moderate hazard' zones in Taleghan basin and surely the accuracy and reliability of this method can be improved by precise definition of the rule bases. * The method applied for translational landslide hazard zoning, Revised Nilsen model, Nilsen method and models based on Mora & Varson can be considered respectively to apply in the areas like Taleghan basin. Although those methods were evaluated in Taleghan basin, but still it is really difficult to choose a certain method and propose to apply it in such areas. It implies that none of the evaluated methods are enough qualified for accurate zoning of each, hazard classes. So the conclusion can be drawn that the integrated method should be taken into account. In the integrated method, the advantages of the evaluated models are accepted to use and the weak points of models are rejected, i.e, all of the advantages of the evaluated models are combined and integrated and introduced as a unique method. In the other words one -of the evaluated models might be used for mapping of only low hazard class, while the high hazard class can be mapped out through another known model. Application of the integrated method in Taleghan basin showed that there is a strong logical relation between the predicted hazard classes and the landslide indices and they are highly correlated.