A statistical approach for logical modeling of landslide hazard zonation

This paper presents a logical method for landslide hazard zonation, based on statistics. The multiple regression equation as the final model, describes the correlation between slope instability index as dependent variable (y) and five effective factors; elevation (x1), fault influence (x2), slope gradient (x3), lithology (x4) and slope aspect (x5) as independent variables. More than 140 major landslides were studied and the location and area of every single one of them was exactly indicated on a map called landslide location map. The maps of five independent variables are often available or can be prepared easily. The foundation of regression model is based on the combination of these five maps and the landslide location map.