Abstract
In a disaster area of 814ha affected by two typhoons, distinctions between points with and without small landslides of less than 0.1ha, were made by using the “Quantification II” method. Only four geomorphic factors, easily measurable on a topographic map of 1/5, 000 scale, were used as “Items” in this method. The partial correlation coefficients of the “Items” in reverse order were longitudinal-slope form<inclination<exposure<lateral-slope form. The ranges in reverse order were longitudinal-slope form<exposure<inclination<lateral-slope form. The susceptibility to small-sized landslides was found to be great at points that were on laterally concave, steep south slopes. Longitudinalslope form was not a significant factor. The stand condition, which had been proved to be a large contributer to distinguishment between the presence or absence of landslides on areas of one hectare, was not used as an “Item” in this analysis. Nevertheless, the correlation ratio indicating the accuracy of the distinguishment clearly was larger in this analysis than in most of the usual analyses using stand condition as an “Item”. This improvement in the accuracy of distinguishing was thought to have been brought about partly by limiting the distinguishing to landslide of less than 0.1 ha and partly by reducing the unit of examination from an area to a point. The availability of the results for the planning of the disposition of protection forests was thought to be enhanced by excluding the stand condition from the analysis.
