Abstract
In this paper the authors discuss not only a prediction method but a risk evaluation method for landslide occurrence by applying the vegetation index which is calculated from LANDSAT/TM data. We have examined the variation of vegetation index between the years of 1986 and 1989. The result shows the validity of this prediction method. The risk evaluation has been accomplished by using the Hayashi's quantification analysis of classes II and III in the following cases: (1) not applying the vegetation index, (2) applying the variation of the vegetation index and (3) applying the vegetation index. Finally, we have reached such conclusion that the result in case (3) has been appropriate to the risk evaluation for landslide occurrence.
