2002 Volume 54 Issue 6 Pages 4-15
In this study, data mining based on the rough set theory is used for specifying critical factors that would cause slope disasters, such as, hillside landslides and debris flows, and for determining the relationship between these factors and disasters. Data on debris flow from the Hiroshima disaster on June 29, 1999, was used to illustrate this methodology. The area of the disaster geologically consists of granite. First, using the data from the primary valleys, we defined the critical geological and rainfall factors that caused the disaster. Then we identified the steps of sediment movement. The critical factors that contribute to transporting sediments are the maximum hourly rainfall, the steepest riverbed slope, and the width of a stream. We also identified particular steps for hillside landslides that resulted in debris flows. It was shown that the area of collapsed hillside and the slope directly under it are the critical factors for an occurrence of debris flow.