In reaction to storm and flood damages which threaten inhabitants' life and their properties, local governments give out evacuation calls and directives, and most of the inhabitants start evacuation by these information. However, the awareness for evacuation of the inhabitants is very poor, and the rate of the evacuation is low at present even if self-motivated evacuation is needed. The evacuation behaviors of the inhabitants upon an outbreak of a disaster are strongly influenced by psychological factors, such as, “normalcy bias”. Therefore, to clarify the mentality and action process of the inhabitants for evacuation and draw up the formulation of evacuation action plan which considers well the psychological action of the inhabitants are required at the same time.Therefore, this study extracted psychological factors such as the attitude and responsibilitiness which concerning evacuation and assumed process of evacuation. This study classified these processes in four psychological levels, 1) information transfer, 2) awareness of importance, 3) awareness of needs, 4) intendment by combining some psychological factors. For verifying the appropriateness of our hypothesis, we conducted a questionnaire survey about landslide. And also tried to attempt the classification of the disincentives which ruled out the process of evacuation behavior in accordance with assumed evacuation behavior process and surveyed its validity and effectiveness in flood damages, such as sediment disaster and flooding. As a result of questionnaire survey, it is cleared that there is a tendency of the inhabitants' psychological levels for their evacuation behavior process as we assumed. And we classified disincentives of inhabitants' evacuation behavior process, and suggested the applicability and the validity of applying psychological levels for the evacuation action plans.
Many hillside landslides and debris flows occurred by typhoon T 0421 in Miyagawa village （the present Odai town） on September 29, 2004. These caused severe sediment-related disasters. In this study, several findings on the influence of precipitation, slope angle and forest practice on hillside landslides caused by typhoon T 0421 were shown in Miyagawa-dam basin as follows : 1） Correlation between precipitation and collapsed area ratio was not clear. 2） Angle of collapsed slope in deforestation sites was 5-10° lower than those of non-forestation and reforestation sites. 3） Collapsed area in deforestation sites occupied 57% in total collapsed area. 4） Most of landslides were occurred since 10 to 20 years after deforestation. 5） More hillside landslides observed in larger deforestation area. 6） A prediction model of slope stability changes considering forest practice （Kuroiwa and Hiramatsu, 2004） was applicable to Miyagawa-dam basin. As the result of this simulation, collapsed area is generally in proportion to collapsing possibility area ratio index （APhr（t））. 7） Angle of collapsed slope and APhr（t）and collapsed area ratio showed significant correlation.
A very rare and heavy rain-storm for a few hours in Shobara-City, Hiroshima Prefecture, induced a lot of slope failures and debris flows in the afternoon on July 16, 2010. Damaged area was located within only a square of about 5km. This mountainous area is geologically covered by Takata rhyolitic rocks, Kisa andesitic rocks, and Bihoku sedimentary rocks. Because the surface soil layer was not thick, most of failure-depths were 0.5-1.5m. Countless small holes and fissured bed-rocks, through which water-flow pushed out surface soil layer, were observed around most of slope failures. Almost displaced materials turned to debris-flows or a kind of liquefied state and roared down into main stream, and then flowed down as a large hyper-concentrated flow. The rainfall as a cause of this disaster was statistically about 240 years of return-period to two-hour-rainfall, and about 5,700 years to three-hour-rainfall in this district. Areal ratio of slope failures steeper than 30° to this same watershed area was 7.0% for Shinodo-area, and 4.7% for Saki-Oto-area.