数値地理情報と降雨極値データを利用した土砂災害発生確率モデルの構築

概要

論文の詳細を見る
A probability model of sediment hazard was made by multiple logistic regression analysis. Relief energy, hydraulic gradient due to extreme precipitation of return period and geological features are used as parameters in this model. The risk has been analyzed quantitatively by giving a probability through setting the cause of the sediment hazard. The evaluation of the probability distribution in Japan has been performed by using those conditions that are expressible by digital geographic information. The probability distribution is composed of the 1 km× 1 km resolutions that can reflect the social risk and global changes. Dangerous areas could be extracted according to the probability of showing quantitative risk. These results are useful for the decision of the areas for the countermeasure plan. In summary, the results are as follows. (1) The urban area that extends a Chugoku mountainous district is at high risk with sediment hazard by extreme precipitation of 10-year return period. (2) The high risk areas have been extracted from the urban areas in Japan by extreme precipitation of 100-year return period, and an enormous economic loss has been predicted. (3) Probability of the resolutions has been verified to reflect information on probability of 50m×50m resolutions.
日本自然災害学会の論文
2008-05-31