Abstract
Existing risk assessment systems for sediment-related disasters, such as“Sediment Disaster Warning”, usually evaluate risk from precipitation data and do not incorporate premonitory phenomena. Premonitory phenomena have three uncertainties:1)when a premonitory phenomenon occurs, it is uncertain whether it will be seen;2)if a phenomenon is assumed to be a premonitory phenomenon, it is uncertain whether a sediment-related disaster will really occur;and 3)when no premonitory phenomenon is reported, it is uncertain whether the inhabitants actually saw the phenomenon. In this study, the authors built a risk assessment system for sediment-related disasters that utilizes data on premonitory phenomena and precipitation. The system considers the uncertainty of premonitory phenomena. It uses precipitation data as the main factor for evaluating the degree of risk of a sediment-related disaster, and the information on premonitory phenomena is used as a sub-factor. The system evaluates the risk of a sediment-related disaster for each evacuation unit, when ongoing sediment disaster warning systems indicate a degree of risk for each local authority. Two cases were evaluated using the system:the case of Typhoon MEGI at Niihama City in 2004 and the case of Typhoon Nabi at Tarumizu City in 2005. The following results were obtained:1)the system can specify high-risk areas in a local authority;and 2)by using information on premonitory phenomena, high-risk areas were specified 2 hours before the sediment-related disasters occurred.
