Predicting flood-vulnerability of areas using satellite remote-sensing images in Kumamoto City, Japan

Flood is a natural disaster that occurs almost every year in Japan. According to the ood record, it occurs during the rainy season, around July of each year. e aim of this paper is to predict areas vulnerable to ooding in the Shiragawa watershed. is study was carried out using DEM data, ALOS AVNIR-2 imagery and Amedas data to produce a watershed area, a vegetation index, a land-cover map and an isohyet map. DEM data, with spatial resolution of 10 meters, were derivedfrom the Geospatial Information Authority of Japan (GSI) to show topography and the watershed. e ALOS AVNIR-2 image was used to create the land-cover map and the vegetation index. e land-cover map was created by the unsupervised method and then veri‰ed by use of the landcover map of the Geospatial Information Authority of Japan (GSI). e vegetation index was created by use of the Normalized Di‹erence Vegetation Index (NDVI) algorithm. e isohyet map was obtained by use of datafrom rain-gauge stations in Kumamoto Prefecture and then interpolating by the application of the kriging method. All spatial data were overlaid to create the ood-vulnerability map by application of the Geographic Information System (GIS) model. is study combines all the data to predict areas vulnerable to ood. e result indicates that ood occurs in the middlepart of the Shiragawa watershed.