Author(s): Hajime Shirozu; Koji Asai
Linked Author(s): Koji Asai
Keywords: Remote sensing; VNIR; SAR; Agricultural reservoir; Slope failure
Abstract: Heavy rain in July 2018 in West Japan caused serious damage in small agricultural reservoirs. It is difficult to know the damage situation of many reservoirs in case of sudden disaster. Therefore, the utilization of the satellite remote sensing is very useful to detect the damaged reservoirs. However, it is difficult to directly extract the damaged reservoir because the reservoir is too small. In this study, quick way to detect the slope failure behind the reservoirs is tried by analyzing SAR satellite images and optical satellite images. The analyses were carried out using the multiband image after the disaster obtained from World View -2 and HH single polarization image before and after the disaster obtained from PALSAR-2. In the analysis by optical image, the difference between GSI and NDVI was used to extract the slope failure areas. The slope failure areas behind the reservoir was detected by extracting only the slope located at a slope larger than 30 degrees and the radius of the reservoir within 30 m. According to the official report from Yamaguchi local government, 4 slope failures behind the agriculture reservoirs in Iwakuni City were confirmed. Our extraction method with the optical image can detect 6 slope failures which are suspected to damage the reservoirs. In the analysis by SAR image, color mixing method was used. 13 slope failures were detected. 4 slope failures occurred behind reservoirs.
Year: 2020