A landslide detection based on the change of scattering power components between multi-temporal polsar data

There is currently a great deal of interest in the use of radar polarimetry for disaster monitoring. In this context, this paper presents a result of numerical assessment of landslide detection methodology based upon the change of scattering power between multi-temporal fully polarimetric synthetic aperture radar (POLSAR) data. There are several types of natural disaster. Among them, we tried to detect sediment disaster including landslide, slope failure, debris flow, from POLSAR data. Sediment disasters are those that occur on mountain slopes. Mountain slopes in warm and humid regions such as Japan are usually covered by vegetation. Sediment disasters are triggered by heavy rainfall or an earthquake. Soil and vegetation flow down the slopes. Therefore the original slope changes to a slope with bare soil. The scattering process of the microwave comes to the surface scattering from the volume scattering. Therefore, we select a mountainous area stricken by a large earthquake. The four-component scattering model decomposition was applied to the POLSAR data over the area. The data acquisitions were conducted before and after the earthquake. Compared sediment disasters interpreted from aerial photographs with the scattering power decomposition images; it is revealed that the surface scattering increased after the event as compared to the volume scattering as expected.