We proposed a distributed model to simulate occurrence of snowmelt type volcanic mudflow in upstream hillslope areas. The model included processes of heat transfer between snow and erupted sediment layers, snowmelt, vertical infiltration of snowmelt water in the snow layer, slope failures at the snow-soil and soil-bedrock boundaries, erosion due to surface runoff including sediment (i.e., mudflow) and deposition by the surface runoff with excess sediment concentrations. Mudflow discharge from the upstream hillslope area was employed as an input in simulations of two-dimensional unsteady flow in the downstream area. Simulations were conducted for a steep catchment located at northwest of the summit of Mt. Yake, which is an active volcano, and its downstream area, Takahara River. Simulation results under uniform snow thickness and actual distributions of snow thickness in the hillslope area suggested that the spatial distribution contributed to differences in runoff timing of mudflows from the sub-catchments of the hillslope area. Snow density generally varies even within a winter season. Our simulation also revealed that relationship between snow density and mudflow volume was not always linear and showed a convex shape. The simulated mudflow discharge from the hillslope area and flow depths and sediment thicknesses at residential areas in the downstream area were maximum when the snow density was 330 kg/m³, which can be assumed snow conditions in mid-January.

Urgent works, which should be done just after large landslide dam formation, are very important to prevent landslide dam brake, therefore we have to conduct and complete the urgent works in short time although facing many difficulties. But unfortunately we had not haven effective method about like this urgent works against large landslide dam until the terrible disasters, typhoon Talas, Sep. 2011, in which the large landslide dams were caused in Kii peninsula, Nara and Wakayama prefecture, so that it was very hard to get progress and manage the works, and we had to have many trials to complete the urgent works. The review of these trials is necessary to establish the effective way of urgent works against large landslide dam. Therefore I reviewed these works and what methods should be applied to urgent works against large landslide dam. As results, to construct temporary channel effectively, and to get the method to reduce water level of landslide dam lake are important.

We verified the detection capability of landslides using SAR images taken before and after the heavy rainfall in northern Kyushu in July, 2017. We have successfully detected about 70% of the landslides that occurred on the slopes facing away from the radar. The landslide detection accuracy tends to be higher where the local incidence angle is large. If these conditions are met, we are able to detect a relatively small landslide with the area of about 2,500 m².This is a quarter area of a landslide that was detected in previous studies. By using multi-temporal SAR images, the changes on the ground surface conditions can be detected with 90% accuracy.

In this report, we introduce a new method for automatically detecting flow-regime changes using unit vectors and the normal unit vector of RGB values from camera imagery. Although the method is still under development, it is expected to lead to further enhancement of the monitoring system for sediment disasters. From the results of a casestudy at the Inari river in Nikko, Japan, it is suggested that the RGB unit vectors may have eigenvalues according to different types of water color : e. g., clear water, turbid water, and debris flow color. It is also suggested that the RGB unit vectors can draw a curve among the unit vectors of pre and post-turbidity colors. Furthermore, even if the amount of solar radiation changes, the RGB unit vectors can indicate almost stable values.