The viable countermeasures to prevent damages from overtopping erosion of landslide dams depend largely on duration time until overtopping erosion. In this study, we collected and organized resources relevant to the duration times of past landslide dams and defined a clear relationship among basin area, reservoir storage, and duration time. Therefore, the viable countermeasures were organized by basin area and reservoir storage of the dam-forming position based on duration time. Duration time was classified into five stages based on basin area and reservoir storage. According to this classification, it is possible to predict approximately duration time until overtopping erosion for the scale of landslide dam in each point of the basin. Then, it is possible to consider the viable countermeasures of previous landslide dams and necessary advance preparations. However, the duration time of landslide dams caused by earthquakes had a low accuracy of classification compared to by heavy rain regarding consequences to factors other than the basin area and reservoir storage of the dam-forming position.

During sediment disasters, landslide dams rarely form compared to landslides and debris flows. However, when a landslide dam forms and an outburst occurs, it incurs significant damage to downstream residential areas. Therefore, researchers need to estimate the area and degree of damage to develop effective countermeasures. Recent research has revealed that peak discharge at landslide dam outbursts varies depending on the conditions of landslide dam formation, such as constituent materials, their shape and grain size. River conditions such as slope and width also affect peak discharge. It has been reported that multiple landslide dams have formed continuously in one valley stream, for example, Imokawa basin after the Niigata Chuetsu earthquake. However, most of the recent studies and reports have considered only a single landslide dam, and have not taken the water saturation of landslide dam materials into consideration. In this study, we performed a channel experiment considering the conditions of landslide dam formation and riverbed conditions focusing on the number of landslide dams, and observed the difference in peak discharge. The results showed that landslide outburst process was different in single and two landslide dam condition. In different grain size condition, the water saturation of landslide dam affected erosion process and peak discharge. Some cases of two landslide dams showed larger peak discharges compared to cases of one landslide dam. However, it was difficult to indicate the factor controlling the discharge ratio, such as height ratio or amount of submerged water.

In steep mountainous regions, not only soils but also weathered bedrocks were sometimes sliding simultaneously. These landslides often move rapidly and triggered debris flow and sometimes induced landslide dam. In this study, these landslides are referred to “deep-seated rapid (catastrophic) landslide”. This study excludes slow failures of a more chronic nature, such as deep-seated chronic landslide, deep-seated gravitational creep or rock flow, from the deep-seated rapid landslide. To date, there is no adequate information about probability of damage occurrence due to deep-seated rapid landslide. Since there is a variety of sediment movement type, it is important to assess what type of sediment movement may occur. In this study, we compiled existing three databases about deep-seated rapid landslide inventories, which developed by Public Works Research Institute. So, we used about 300 landslide inventories to characterize volume of landslide, sediment movement types and location of damage occurrence. We roughly classified the sediment movement which caused damage into three types, landslide mass, debris flow and landslide dam breach and showed the ratio of occurrence of each sediment movement type. Consequently, we found that, when the landslide volume exceeded 1 million m³, the ratio of disasters due to landslide dam breach was high. In contrast, if the landslide volume was smaller than 1 million m³, the ratio of disaster due to debris flow became large, instead of landslide dam breach. Then, we tested the correlation between deep-seated rapid landslide scale and the extension of damaged area. We found positive correlation between volume of landslide dam lake and the distance from landslide to the lower end of damaged area due to landslide dam breach. Also, we showed positive correlation between landslide area and the distance from landslide to the lower end of damaged area due to debris flow. Finally, we proposed empirical relationship between probability of damage occurrence, landslide scale and distance from landslide.

When landslide dam retaining a large amount of water bursts, it causes floods and catastrophic damage downstream. Therefore, the study of landslide dam deformation and the prediction of the outflow discharge of dam bursts are essential for predicting flood potential and the risks associated with flood. This study focuses on understanding the outflow discharge because of landslide dam deformation by overtopping. The outflow discharge depends on the soil type of the dam materials and discharge of inflow from upstream. However, the influence of different in grain size of sand and temporal changes of inflow discharge for outflow discharge has not been fully understood in existing studies. In this study, we conducted flume experiments to observe the above-mentioned factors on outflow discharge. The effects of these influences on landslide dam deformation and outflow discharge are discussed in this study.

Many large-scale landslides occurred due to heavy rain influenced by climate change and earthquakes. Landslide dams are caused by large-scale slope collapses. Predicting dam failure is important as hazardous flooding may result downstream when landslide dams burst. Accurate prediction of the length and height of a dam, from two-dimensional landslide maps, is necessary as a countermeasure against the disasters they cause. In this study, previous landslide dam failures were investigated, the relationship between dam scale and the topographical characteristics of the region was established, and novel analysis for the accurate prediction of structural integrity was proposed. Finally, we demonstrated prediction accuracy using data from the dam height and length of recent landslide dams.

Mt.Sinabung, which is located in the north side of Sumatra Island, has made the first steam explosion in the historical record in August 2010. After three years of the lull, the volcano became active again in September 2013 and it has continued until now. This report describes the field reconnaissance results conducted in December, 2015. The pyroclastic flow deposits were widely spread from the east to south-southeast slopes of the volcano, and the tongueshaped deposit of lava flow has been accumulated in the center of the pyroclastic flow deposits. The riverbed of Borus River rose remarkably at the downstream. It is thought that the sediment supply of pyroclastic deposit has continued flowing into the Borus River until now. On the other hand, debris flows have occurred frequently in the rivers/stream located in south to south-west slopes, and have caused damages to villages, main roads, and the bridges. It is assumed that debris flows easily occur after the eruption. The residents in the village of the east to south slopes have been already evacuated, and gabions and check dams are currently under construction as the emergency measures. Permanent measures are required to be implemented subsequently.