Establishing chronologies for large-scale landslides is crucial to understand the cause of the mass movements and to take measures against potential hazards in future. We discuss the applicability of dating methods for determining landslide chronologies in relation to the type of samples and the stratigraphic setting of sampling location. Case studies are carried out with fossil wood samples buried in the deposits of large-scale landslides in two areas of the Japanese Alps region in historic times ; Dondokosawa rock avalanche (DRA) and Ohtsukigawa debris avalanche (ODA). Ages are determined by accelerator mass spectrometry radiocarbon dating and dendrochronological analysis using the oxygen isotope composition of tree ring cellulose. We report seven radiocarbon ages and five dendrochronology data for the wood samples taken from outcrops and excavated trenches in the lacustrine sediments of dammed lakes formed by DRA, and two radiocarbon ages and two dendrochronology data for wood samples of ODA. Two sets of data for DRA are crosschecked independently to ensure the accuracy of results. Most of ages in the DRA area are concordant with the period of AD 887 Ninna (Goki-Shichido) mega-earthquake as proposed in previous studies. In the ODA area, ages are not concentrated in a specific period. When the preservation condition of buried wood trunks is good enough to date the exact or approximate tree-death years dendrochronologically, it is possible to estimate landslide occurrence periods in further detail by comparing the landslide chronology with historical records of heavy rainfall and large earthquakes.

It is indispensable to have scientific estimates of the runoff sediment volume and the peak discharge of debris flow in streams when considering both active and passive measures against debris flows. The purpose of this study is to propose an estimation method for the design runoff sediment volume from debris flow and the peak discharge of debris flow by means of the probabilistic safety assessment based on about 450 data on debris flow throughout Japan listed in a literature. As a result on statistical analysis, there were relatively strong correlations both between the runoff sediment volume and the drainage basin area and between the peak discharge and the drainage basin area, but on the other hand, there were no significant correlation between both of the runoff sediment volume and the peak discharge and both precipitation and geology. An estimation method for the design runoff sediment volume and the design peak discharge via the drainage basin area by the 95% or 99% prediction interval of the data obtained by using the past survey results was proposed.

In the event of large-scale sediment movement, it may be possible to detect seismic waves using a high-sensitivity seismic network. Use of seismic data to detect large-scale sediment movement is expected to become a standard disaster monitoring method. This paper systematically presents the results of recent studies on the relationship between large-scale sediment movement and the characteristics of seismic waves detected by high-sensitivity seismic networks. We discuss (1) the monitoring position for detection of large-scale sediment movement, (2) the frequency characteristics of the seismic wave at the time of the sediment movement, (3) the relationship between the type of sediment movement and seismic characteristics, (4) the relationship between underground structures and the likelihood of detection of seismic waves, (5) the relationship between sediment movement and the timing of the increase in wave amplitude, (6) the relationship between the displacement waveform of the extremely low-frequency wave (0.01-0.1 Hz) and sediment movement, and (7) methods for detecting the hypocenter.

When conduits under roads were blocked by cobbles discharged from mountain area during heavy rainfall, flood started to flow on the roads and damage them such as the disasters in Fukuoka and Oita Prefectures in 2017. In such situations, it would be very difficult for residents to evacuate to their shelters safely. However, the process and mechanism of sedimentation in a conduit is still unclear, so that we aim to make clear the process and mechanism of sedimentation in a conduit by flume experiments and three-dimensional numerical simulations. We used beads instead of cobbles in our experiments. We found followings ; (1) the beads suddenly became slow after passing the hydraulic jump, (2) the following beads collided the slowed beads and they started to deposit, and (3) the direction of forces acting on beads such as pressure and drag force was to upstream, so that the beads became slow and stopped.

Seismic sensor networks detect ground vibrations caused by the movement of large masses, predominantly at low frequencies of 1-3 Hz. We modeled sediment flow down a slope and measured ground vibrations in cases with and without a concrete wall at the end of the slope. The sediment weight, amount of water, and presence of iron balls were varied, and the waveform and amplitude spectrum were analyzed. As a basic, small-scale experiment, physical dissimilarities were not adjusted for. Nevertheless, the ground vibration amplitude and low-frequency waves mimicked real-world seismic vibrations.