Rainstorm control of the scale and the type of debris flow was analyzed using the 14 sets of rainstorm and debris-flow data obtained from 1980 through 2005 at the Kamikamihorizawa Creek of Mount Yakedake. The principal component analysis on the parameters of debris flow : its frontal velocity, peak velocity, peak flow depth, peak discharge and total discharge, combined with the correlation analysis among these parameters and the pre-event rainfall intensities indicates that these 14 debris flows should be divided into three groups. The flow in the first group shows the higher velocity and the larger total-discharge, and have a massive and turbulent boulder-dam filled with muddy matrix, while the flow in the second group shows the lower velocity and the smaller total-discharge, and a massive boulder-dam scarcely filled with muddy matrix, and an alternation of halt and re-start of the dam, and the flows in the third group take the lower velocity and the larger total-discharge. The analyses assert that the combination of the rainfall in longer time duration such as 24 hours and the rainfall in shorter time duration such as 10 minutes before the events controls this debris-flow grouping, namely the larger rainfall in longer time span generates flows in the 1 st group and the smaller rainfall in longer time span with the higher rainfall in the shorter time span generate ones in the 2 nd. Ones in the third group are generated by the larger rainfall in longer time span with the smaller rainfall in the shorter time span. The characterized relationship between rainstorm condition and debris-flow parameters might be applicable not only to the similar volcanic slopes but also to the torrents in other geologic regions, and be significant both to mitigation of debris-flow hazards and control of fluvial sediments.
Upon an outbreak of a natural disaster, specially under the circumstances where information of weather, sediment disaster as well as premonition of further disaster frequently changes minute by minute, it is envisaged that decisions on evacuation and evacuation preparation may be made and most probably be influenced by the psychological factors. Therefore, this study is intended to construct a Model for evacuation behaviors upon disasters from psychobehavioral aspects. The suggested model illustrates the evacuation process by four stages, namely 1) confusion, 2) perception, 3) judgment, and 4) action. A certain score is set forth for one perception and the scores shall be accumulated as time passes. Thus, it is assumed if such accumulated scores cross the threshold level, a certain action will be taken. A score given to each influencing parameter differs between disaster and normal circumstances. Scores shall be given to the parameter by chronological analysis under disaster while they shall be added to the parameter under the normal circumstance. The influencing parameters were set forth following the results of earlier studies and the threshold level was considered and decided in conformity with the results of public survey. A public survey was conducted in the form of a questionnaire at three municipalities in Nagano Prefecture where Typhoon No.4(man-yi) & No.9(Fitow) hit devastatingly in 2007 to ascertain applicability of the model that has two threshold levels of evacuation activities and evacuation preparation activities. As the result of analysis, applicability of the model has been improved by considering the regional differences as underlying potentials. Furthermore, the threshold level that covers all the refugees has been established by optimizing scores by quantification methodⅡ. In addition, it is confirmed that direct evacuation alert announcement through public address system of the community is much more effective than warnings of sediment disaster broadcast by TV or Radio.
There are direct method and indirect method in bedload measurement. It is possible to surely measure in direct method. But the work of this method is the labor, and the measurement is not possible in the long-term. In the other, the work by indirect method with sensor is easy, and long-term measurement is possible. Therefore, the expectation which is caught in indirect method is big. But there is a study case of this method only on hydrophone. There seems to be some the necessity of investigating the usefulness of the sensor by trying bedload measurement using other sensor. And then, we tried bedload measurement using impulse radar system. In this study, the experiment which confirmed the relation of concentration of sediment which flows in flume and signal strength was carried out. It was proven that there was the clear and positive correlation from the experimental result to both. Therefore, there was the possibility of bedload measurement using radar.
We quantified the effect of tree root systems on slope stability in a natural cool-temperate deciduous broad-leaved forest. We established three study plots on a mountain slope in Tomamae Town, Hokkaido, and measured tree diameter and position and identified the trees to species. We dug out the root systems and measured the root distribution for three to four trees per plot. We measured the root diameter and the force required to pull the roots from the soil after removing the stumps for two to four trees per plot. We measured cohesion, angle of internal friction, and bulk density of soil. We estimated the shear plane using core boring. We estimated the pullout resistance of the roots on the shear plane for all trees in each plot using root distribution data, the relationship between root diameter and the pullout resistance of the root, and the ratio between the diameter of each tree and that of dug out trees. We calculated the safety factor using a slice method (the Janbu Method). We added the pullout resistance of the roots to the soil cohesion in the formula used to calculate the safety factor. The pullout resistance of the roots was 2.29, 1.16, and 0.50 kN/m2 in the three plots. The safety factor increased by 0.28, 0.1, and 0.04 in the respective plots by adding the pullout resistance of the roots to the formula for the safety factor.
This study aims to determine empirical rainfall intensity-duration thresholds for landslides and debris flow and to find standard rainfall warning for debris flow overflows in Mt. Bawakaraeng Caldera, Indonesia. The debris flow overflow happens when excess sediment (debris) flows over the top of channel works and check dams. The study is based on the hourly-rainfall data recorded from 1997 to 2007. The threshold, as defined by the lower boundary of the points representing landslides and debris-triggering rainfall events, is expressed as I=41.85 D-0.85 before the large scale landslide on March 26, 2004 and I=37.71D-0.90after the large scale landslide, where I is the rainfall intensity (mm/hr) and D is the duration of rainfall (hr). The Debris flow overflow warning in the study area may be issued when rainfall is expected to exceed the warning line.
Generally, debris flow occurs by heavy rainfall or increment of surface water. But the occurrence condition of debris flow may be influenced by conditions of stream bed deposit, such as shape, gradient and its deformation. In past studies, through the hydraulic model experiment for analyzing debris flow occurrence, experimental flume is steep and equal gradient, huge amount of water must be supplied, and stream bed deposit has been saturated. But in natural torrent, riverbed gradient changes gentle toward downstream, stream bed deposits are not uniform shape, and also deposit sometimes unsaturated. We observed and analyzed the condition and deformation of stream bed deposit at initiation zone of debris flow through the hydraulic model experiment. The experiment was carried out on a experimental flume whose bed slope is 30° at the upstream end and 12° at the downstream end, supplying a small amount of water nearly infiltration flow at first and rises as surface water level during rainfall. In the steep section as 27°-30° with small amount of surface water, a moving layer has been formed and flow down gradually. This moving layer starts to move when the layer is saturated. Gradient turns small as 21°-24°, moving layer start to deposit and form a dune. Upper part of the dune has no surface water but lower part has it and flow down. This dune starts to move or deforms when surface water reaches and flow over the edge of the dune.
Bili-bili Dam located in Sulawesi Selatan, Indonesia was built as a multipurpose dam. However, it has lost its sediment capacity by deposit resulted from the collapse of Mt. Bawakaraeng. Since then, there has been a criticism that the construction of Bili-bili Dam was not appropriate at the very place. Without Bili-bili Dam, sediment in the river by the collapse would have possibly triggered serious sediment disasters in downstream areas. We evaluate effectiveness of Bili -bili Dam as a sabo dam which prevents sediment from flowing downstream. We simulate one-dimensional riverbed variation of the Jeneberang River and two-dimensional flood in the downstream areas of Bili-bili Dam. The results of one-dimensional riverbed variation simulation show that 22 million m3 of sediment would flow downstream in twenty years and the riverbed would rise up to 10 m. The results of two-dimensional flood simulation show that the floodwater would inundate 68 km2 of the areas including highly populated residential areas. Therefore, Bili-bili Dam functions effectively as a sabo dam.
Typhoon Morakot is the deadliest typhoon for Taiwan over the past 50 years. In Taiwan, more than 600 people were killed during the period of Typhoon Morakot. Different kinds of disasters which were resulted from Typhoon Morakot occurred simultaneously. These disasters, including flood, landslides and landslide dams, are briefly explained in this paper. Simultaneous occurrences of these disasters resulted in the composite hazards. The devastation of Siaolin village is an example of the composite hazards and is described in this paper. The rainfall, which is regarded as the major cause of disasters, is compared to that of six typhoons, including Typhoon Isewan struck Japan in 1959. From the comparison, the rainfall of Typhoon Morakot is found to be high-intensity and long-duration. The extent influenced by the rainfall is large. The landslides and landslide dams in Taiwan, which are recognized using the satellite imagery, are overlapped with the isohyet. It is found that most of the landslides and the landslide dams appeared in which the values of accumulated rainfall depth are larger than 800 mm. It is concluded that the present system is not sufficient to handle such composite hazards. It is necessary to develop a comprehensive system for decision makers to handle the composite hazards.