Prediction of ground water movement is very important and critical to forecasting slope failure because many slope failures are triggered by the surcharge of water weight and drop of suction when precipitation infiltrates into soils. It is necessary to consider both saturated and unsaturated flow phases, particularly for sequentially estimating ground water flow to specify the time and location of the failure. Richards' equation has strong nonlinearity of coefficients. As a result, its numerical analysis tends to be much more difficult than that under moderate conditions, especially with very dry soils or a perched water table. Recently, new methods to solve Richards' equation that focused on diminishing water balance errors and less limitation of time step size were presented (Celia 1990, Pan 1995). In this paper, those new methods are first introduced with explanation of their algorithms specified for balance errors or time step size. Second, based on the fact that both steady and unsteady phases are combined in Richards' equation, the consequent oscillation problem is demonstrated and examined. The algorithm of the oscillation is explained with the analytic method presented in this paper.
The influence of soil water fluctuations on soil reinforcement by roots was examined with a model based on the de-formation of a pile inserted into soft subsoil. This model was applied to the results of a direct shear test using two materials that simulate a root system. One was a bamboo stick (0.27cm in diameter 10cm long), the other was a ny-lon net (0.05cm thick, 10cm wide, 15cm long). As soil sample, Toyora standard sand was used. The range of suc-tion in this test was from 0 to 50cm H2O. The calculation results from this model corresponded well with the increase of shear stress by the direct shear test. The difference in the soil reinforcement between two materials was described in the relationship between the horizontal displacement and the shear stress increase. This difference was explained by the interaction between the subsoil deformation and the difference of the moment of inertia of the area between two materials in this model. This model has a parameter related to the elastic modulus of soil. The relationship between the parameter and the soil suction, the normal stress, explained the influence of suction on the shear strength of soil with a root system.
A deep-seated landslide occurred at Izumi City in northwestern Kagoshima Prefecture on July 10, 1997. Debris flows which followed the landslide killed 21 persons. The purpose of this study is to examine the hydrogeomorphological characteristics of the landslide in the Harihara River basin on the western flank of the old volcano of Mt. Yahazu-lake. We measured the base flow at the observational stations of 14 streams around Mt. Yahazu-lake. The considerable amount of base runoff of the Harihara River originates from the groundwater runoff at the foot of the hillslope where the landslide occurred. The amount of base runoff depends on the spatial distribution of gentle slope and deeply weathered rocks, and the degree of dissection. The stream discharge is an important factor in predicting the potential deep-seated landslide sites.
The 1927 Kita-Tango Earthquake triggered thousands number of slope failures in the Tango Peninsula, Kyoto Prefecture. The area and number of slope failures caused by the earthquake had been investigated and reported by the Kyoto Prefectural Government in 1930. The rate of slope-failure area at each of 59 municipalities in the Kita-Tango Region was obtained based on the report. Analyses were made on the relations between the rate of slope-failure area, geological and morphological features susceptible to slope failures during seismic shaking. Results derived from these analyses can be summarized as follows. There is a strong relationship between the rate of slope-failure aea and the distance from the earthquake faults. It is apparent that the average rate of slope-failure area is less than 1.0% when the distance from the earthquake faults ex-ceeds 10km, and almost 0% when the distance from the earthquake fault reaches about 15 km. Although the rate of slope-failure area decreases as the distance from the epicenter increases, the reationship is weeker than those of the disatnce from the earthquake faults. There are no apparent relationships between the rate of slope-failure aea, the morphological features and the relief energy. There is an apparent relationship between the rate of destroyed houses and the distance from the earthquake faults as like as the rate of slope-failure area. These results suggest that the distance from the earthquake faluts is available for prediction of the rate of slope-failure area.
Slit sabo dams have been employed more often to date because they allow sediment, fish, animals etc. to pass through or over them freely during usual times. The sediment control of these slit sabo dams has been studied through hydraulic model tests in conjunction with computer simulation techniques. Their effects, however, have not been verified through field tests. A slit sabo dam built in the Baba-dani of the Kurobe River experienced a huge amount of sediment inflow in July, 1995. Changes in river bed elevation during the storm and photos were analyzed to estimate how well the slit dam worked. It was confirmed that the slit sabo dam checks the sediment discharge well when the water level rises high and releases it when the water level drops. It was also found that a sabo dam located down-stream of the slit sabo dam affects the sediment deposition of the slit sabo dam.
Introduced herein is a type of debris flow that hit a torrent equipped with one closed type and one open type sabo dam. Those two sabo dams could not check the debris flow as was expected. As a result, certain types of sediment flowed over them and damaged a number of houses standing in the downstream alluvial fan. We gained some knowledge from the analysis of the event. The closed sabo dam must be kept empty through excavation to be capable of holding the utmost sediment deposit. The opening of a grid type open dam must be made narrow at its upper part because the grain size of the sediment of the subsequent flow is so small that the sediment cannot be trapped if the upper section of the opening is wide. When designing the debris flow control structures, it is necessary to consider that debris flow starts not only at the uppermost stream, but also at various slopes.