Landslides are one of the most important processes supplying debris flow and bedroad materials into channels. Therefore, we need to predict occurrence of landslides for better estimation of the bed deformation of mountainous rivers. Spatial and temporal changes in the pore water pressure determine timing, location, and volume of landslides. However, most prior studies simply substitute the hydrostatic pressure for the pore water pressure and ignore existence of hydrodynamic pressure. In addition, depth profile of the pore water pressure and the effective stress in the slopes with multi-layer soil structure are poorly discussed. We, therefore, derived the pore water pressure and the effective stress on the basis of the equations for the vertical infiltration process（i.e., continuity equation and Darcy's law）in the multi-layer soil structure. We also discussed their depth profiles by considering boundary conditions. Our study clarified that the water velocity as well as depth profile of the pore water pressure are affected by depth profile of the hydraulic conductivity in the saturated zone. Pore water pressure agrees with the hydrostatic pressure when saturated zone develops on an impermeable soil layer. In case that the hydraulic conductivity is almost constant throughout the saturated zone, pore water pressure is much lower than the hydrostatic pressure. On the other hand, pore water pressure approaches hydrostatic pressure when the hydraulic conductivity in the saturated zone changes significantly in the depth direction. Peak of the pore water pressure locates at bottom of the soil layers in which ratio of the flow velocity to the hydraulic conductivity exceeds -1. When two separate saturated zones develop at different depths, existence of the upper saturated zone does not affect pore water pressure of the lower saturated zone ; upper saturated zone just increases effective shear stress of the lower saturated zone as increases in degree of saturation.

View full abstract

Tree roots with an underground lateral network play a very important role in slope stability. In studies such as those by Tsukamoto（1987）and Abe（1997）, the authors focused on roots that have a vertical connection between the surface soil layer and below, but they did not place much emphasis on the lateral spread of roots. Because slope failures occur in three dimensions, horizontal roots fully demonstrate their resistance against the failure. This paper aims to describe the resistance force Δ C that lateral roots provide. Two types of field investigations were made. One was the root pulling test to monitor the strength of a root pulled by a clamp. The other surveyed the lateral root distribution exposed in vertical sections excavated in the plots of Japanese cypress and cedar forests. The observed root distribution was used to calculate the resistance force Δ C. The field data show that as the distance between two trees increases, the Δ C value gets smaller. Their relationship can be described by a negative power function. It has also been proven that as the age of a tree increases so does the Δ C, but its increment gets smaller in the later stages of the tree's lifecycle. Intensive work with the field data has shown that forest thinning operations might decrease the root volume after felling and bring about an increase of Δ C that lasts for several decades. Moreover, thinning operations produce good results in slope stability when lateral roots are effectively increased. However, if the increase is insufficient, it might not contribute to slope stability.

View full abstract

Mass conservation equations for a debris-flow body and for bed sediment, in addition to an equation of motion, constitute the governing equations for debris flow. The erosion rate equation is generally added to close the governing equation system, However, it is impossible to conduct numerical simulations with just the governing equation system, because points that are highly unsteady locally and are not assumed in the process of developing the existing erosion rate equation occur in numerical simulation. Analysis of the existing erosion rate equation clarifies that there is a high possibility of occurrence of abnormal values at the front of debris flow because the deposition volume calculated using the existing erosion rate equation becomes larger than the existing volume. Therefore, a processing method for the erosion rate equation is necessary. Use of a method that only modifies abnormal values to normal values results in problems related to the conservation law, such as an increase in flow volume or sediment volume. Thus, two methods were examined in this study. One method entails limiting the erosion rate so as to avoid introducing abnormal values and the other entails adjusting the riverbed level corresponding to the extent of modification of abnormal values to normal values. These two methods were evaluated by numerical simulations performed at the transition point of the riverbed gradient. The simulation results clarified that the method of limiting the erosion rate did not obey the conservation law in the case of a large time step value. In comparison, the method of adjusting the riverbed level was able to obey the conservation law almost completely. Moreover, the minimum concentration and equilibrium concentration, taken as target modification values of the sediment concentration, were examined by the numerical simulation. Then, the simulation results indicated that numerical oscillation occurred with the minimum concentration but not with the equilibrium concentration. Therefore, the method of adjusting the riverbed level using the equilibrium concentration as the target modification value was the most effective method under the calculation conditions considered in this study.

View full abstract

On March 12, 2011, the North Nagano Prefecture Earthquake and subsequent aftershocks triggered several sediment-related disasters, including two large-scale landslides that dammed the Higashi-irisawa River, a left tributary of the Nakajo River in Sakae Village, Nagano Prefecture. The volume of the landslides was about two hundred thousand cubic meters and one million cubic meters. In this report, we summarize the characteristics of the earthquakes, landslides and subsequent debris flow, and emergency works following the disasters. Based on boring surveys of the sediment accumulated on the streambed and other field measurements, we discuss the geological features and the hydrologic response in the borehole and the dammed pond.

View full abstract

Tree roots with an underground lateral network play a very important role in slope stability. Since failures occur in three dimensions, lateral roots fully demonstrate their resistance against failure. This paper describes the resistance force that Δ C lateral roots provide. Intensive field investigations and a model description were made. Since the field data show that the root strength Δ C varies considerably among the surveyed plots, predicting the Δ C from the plot information such as tree species, tree ages, slope and soil thickness can be difficult. To assist with prediction, this paper proposes a Δ C model to calculate the resistance force shown by trees of any age when it is given information about the species, forest operation history, and initial tree density. The Δ C model constitutes four components : a forest yield calculation table ; Karizumi's root volume regression curve ; a root distribution model proposed by Tsukamoto ; and a summation process of Δ C. The model highlights the importance of a particular parameter, which is the weight ratio of lateral roots to full roots. The model shows that the ratio of lateral roots directly affects the size of Δ C. By inspecting the field investigation data and the output of the Δ C model, it has been proved that the change of Δ C over time is well represented by a simple logarithmic curve which has only one coefficient multiplied by the logarithmic term itself. In comparing logarithmic curves to the output of the Δ C model, the logarithmic coefficient corresponds well to the weight ratio of lateral roots. The variance of Δ C shown by the field data is lumped into the coefficient. It has the potential to describe lateral root strength in a simple manner and could be a good index of Δ C. Also, when considering a forest thinning operation, the change in lateral root strength might easily be represented by the coefficient.

View full abstract

About 90% or more of the slope disasters occur as surface collapses. As a case study,we report debris disaster coused by heavy rain in Nothern Kanagawa, September 2010. This report proposes Handy type method of Soil slide survey and countermeasure without removing vegetations.

View full abstract