In Okinawa, red soil erosion and its sedimentation in reef areas have caused serious environmental problems. The major sources of red sediment are pineapple fields which consist mainly of cohesive soil. When the surface of a pineapple field is saturated by rainfall, the soil surface swells and forms an over-saturated layer. The authors carried out soil tests and hydrological experiments to evaluate the over-saturated layer strength using a two-dimensional flume and the ground-erosion resistance tester developed by the Public Works Research Institute, Ministry of Construction in 1992.
The following conclusions were drawn from this study.
1) According to artificial rainfall and submersion testing with field red soil obtained on the site, an over-saturated layer with apparent degree of saturation over 100% forms at a depth of about 5mm from the surface of the specimen within 30 minutes.
2) The shear strength of the over-saturated layer falls in inverse proportion to the rise in the moisture content. The shear strength of the over-saturated layer and that of the non over-saturated layer (apparent degree of saturation of 100% or less) are 1kPa (10gf/cm²) and 0.51 kPa (5gf/cm²) respectively.
3) A ground-erosion resistance test showed that when the rotation rate exceeds 10 to 20rpm, the erosion rate of the saturated layer increases proportionally to the rotation rate, and if the viscosity falls, the erosion rate tends to rise.
4) The critical tractive force of the over-saturated layer was calculated as approximately 0.65 Pa. This value corresponds to the critical tractive force of the thin sheet flow produced by the critical rainfall intensity of 10mm/h, and is close to the red soil runoff critical rainfall that has been measured by field observations.

View full abstract

Soil pipes developed in the forest slope function as collector and/or drain of soil water and play an important role in the runoff process. Soil pipes often detected on the collapsed slope suggest that these pipes influence the slope stability. However, this problem has not been examined in detail. In this paper, the influence of the soil pipes on slope stability is examined using the model slopes with a simplified pipe system and artificial rainfall equipment. Studies on the change in runoff components and pore water pressure and the characteristics of soil mass movement are summarized as follows: (1) While pipe system function as a drain system, the pore water pressure is depressed and the slope increases stability; (2) When the pipe system is plugged, the pore water pressure in the vicinity of the plug is elevated and the slope becomes unstable; (3) The position of the deepest slip plane corresponds to the highest elevation in the pore water pressure; (4) When the pipe system is plugged after the pore water pressure become steady, the soil mass moves rapidly and to a distance.

View full abstract

Physical models, which combines simplified hydrological processes and a mechanical analysis of slope stability, have been proposed. Although they can be practical tools for predicting shallow-landslides, model performances were not adequately evaluated in comparison with observed groundwater generation processes. Accuracies and limitations of one of such models were examined by using groundwater levels observed at a forested watershed for 7 years. For the model calculation, whole watershed was divided into rectangular parallelepiped elements. In each time step, water balance in each element and both saturated and unsaturated lateral flow between two adjacent elements were computed. The model assumes a new convex H-S_e relationship, where H is the groundwater level and S_e is the effective saturation, instead of the previously-suggested linear H-S_e relationship. Results showed that the model can regenerate peak time and level of groundwater table observed at a main hollow and middle part of a 0-order valley during heavy storms. Model performances were improved by assuming the convex H-S_e relationship instead of the linear H-S_e relationship. However, the model did not succeed in estimating groundwater dynamics measured at the lowest part of the 0-order valley, and the computed safety factor at the main hollow often suggested occurrences of landslides although no landslide was actually happened. These contradictions were attributable to the difference between the digital topography used for model calculations and the actual topography. Observed groundwater tables around the outlet of the watershed occasionally exhibited long-lasting recession tails, and the model could not explain such dynamics of groundwater.

View full abstract

Debris flows are classified inside the ternary phase diagram whose three apices represent the hundred percent of the total shearing stress within flow is shared by the stresses due to visco-plasticity of the material, brief and enduring particle contact, and mixing of material, respectively. The inertial debris flow, a general term for stony, immature, hybrid and turbulent muddy debris flows, exists close to the opposite side of the apex representing the hundred percent share of the effect of visco-plasticity. A generalized theory of inertial debris flow is developed imagining a hybrid debris flow whose lower part is the densely concentrated particle-colliding layer and the upper part is particle suspension layer. The ratio of these two layers depends on the hydraulic conditions, and in stony debris flow the particle-colliding layer prevails in the entire depth and in turbulent muddy flow the particle suspension layer occupies almost all the depth. The theory is based on the constitutive equations for an inertial granular flow and the turbulent suspension equation. The experimentally obtained characteristics, for example, the solids concentration in flow for a certain bed slope becomes large if comprising particle diameter becomes small, are well explained quantitatively. The theory is modified to give the solids concentration and the resistance to flow in the convenient forms easily applicable to the estimation of debris flow under actual conditions.

View full abstract

The influence of cement in concrete on water quality and fish was studied experimentally. In the summer of 2001, an accident occurred where fish died in a torrent. It was said that the accident was caused by fresh concrete used for sabo structures partly running out into the torrent. That kind of accident has, however, not been reported to date in the field of sabo works. Mortar test pieces were used to study the influence of cement on water and fish in the experiments. The mortar test pieces raised the pH of water up to 11.54 at maximum. When cement was mixed into water directly, the pH of the water went up rapidly even if the quantity of the mixed cement was little, resulting in damage to the Japanese goldfish used in the test. A number of Japanese goldfish died when the pH of water exceeded 10.80. Air entraining agents do not affect the pH of water. Since there is a fear of cement greatly affecting the quality of water and the lives of fish, when concreting sabo structures, it is necessary to take great care to avoid fresh concrete running into the water and increasing its pH to the point of killing fish.

View full abstract

Taking the Ichinosawa catchments of the Ohyakuzure valley located in the uppermost of the Abe River, the formation and dynamic process of debris flows were observed with a VTR system. The VTR system was able to take a 0.75-second-long movie at fixed intervals of 3-5 minutes.
The results are summarized as follows: 1) There is great possibility of forming a debris flow on this debris torrent under condition of the rainfall exceeding 30 mm in total amount and 5 mm in rainfall per 10 minutes. 2) Debris flow consists generally of three phases, the primary flow with containing relative an abundant mad, the main flow, and the following flow with reducing the mud concentration gradually. The main flow appears two types, one is muddy flow with a lot of crushed shales, and the other consists of abundant debris of about 20 cm diameter. 3) The observed main flow had moved down on the debris torrent floor with a succession of deposition and erosion alternately.

View full abstract