Hikaru Kitahara

The effects of rockfall volume on runout distance

The main goal of this paper is to clarify the effects of rockfall volume on its fluidization. For this purpose, outdoor rockfall experiments were carried out to analyze runout distances and individual movements of rockfall blocks and numerical simulations were conducted for these experiments to learn more about the mechanism of rockfall fluidization. The rockfall experiments were conducted using an artificial slope on which granite slabs were surfaced and overlaid with cubiformed granite blocks. The size and number of blocks were varied in this series of experiments. Further, numerical simulations were carried out in which the coordinates of individual rockfall blocks were traced in three-dimensional space from rockfall initiation to final deposition. It became clear from the experiments and simulations that the runout distance had a positive correlation with the rockfall volume (number of rockfall blocks) and the runout distance of the gravity center of deposited rockfall mass had a negative correlation with the rockfall volume. To clarify the mechanism of these two phenomena, the positions of individual blocks from initial arrangements to final depositions were traced in the experimental and numerical simulations. This revealed that the relative positions of each block along the slope direction were not changed during rockfall movement. The reason for the block-sequence preservation was that front facing blocks in the initial cube arrangement accelerated and rear facing blocks decelerated along the slope by internal collision between blocks. Further, as the rockfall volume increased, the opportunities for impact among the rockfall blocks increased with the front facing blocks pushed farther. Whereas the gravity-center of the deposited rockfall mass had an inclination to travel shorter distances as rockfall blocks increased according to the increase of kinetic energy dispersed by the collision of blocks.

Landslide fluidization process by flume experiments

Abstract An almost real-size slope model was used to study the generation process of landslide fluidization during torrential rain. Experiments were conducted by filling an inclined flume with loose sand and spraying water over the flume with a rainfall simulator to induce the sand to collapse. Both the movement, volumetric strain and the pore water pressure of the sand were monitored throughout the experiments, from the start of spraying to the cessation of the landslide. Our experiments showed the following. (1) Landslide fluidization caused by undrained sudden loading undergoes three stages: compaction of the sand layer by the sliding mass from upper slope, generation of excess pore water pressure in saturated zone, and induction of fast shearing. (2) Fluidization at the collapse source area also undergoes three stages: destruction and compaction of sand layer skeleton by outbreak of shearing, increase of pore water pressure in saturated zone, and shift to fast shearing. But these three stages take place almost simultaneously.

Fluidization in dry landslides

This paper aims at determining the physical properties affecting the distance travelled by landslides with dry particles, and elucidating the mechanism of the properties affecting landslide fluidization. The procedure is as follows: laboratory landslide experiments were conducted to verify the simulation model which was proposed to represent the movement of landslides. Further, sensitivity analysis for some physical properties affecting the travel distance were conducted using this model. The simulation model could identify the coordinates, velocity and angular velocity for every particle in three-dimensional space, and the kinetic energy of particles consumed by inelastic and frictional collision with each other in the model. The travel distances of landslides simulated by the model were verified by laboratory experiments statistically where the physical properties for the particles and slope angle were changed. Then, sensitivity analysis for the physical properties were conducted using the model to clarify the effect of the properties on the travel distance. It was proven that there were no significant differences between the travel distances represented by the simulation model and those found experimentally using statistical analysis (i.e. the simulation model could represent virtual real landslides). From the sensitivity analysis using the simulation model, the travel distances were positively correlated to volume, or number of particles, and negative correlated to the slope angle, kinetic friction and rolling friction of particles when the initial potential energies were the same. The mass of particles of equal size did not affect the distance travelled. As the number of particles increased, the simulated travel distances tended to be longer compared with theoretical travel distances calculated using the friction between particles and the slope. Consequently, we could determine the distance of the lumped mass model as being critical between fluidized and non-fluidized landslides.

Application of Universal Soil Loss Equation (USLE) to mountainous forests in Japan

Universal Soil Loss Equation (USLE), originally developed by the USDA for agricultural lands and then used throughout the world, was applied in mountainous forest terrain in Japan. The slope length and steepness factors were extended for long slope lengths and steep slopes by recalculating the data from available literature in Japan and by using data from supplementary erosion experiments. As such, we conclude that the USLE can be used successfully to estimate surface erosion on long, steep mountainous forest slopes. The cover and management factors of many kinds of natural and artificial forest stands, cutting types, and disturbance regimes were calculated from the literature. The support practice factors of many types of soil and water conservation practices were also calculated from published information. Based on these results, we showed that the USLE can be applied to estimations of soil erosion from wide areas including many kinds of agricultural and forest lands.

Topography and volume effects on travel distance of surface failure

The authors investigated factors affecting the travel distance of a surface failure and developed a simple model for estimating the distance based on soil properties and topographic factor. The authors conducted field surveys and various soil tests for a number of surface failures that occurred during torrential rain at the end of August 1998 in southern Fukushima Prefecture, Japan. The authors studied the effects of various factors such as landslide volume, pore-water pressure, slope inclination, and the internal friction near the slip surfaces on the travel distances. The analyses showed no clear relationship between landslide volume and equivalent coefficient of friction, which was likely attributable to the very small range of volume compared to other studies on catastrophic landslides. The effect of excess pore-water pressure was likely negligible because undrained conditions were not to be maintained at the shallow flow depths. A positive correlation was shown between slope inclination and equivalent coefficient of friction. This correlation was attributable to two factors, one was positive correlation between the internal friction on the slip surface and slope inclination, and the other was the kinetic energy dissipation of moving mass that occurred at the inclination changing point between a slope and a sedimentary flat surface. The authors then developed a predictive model for critical flow inclination of landslide by installing the factors of soil properties and slope inclination factors. The model predicted equivalent coefficient of frictions, which were very similar to the observed values, thus verifying the effectiveness of the model.

Baseflow concentrations of nitrogen and phosphorus in forested headwaters in Japan

A comprehensive investigation on all dissolved nitrogen and phosphorus components at both local and regional scales in the headwaters from forested watersheds is valuable to improve our understanding of the factors controlling water quality. Here, we investigated the baseflow concentrations of dissolved nitrogen and phosphorus components, N:P ratio, and their associations with region and vegetation type in forested headwaters in fives regions of Japan. We found that inorganic nitrogen and phosphorus were the dominant components in the 26 temperate forested streams, rather than organic forms. There were significant positive correlations between the concentrations of N and P components. Furthermore, the regional patterns of the concentrations of nitrate, dissolved inorganic P (DIP), and dissolved total N (DTN) and P (DTP) were similar. Our results suggest that the regional patterns of the concentrations of N and P components should be related to the regional atmospheric deposition of both N and P nutrients. We also found that the nitrate and DTN concentrations were higher in man-made evergreen conifer (EC) than those in the natural deciduous broadleaf (DB). In contrast, the DIP and DTP concentrations in EC were lower than those in DB. The uniformly higher N:P ratio in EC- than in DB-forested streams for each region suggest that EC-forested streams could be more affected by P-limited than DB-forested streams when N inputs from atmospheric sources increased.

Forest soil and litter as filtering media for suspended sediment

A series of two filtration experiments were conducted to evaluate the filtration function of forest soil experimentally. The first experiment evaluated the differences between the filtration capabilities of the A0 horizon and A horizon, and the effect of overstory species on the filtration function of the A0 horizon. Undisturbed A0, A and A0+A horizons were collected for the filter mediums with cylindrical samplers. Leaves ofQuercus serrata, Quercus myrsinaeforia, Sasa senanensis Pinus densiflora, Chamaecyparis obtusa, andCryptomeria japonica were also packed in the samplers. Various suspended sediment concentration of water were sprayed at constant intensity on the surface of samples. Filtering coefficients were not affected by SS concentration in all samples, and the order of filtering coefficient was: A0 horizons>A0+A horizons>A horizons in undisturbed forest soil samples, andQuercus serrata>Sasa senanensis>Quercus myrsinaeforia>Pinus densiflora”Chamaecyparis obtusa>Cryptomeria, japonica in leaf samples. These results led to the conclusion that SS from managed forests can best be prevented by buffer zones where a thick A0 horizon is maintained. The second experiment evaluated the effect of turbid water supply rate on the filtration capacity. Undisturbed A horizons and four leaf types,Quercus serrata, Pinus densiflora, Chamaecyparis obtusa, andCryptomeria japonica were used as filter mediums. Filtering coefficients were inversely proportional to supply rate of turbid water in all samples.

Effect of climate and structure on the progression of wooden check dam decay

To gain better understanding of rates of decay of wooden check dams with different structures under different climate conditions, several dams of this type were examined under different environmental conditions over a 3 to 5-year period post-construction. Because a linear relationship was found between the mean (μ) and standard deviation (σ) of pilodyn penetration depth, mean penetration depth was taken as the indicator of deterioration, and relationships with check dam structural features and climate data recorded by AMEDAS were investigated. Multiple linear regression analysis revealed that temperature, climate index (CI) computed from rain days, warmth index computed from daily and annual mean air temperature, and altitude were the climate variables with the most effect on the rate of decay. With regard to dam structural features, factors such as specific discharge rate, water through width, dam length, and dam height had the most effect. Accordingly, in an effort to summarize the effects of climatic conditions and structural features, CI, altitude, and dam height were extracted as the most significant explanatory variables, and a formula for prediction of μ was obtained for each factor for up to 5 years post construction. The results showed that by taking into consideration regional conditions and calculating CI values from AMEDAS data, it is possible to predict the extent of decay of wooden check dams.