Since a seismic wave associated with a snow avalanche may contain considerable information about the avalanche itself, seismic observation using 4 seismographs was carried out at Toikanbetsu (N45°, E142°), Northern Hokkaido, Japan during 80 days from January to April 2001. This paper first discusses the characteristics of the sesmic signals. Second techniques to estimate the avalanche release point and its scale are developed. As a result of this observation, totally 50 seismic waves from full-depth avalanches were obtained. They correspond to 86% of all observed avalanches, which indicates that monitoring of avalanches with seismographs is useful. The shapes of waves from avalanches released at the same point are similar to each other. From frequency analyses of the seismic signals, a negative correlation is found between the dominant frequencies of the seismic signals and the distance from the avalanche release point to the location of the seismograph. On the other hand, the positive correlation is found between the total energy of the seismic wave and the avalanche mass. The results indicate that the avalanche release point and avalanche mass can be estimated.

View full abstract

When constructing new buildings, bridges and other facilities in snowy areas, it is necessary to define the maximum snow load at the site beforehand. In this paper, to estimate maximum snow load on the ground from maximum snow depth, equivalent snow density in extremely heavy snow areas is conducted. The equivalent snow density is logically calculated from the maximum snow load divided by the maximum snow depth at a site in a winter. About ten-year data of the maximum snow depth and maximum snow load were obtained from the snow observation network in mountainous areas, which was installed by National Research Institute for Earth Science and Disaster Prevention. The equivalent snow density conducted by this study is greater than that calculated by three previous guidelines, when the maximum snow depth exceeds 3 m in the Honshu Island.

View full abstract

This paper describes a method for calculating the time variation of snow depth and depth-density profiles on “Excel” software by using precipitation data, based on a viscous compression model (Table 1) . This method is applied in a period when the surface snow hardly melts. For improvement of accuracy in calculation, equations showing the viscosity-density relation are used with ρ≤200kg/m³ and ρ>200kg/m³, respectively. The results calculated from daily precipitation data at Tohkamachi, Niigata in 1983-84 coincided with the measured snow depth and density profiles (Figs: 5 and 6).

View full abstract

In order to survey snow water equivalent over vast areas and to study the internal structure of the snow cover, a ground-penetrating radar (GPR), which has been popular in engineering applications, was used. First, we carried out calibration experiments in an open space to determine the noise level of the system (800-MHz GPR) . Second, we made observations of winter snow-cover in Hokkaido and Murodou-daira, Toyama prefecture. As a result, it was found that the GPR is capable of detecting numerous thin ice layers within the snow-cover. Finally, based on analyses of snow survey data in Hokkaido and GPR measurements, simple equations were derived for various types of snow in Hokkaido, which are able to estimate snow water equivalent only from two-way travel time (delay time) by GPR surveys.

View full abstract

In Hokuriku, horizontal distribution of snowfall intensity is heterogeneous and the snow compactive viscosity coefficient significantly changes over time because the temperature fluctuates near zero degrees Celsius in the snowy season. Comparison between observed snow depth and hourly calculated snow depth using a multi-layer densification model of the snow layer with data from meteorological observations is performed to obtain the horizontal distribution of some characteristics related to the snowfall and the snow densification. It is found that the temporal change of the snow compactive viscosity coefficient is important to estimate the correct snow depth in Tohkamachi; 18% overestimation of the seasonally averaged snow depth results from failure to consider the temporal dependency of the snow compactive viscosity coefficient related to the moistening and the granulating of the snow layers. The following results are also found from comparison between the observed and calculated snow depths at the meteorological stations: 1) hoaring or sintering often occurs in mountainous regions of Nagano and Fukushima Prefectures; 2) the snow depth cannot be correctly estimated without humidity data at Fukui Prefecture and the southern part of Niigata Prefecture; 3) knowledge of the correct compactive viscosity coefficient of the snow layer is important for estimating the snow depth in other areas.

View full abstract

A field survey and numerical simulation were carried out to study avalanches which occurred at Kamikouchi-Norikura Super-rindo on 5 January, 2003; they occurred in the forest and more than 20 cars were involved.
A weak layer, which caused these avalanches, was observed 30 cm below the snow surface and was composed of solid type depth hoar. It is conjectured that this weak layer was formed under the following condition; fresh snow deposited between midnight of 1 and the morning of 2 January changed into a solid type depth hoar layer under the weak wind and cold condition on the morning of 3 January.
A simulation using a numerical snowpack model with meteorological data also revealed a solid type depth hoar layer and demonstrated the availability of the model to forecast avalanches.
Although forests are considered to contribute to avalanche protection in general, this avalanche implied that we need to evaluate their effectiveness carefully.

View full abstract

[in Japanese]

View full abstract