The snowfall observed on 1st April in 1956 at Tokyo was a very interesting case because of its giant snowflakes. Analyses of the size-distributions of flakes were done using meteorological and radar data. In default of a close network and accurate physical measurements, the analyses were not quite satisfactory. Though other observations concerning size-distributions of precipitation particles were used for the analyses, yet conclusions may retain some unreliabilities. But, as some interesting results which lead to relatively significant conclusions were obtained, the report will be offered to the public in spite of its deficiencies.
Some synop t i c analyses about the snowfall that occurred on 1st April last year were made. The chief cause of the setting in of the snowfall in the Kanto district was concluded to be NE cold-streams on the surface. Cooling of the atmospheric layer by melting of snowflakes may lower the melting layer and accelerate the appearance of snowflakes on the ground. From the distributions of temperatures and phases of precipitation particles over the Kanto district it is found that the melting of the precipitation particles takes place at temperatures from 0°C to 3°C. RHI radar photographs reveal some particular motion of generation-cells of the precipitation. And the generating-cells were seen on the warm frontal surface. A convective cell of the frontal surface was formed in a shape similar to that of dynamical cumulus commonly seen in windy weather. It was found by tracing contours of the echoes that the maximum velocity of the up-current was about 5.5m/s.
Considering the echo intensity on th e RHI scope and the ascending velocities of the echoes a simple comparison of the rates of icecrystal growth was made, and IsoNo et al's experimental result on ice-crystal growth which considers it to occur more rapidly than is inferred from HOUGHTON'Sth eory is supported.
The procedure of observations o f snowflakes and the expected errors were described. For the low concentrations, correct sampling was very difficult in larger sizes. But it is clear that an extreme extension of the breadths of size spectrum resulted in up to 7.3 mm in melted diameter.
The relation between variation of the cut-off sizes of the spectrum in larger ends and the air temperature on the ground was examined. It was found that a large extending of cut-off sizes occurs within a narrow temperature range 0°∼1.3°C. This temperature range will correspond to the height range about 250 m of normal atmosphere. This agrees with the evidence given by many radar researchers that bright bands have the thickness about 300 m.
On an average, the size distribution of snow-flakes in the present observation has a considerably long tail than that of raindrops presented by MARSHALL and PALMER as representative. This is attributed to re-formation of the distributions occurring within the melting layer.
Consideration on the relation between aggregate forms of snow and synoptic weather situations in which snow is generated is made referring to other investigators' works.
Variations of the size spectrum of snowflakes were described using not only the present data but also those obtained in other cases. Generally in snow, it was found that gravitational separation of small and large particles in the snow streaks was hardly observed in every case, contrary to the case of rain streaks. This is attributed to some successive modifications of the spectrum during the fall due to extremely frequent collisions of snowflakes.
On the collision frequencies of snowflakes, simple theoretical estimations were made and the results were compared with these for rain.
The ratios between the collision frequencies of snowflakes and raindrops for the same falling depth, when the precipitation is stationary, were as follows: 423-dry flakes,126-rimed dendrite flakes. From the extreme values, even if fractional collisions may result

View full abstract

The author observed “Diamond Dust ” or ice crystals in the air of a unusual crystal form at Hailar (49°13'N, 119°44'E, 619m height). Some microphotographs of these crystal forms are shown in Figures 1,2,...... and 15.

View full abstract

The accumulated amount of Sr-90 deposition and the infinite gamma ray dose due to the artificial radioactive substance originated from nuclear tests were calculated. The amount of Sr-90 was derived from the values of gross activity of rain water and fallout which have been observed daily since May 1954. The gross activity is reduced to the value at the time one year after fission and Sr-90 content is calculated based on Hunter-Ballou's Tables. The contribution of the older fission products falling slowly from the stratosphere was taken into consideration under a few assumptions in which the values of five years and two months as the residence time of dust particles in respectively the stratosphere and the troposphere were given. Agreement among the calculated values by the present authors, those by US Atomic Energy Commission and the observed monthly deposition of Sr-90 by the chemical method (by IZAWA,1957) was fairly good (Fig.1). The prediction of the probable level of Sr-90 deposit in future was also given (Figs.2 and 3). Calculation of the infinite gamma radiation dose at a given point one meter above the ground was made taking also the stratospheric fallout into computation as above. The result of calculation is given in Table 4 and Figs.4 and 5.

View full abstract

Eight monoplane wind van e combinations with two arm lengths and four vane weights are tested in a wind tunnel concerning their time constants of performed damped oscillation in the tunnel air current. The time constants are investigated with respect to the arm length and moment of inertia, the vane form being kept constant. The tested air current speed ranges from 5 to 20 m/s.

View full abstract

TOMODA'S method for calculating the correlation coefficients R(t)has been applied to temperature analysis in the atmosphere and oceans. The results show that the temperature fluctuations are caused by turbulence m=4/3, diurnal variation m =2, and some factor related to temperature durability m=1, being represented R(t)∼1- (t/T₀)^m,The maximum passage time of turbulence was T₀=30 hr.

View full abstract

[in Japanese]

View full abstract