A case study was made on the snowfall from a cloud layer under subsidence inversion, as the most simple case of snowfall. The observed shapes of snow crystals are shown in Photos. 1-12. Formation of such various snow crystals can be explained from the Ta-S diagram by Nakaya on artificial snow crystals, if the distribution of humidity in the cloud layer is assumed as shown in Fig. 7.
A detailed study was made on the horizontal distribution of snow crystals during the snowfall on January 30, 1959, observed at 14 points in an area of about 5200 km2 in the Ishikari Plain, Hokkaido. The relation between the shapes of observed crystals and the meteorological conditions in the upper air; from which they fell was discussed, comparing it with the Ta-S diagram of artificial crystals. As the result of such study, it was concluded that meridional variation of the shapes of observed crystals could be explained as being determined by the meridional distribution of air temperature within the transitional layer.
The leeward profiles of the thickness of smoke plume emitted from a continuous point source on the ground surface and spreading into diabatic atmospheric surface layer is discussed. The grade of the profile along leeward axis is derived from Lagrangian correlation of vertical fluctuating wind velocity where the grade is considered, namely the root mean square of fluctuating wind velocity at that position, and mean wind velocity. The equation for the root mean square of fluctuating wind velocity under diabatic condition is given in this report from the equation of diabatic wind profile by applying the generalized mixing length theory. Then the theoretical profiles of the smoke plume are compared with experimental and empirical results.
We derived a finite-difference scheme of primitive equations for the staggered coordinate system as well as for the ordinary one. This scheme is supposedly reasonable from the standpoint of compatibility with the system of original differential equations. We also formulated computational boundary conditions with a great care. In order to test the usefulness of this computation scheme, the forecastings of wave motions in a channel with the balance Barotropic model and with the non-balance barotropic model of the atmosphere were attempted. The finite-difference form we used was computationally stable for both models and caused no trouble in 48-hour forecast of the height field which was given mathematically at the initial time. The budget of energy within the forecasting domain could be estimated accurately. Besides, the results of the forecasting agree well with what can be expected theoretically from the model atmosphere. Since the numerical test of the finite-difference form gave fairly satisfactory result in the forecasting of simple pattern, we tried to apply the scheme for the non-balance model to the barotropic forecasting of 500-mb height over the northern hemisphere. There occurred no trouble in this case, either.
Instability properties are compared among multi-level geostrophic models with various number of freedom, in order to obtain some informations concerning the number of levels by which the baroclinic development is well forecast within the hierarchy of quasi-geostrophic models. The growth rate is primarily a function of static stability applied and is not affected very much by the interaction between layers. The effect of horizontal gradient of static tability is also studied numerically and the convex type distribution of the vertical profile of zonal flow is shown to have a destabilizing effect on certain type of disturbances. A multiple parametric resolution is compared with corresponding multiple level resolution.