Freezing experiments were carried out from a point of view of the mechanisms of origin of polycrystalline snow crystals, namely, spatial dendrites, radiating assemblage of dendrites, combination of bullets and so on. Especially, a detailed examination was made on the number of component crystals and the angle between c-axes of neighbouring crystals when water droplets of 0.4-1.7mm in diameter were frozen under controlled rates of cooling of -12°C/h, -30°C/h and -60°C/h. The results were summarized as follows; (i) Most of the water droplets froze in a range of -17--23°C among the rates of cooling selected. (ii) The number of the component crystals was 2-3 for the cooling rate of -12°C/h and -30°C/h and 5-6 for the cooling rate of -60°C/h. (iii) There were two peaks on the frequency distributions at 20°-0°and 60°-80°for the angle between c-axes of the neighbouring crystals. (iv) Under the condition of annealing, a remarkable peak around 20°-30° diminished and another peak at 60°-80°remained showing an increasing prominent. (v) Examples suspected to be penetration twinning were obtained and the angle between the neighbouring crystals of 13 out of 18 examples was found concentrated at approximately 70°. These experimental results seem to explain the 70° of the angle between spatial branches and bullets of natural polycrystalline snow crystals and further appear to support the calculated results based on twinned structures of snow crystals by Kobayashi et al (1976).
Free-fall pattern and velocity of columnar snow crystals were observed by the stereophotogrammetric method. The experimental formula between falling velocity and mass was obtained for the various types of columnar crystals. The rotation about vertical axis and the oscillation in vertical plane were observed for all of the rimed columnar crystals. The period of oscillation was about one-half that of rotation and the non-dimensional frequency of rotation increased in proportion to Reynolds number. Although the rimed columns fell with vertical path, about 50% of the rimed needles and sheaths fell with clear spiral trajectory.
Ninety nine examples of electric field variations were chosen from the continuously observed data in Kyoto with a field mill during three summer seasons from 1965 to 1967, and were classified into 12 types of field variation patterns. A composite of the field variations which appear frequently has the following representative features: the negative field appears first in time, the time scale is about 1 hour and the field intensity is about ±5000V/m. A thunderstorm charge distribution model which gives the observed field patterns is presented, in which the charges are +120 Coul at the height of 8.5km, -120 Coul at 6.0km and +24 Coul at 3.0km where the negative main charge locates 0.2-0.6km ahead the other charges in the storm moving direction.
A comprehensive calculation of the electric field changes was made for the tilted streamer model in the cloud discharge, in which the streamer proceeds with various angles from the cloud charges. The field changes calculated as a function of distance from the cloud charge to the observer are shown systematically for the positive and negative streamers in the positive dipole. The result shows quite different features from the case of the vertical streamer. The observed electric field changes due to cloud discharge may be interpreted more clearly with this tilted streamer model.
The reaction system and sources of stratospheric chlorine compounds are reviewed. Sensitivity of the reaction system to rate coefficients for the reactions of ClO with O3, CH4 and HO2, and of Cl with HO2 and O3 + M is examined. It is found that these reactions could affect the ClOx content and the ozone destruction rate. Contemporary source strengths of ClOx due to chlorocarbon compounds such as CF2Cl2, CFCl3, CCl4 and CH3Cl are in comparable magnitude with each other. The stratospheric production rate of ClOx from all of these chlorocarbons is in agreement with the HCl density observed in the stratosphere. The ozone destruction by the Cl-ClO cycle operates by about 10% of the total ozone destruction rate at heights 35-45km in 1974-1975.
Effects of uniform flows on a two-dimensional heat island convection were investigated by obtaining the steady solutions of the linearized vorticity and thermodynamic equations. The calculations were performed for fluids whose values of Prandtl number (Pr) are unity and zero. By introducing the suitable scalings, it was found that the effects of the general flow can be expressed by a non-dimensional parameter F' defined by U/(E'1/3l√agΓ), where U is the velocity of the general flow, l is the horizontal scale of the heat island, √agΓ' is the buoyancy frequency and E' is a non-dimensional parameter defined by k/√agΓ l2, where κ is the thermal diffusivity of the fluid. F' corresponds to γ-1/3, where γ is the non-dimensional parameter used in the linear theory of Olfe and Lee (1971). The convective patterns in a fluid with Pr=1 are classified into three types according to the intensity of the general flow. When F' is less than 1, the dynamics of the convection is essentially the same as that without the general flow. The circulation consists of a pair of symmetric convection cells with an updraft over the heat island. When F' is greater than 3, the effects of the stable stratification is reduced compared with those of the horizontal advection due to the general flow and the dynamics becomes essentially the same as that in the neutral fluid layer. The temperature perturbation takes a form of the bent-over plume in this regime. When F' is between 1 and 3, a convective pattern extending to the upstream direction of the heat island appears in the middle layer. The convective patterns in a fluid with Pr=0 are classified in two types; the bent-over plume for the small values of γ and the upstream boundary layer for the large values of γ. The latter structure corresponds well to the patterns observed in the viscous fluid layer for F' between 1 and 3.
The two-dimensional spectra of wind fluctuations in a horizontal plane in the surface boundary layer of the atmosphere are obtained, which are calculated from a simple model founded upon the experimental formulate of the one-dimensional power spectrum and upon the coherence between velocities at two points. The variations of the two-dimensional spectra obtained from the present model with a height and a decay parameter are discussed. The two-dimensional spectra are . obtained as a function of the non-directional wave numbers, and their features in the wave number range corresponding to the inertial subrange are discussed in relation to the decay parameter. Applications of the two-dimensional spectrum to the engineering are briefly mentioned. The one-dimensional spectrum of wind fluctuations spatially averaged over the lateral distance, which is of use for the estimation of the wind load on the line-like rigid structures, is obtained making use of the two-dimensional spectrum.