While the role of gravitational, thermodynamic and hydrodynamic forces in thunderstorms has been extensively studied, that of electrical processes is still little understood. Measurements of the electric fields beneath storms were conducted during the summers of 1968 to 1970. Electrostatic voltmeters (field mills) were constructed and used in Alberta both in fixed station operation and in mobile operation. Aided with information from radar, hailfall and rainfall, case studies of three storms were performed. The analysis of the results suggested that the dynamical features of storms play a more important part than the microphysical ones in affecting the electrical behaviour of storms, once the conditions necessary for the charging mechanism to function are satisfied. In particular, for storms penetrating into the stratosphere, the organization of the storms becomes extremely important and the amount of lightning manifested by a storm complex is directly related to the number of cells within the complex. Two conclusions on the properties of the charging mechanism were also realized, viz. (i) the efficacy of the charging mechanism is not increased by increasing the updraft speed, and (ii) the charging mechanism should produce lightning of frequency 1-3 flashes per minute for each storm cell, irrespective of whether the storm is a thunderstorm, hailstorm or severe storm.
Microphotographs of rimed snow crystals were analyzed and the riming properties were studied. The onset size of riming on snow crystals and the cutoff diameter of droplets accreted on snow crystals were obtained. It was apparent that when the snow crystals were larger, the riming in the same type of snow crystals was more favorable. It is considered that this is not related to their collection efficiency, but originates from their falling distance. It was shown that it may be possible to estimate the droplet size distribution in cloud, if rimed snow crystals of various types and various sizes are observed simultaneously.
The electrification of freely falling water droplets due to freezing was measured in a laboratory experiment. When distilled water was used for specimen, the frequencies of positive and negative electrification were nearly the same, while the positive electrification was predominant in case of water melted from fresh natural snow. It was concluded that the electrification of natural ice pellets is occurred by the ejection of charged particles of a few micron diameter at the end of freezing state. It was also pointed out that because the impurity of specimen water is very important for the freezing electrification, the experiment by the use of natural water in clouds is desirable, if we undertake to clarify the mechanism of charge generation in natural clouds.
Freezing experiments of freely falling water drops of diameter 45-765 μm were made by usinga large cloud chamber 14.7 meters in height. Frequent shattering of freezing water drops is observed not only when water drops are nucleated by silver iodide in suspension but also when they are nucleated by collision with ice crystals after attaining thermal equilibrium with the environment. The frequency of shattering depends considerably on the nucleation temperature (Tn) and the drop size (d). The frequency of shattering is especially high in the size range d=75-135 μm when water drops are nucleated at temperatures between -14 and -20°C. Spikes are formed when water drops are nucleated at temperatures higher than the air temperature (Ta) and are scarcely formed when Tn_??_Ta. In the case of Ta_??_-27°C, the frequency of spike formation increases with the rise of the nucleation temperature and the increase of the drop size.
To clarify the behavior of the standard deviation of vertical velocity σw under various meteorological conditions, many data obtained by tethered balloon, high tower and light airplane were analized. Relations between σw, wind speed U, and many meteorological elements, for example, insolation, local lapse rate, Pasquill's stability categories, classification of cloud and inversion base height, were checked. It has been found from observations at height of 50 m that σw was proportional to U in case of weak insolation while with increasing of insolation, σw tended to show characteristics caused by free convective fluctuations. Temperature gradient at the observation height of σw can not explain variation of the relation between σw and U with stability. Pasquill's stability categories are not explanatory index either for the σw-U relations, except at 50m. Three different relations between σw and U were found from observations taken at various heights. First is the relation for strong convective layer; second, in the layer below inversion base; and third, in the inversion layer. In the case of second group, there is a critical wind speed Ucγ. When U is weaker than Ucγ, free convection is redominant and σw is independent of wind speed. When U is stronger than Ucγ, however, forced convection is predominant and σw/U tends to be constant. The height dependency of Ucγ was observed.
Formation of thin turbulent layers in a stably stratified fluid is discussed in combination with a critical level and internal gravity waves. A series of numerical simulations by using nonlinear model of an incompressible fluid reveal that very thin unstable layers (local Richardson number <0.25) are formed in the vicinity of the critical level. Mini-K-H billows can be expected to grow up and break into turbulence in the thin unstable layers. Long-time numerical calculations show that such thin unstable layers which contain the mini-K-H billows and turbulence roll up at their edges and form Cat's eye patterns associated with large amplitude K-H billows deformed by the critical level. This result also seems to open a new perspective to the critical level theory itself. Some observations in the planetary boundary layer provide good evidences to the results of the present numerical simulations.
This paper concerns with properties of a thermal convection in a stably stratified Boussinesq fluid caused by partial heating or cooling of the lower boundary. For infinitesimal heating (cooling) the convective motion can be described by a linear theory. Introducing a suitable scaling, it is shown that the convection is controlled mainly by a non-dimensional parameter, R=αgΓl4/κν where Gamma; is the vertical temperature gradient in the basic state, l is the half-width of the heated (cooled) area and the other symbols have conventional meanings. The convective motion is confined to the "frictional depth" introduced by Stommel and Veronis (1957). For a single slab-symmetric heating (cooling) the thickness (the lowest height at which the temperature perturbation vanishes) is given by hT=3.6 R-l/6 and the maximum horizontal velocity by u*max=0.25, √αg/PγGamma;|Th| where Pγ=ν/κ and Th is the temperature difference between the heated (cooled) area and the surrounding area. The width of the upward (downward) motion area is close to the horizontal scale of the forcing area. The convection patterns for the finite-amplitude heating (cooling) were investigated by means of laboratory and numerical experiments for |Th|_??__??_l. It was observed that increase of the thickness of the boundary layer with Th is small. As for the convection caused by heating, however, the width of the upward motion area decreases with the increase of Th until the magnitude of the upward motion becomes comparable with that of the horizontal motion. As for the convection caused by cooling, the width of the downward motion area increases with |Th|, but the essential features of the convection pattern are similar to those for the infinitesimal cooling. These results are compared with available data of the convection due to the urban heat island effect.
The intensity and the degree of polarization of the radiation diffusely reflected by turbid atmospheres are calculated in the visible region of wavelengths, 0.36, 0.40, 0.50 and 0.60μm. The underlying surface is assumed to reflect light in accordance with the Lambert law. Deirmendjian modified Gamma function (1969) and Junge power law (1963) are adopted for their aerosol size distributions. Their refractive indices are 1.34 and 1.50 in the real part, while the imaginary part is assumed to be zero. The values of the optical thicknesses of aerosols are obtained from the work of Rao et al, 1973. The objective of this work is to estimate the optical properties of the aerosols such as size distribution and refractive index by comparing the remote sensing satellite measurements with computed values of the intensity and the degree of polarization of radiation emerging from the top of the model atmospheres. The doubling method (Hansen, 1971) and the adding method (Takashima, 1973) were used for the numerical computations. Computational results show that a strong dependence of the phase matrix of aerosols exists on the diffuse reflection radiation especially at the anti-solar point, showing that (1) the intensity and degree of polarization is a smooth function in the sun vertical if m=1 .34, and if m=1.50, (2) a large intensity value of radiation is observed due to a strong back scattering by the aerosols, and (3) the characteristic polarization curve exists at the anti-solar point. Towards the end of this paper, brief discussions are made for the assumptions employed.
The intensity and the degree of polarization of the radiation diffusely reflected by an inhomogeneous atmosphere are computed by using the adding method (Takashima et al., 1975), where the atmosphere is composed of twelve homogeneous composite layers. The atmosphere-ground system used for the numerical computations is such a realistic model that the effect of scattering due to aerosols distribution with height and the absorption effect by ozone are taken into account. In addition, the atmosphere is bounded by a hybrid surface of a diffuse and specular reflector. In this paper, computations are selected in the wavelength region of 0.65μm which is applicable for the study of VHRR and VISSR data (0.6-1.0μm). Elterman's values (1968) are adopted for the optical thicknesses of aerosol, Rayleigh and ozone constituents. Deirmendjian model L (1969) is used for the aerosol size distribution with a real index of refraction of 1.34. Computational results show that (1) at a low sun elevation, a high intensity value is noted in the forward direction. This is more pronounced in the case of a specular surface than that of the Lambert surface. (2) At a high sun elevation, an intensity peak exists in the specular direction if the surface is a hybrid mode. (3) For a hybrid surface, the lower the sun elevation, the higher the albedo of the atmosphere-ground system will be. (4) The degree of polarization decreases rapidly with the increase of the Lambert component of reflection. These results reflect the application for parameterizing surface characteristics in terms of the diffuse reflection radiation and hence for monitoring the surface by remote sensing satellite measurements.