The long term variations of large-scale atmospheric motions were followed by a simple two-layer model with the nonadiabatic heating and cooling, and viscous dissipation. The computations were made by numerical procedures and partly by an analytical method in which the nonlinear feedback mechanism was taken into consideration. Stability criterion of the growth or decay of the large-scale motion is derived by solving the frequency equation which is a quadratic form with respect to phase speed and with complex coefficients. Kinetic energy spectrum of eddy meridional component was evaluated for the steady Rossby motion. The double maxima appeared in the spectral distribution. Then, the transient motion toward final equilibrium state is followed by numerical process, and the stability of the equilibrium state is examined by means of a variational equation. Time behaviours of the motion are largely characterized by a damping oscillation with well-defined periodicity. The period of oscillation varies widely from about ten days to a month. Finally, conversions between various types of energies, and energy flow diagram are obtained and compared with the results of weather map analyses and numerical experiments. Fairly good coincidences are seen in the direction of energy flow and also magnitude of energy conversions.
In the preceding paper the author has computed the intensity of scattering at 5km level in four wavelength ranges, each of which has the partial energy equal to one fourth of the total solar energy falling on the upper limit of the earth's atmosphere. In the present paper he has investigated the same problem at 75 levels from 1km to 38km height, the interval between neighboring two levels being 500m. Moreover the total solar energy is divided into twelve wavelength ranges, each of which has the partial energy equal to one twelfth of the total. The study is restricted only to the primary scattering in the plane normal to the vertical plane passing through the sun's center in the sky dome. Let T0 be each level point above explained and draw a line passing through T0 and making an angle θ1 with a line passing T0 and the earth's center O. For convenience sake the line thus defined will be hereafter called θ1 line. Let T4 be the intersecting point of θ1 line with the earth's surface or the upper limit of the earth's atmosphere, and T1, T2, T3 be three points which divide the line section between T0 and T4 into four equal lengths. Hereafter the above five points will be called in general Tn . The author has researched and compared the primary scattering intensity which originates at Tn and reaches the level point. Which is the largest of the above five intensities? He has found that the primary scattering is governed by the following four laws in the high atmosphere, concerning the position of the above discussed largest value, hereafter this position being called the maximum position. 1. The distance of the max. position from the level point increases with the height of the level point for the same wavelength, θ1 and sun's altitude (h). 2. The distance in the same meaning as the 1st law increases with the increasing wavelength for the same level point, θ1 and h. 3. The distance decreases with increasing θ1 for the same wavelength, level point and h. 4. The distance increases with increasing sun's altitude (h) for the same wavelength, level point and θ1.
An analysis of the historical series of ice appearance date at Abashiri in northern Hokkaido which on the average falls January 14, was made in terms of the mean monthly January pressure gradient over the northern Okhotsk Sea, and the preceding December pressure gradient upstream over northeastern Siberia, the results obtained showing a moderate relationship based on some 60 years of record and a somewhat closer relationship based on cases when the departure in the pressure gradient from the average equalled or exceeded ±0.8 the standard deviation, σ, suggesting a limited basis for predicting the ice appearance date in northern Hokkaido. Confirmatory evidence of the relationship of the January and preceding December pressure gradients to the north of Japan with the ice appearance date was obtained from an analysis of the frequency of lows crossing Hokkaido and its environs. A similar analysis of the ice disappearance date which on the average occurs April 22, shows a marked relationship with the contemporary April pressure gradient, but little, if any, relationship with the gradient to the north in March, presumably because of the discontinuous change in the general circulation during the spring in that general area. It is further shown that the date of ice appearance at Abashiri taken to reflect the circulation pattern over the general area, correlates moderately with the following mean monthly February air temperature over northeastern Hokkaido.
The photographic observation by aircraft was performed on 20 January 1963 as a part of research project of heavy snowfall in Hokuriku District where is located in the central part of the Japan Sea coastal region of the Japan Islands. The horizontal distribution and the topography of clouds over the Japan Sea near Hokuriku District were obtained bymaking use of the aerial photographs taken from the aircraft at about 8 kilometers level in the way based on the usual aerial photogrammetry. The cloud photographs show a characterisitic pattern changing from scattered cumulus clouds smaller than 1 kilometer in both horizontal and vertical dimensions in the west of the observation area to larger cumulus and cumulonimbus clouds reaching 3 kilometers level in the east where is covered with upper stratiform clouds. It is suggested that cumulus clouds seem to originate over the Japan Sea at least 200 to 300km off the coast of Hokuriku District and a fairly amount of snowmay be released efficiently just in Hokuriku District from the clouds containing sufficient water substance.
In order to include the pressure effect in the concentration of any physical quantity, a potential quantity is defined. If A is a number in an appropriate unit in a unit volume of air and ν the volume of air of unit mass and the relation Aν=A0ν0 holds between two states, α=Ap-l000mb=AFx (P, Pd) is defined as a potential A, where P=1000mb. The functional form of Fx(P, Pd) is given to each process of change, α is conservative for the process. According to this definition, we can define various potential quantities, which are conservative for each assigned process. Potential concentration :ν=n(P/Pd)1/r, where n is the number of particles in a unit volume. Potential vapor amount : P=ρν(P/Pd)1/r, which is conservative for dry adiabatic change and is equal to ξdo, whereξis mixing ratio. Potential water content : W=(ρν+∑n (r) m (r)) Fx (P, Pd), where n (r) and m (r) are the number density and the mass of cloud droplet of radius r. If the process is dry adiabatic Fx(P, Pd)≈(P/Pd)1/r and if the process is moist (pseudo adiabatic and adiabatic), Fx (P, Pd)≈(P/Pd)1/rexp[(r-1)r-1LPd-1∑n(r)m(r)], where L is the latent heat of condensation.
Solving the linearized two level baroclinic model using primitive equations for wave type solution, we examine the relationship between the initial amplitudes of waves and the deviation of wind field from geostrophic balance. The relation indicates that the inclusion of divergent part in the initial velocity field of baroclinic model together with non-divergent component gives the much reduction of amplitude of noise rather than the case of the latter only. The numerical time integration is also performed to show the validity of the theoretical results mentioned above.