There are two types of depressions in China Eastern Sea. One of them grows on a front, which breaks out in an old and unstable cP air mass, and the other on a front of mT air mass which has rushed into cP air mass. The authors have investigated their structures and mechanism of growths by the examples.
The author tried to derive theoretically a hygrometer equation, and explain in a new way that the wet-bulb temperature T' defined by the above equation should change pseudo-adiabatially in the case of the adiabatic vertical motion of damp air while the air temperature T follows a dry-adiabatic change. Then by applying the general idea of entropy, as was done by H. Arakawa, the equivalent potential temperature defined by C. G. Rossby was derived approximately as a function of wet-bulb potential temperature. But in applying those formulæ given by the above authors to air mass analysis, it must be remembered that some errors may enter into the calculation in an order of the difference between dew point and wet-bulb temperature.
It is a question whether the abnormal elevation of sea level in the Oosaka bay, produced by the drift current near the Kii channel and by the decreasing air pressure when a cyclone passed over the western side of the bay, be constant along the coast or not. The writer tried to determine the constants of the mean abnormal elevation of sea level in any place along the north-eastern coast fo the bay by the observational data of Sept. 11th 1937 at Hukae and Oosaka. The results obtained are as follows:- Am=-1.93+1.32 (760-P)+0.4157_??_Wdt……for Hukae Am=-7.74+1.32 (760-P)+0.4577_??_Wdt……for Oosaka where Am means the mean abuormal elevation of sea level in cm in the bay. Using the same data it is found that the most effective wind direction to cause high water is “SSW” and in this case the empirical formula for estimation of abnormal elevation is as follows:- A=-1.93+1.32 (760-P)+0.4157_??_Wdt+0.114υ2_??_(±3.4). Instead of “SSW” components of wind, the pressure gradient components of the direction of 160° are employed as a suitable direction for this purpose and obtained the following formula. A=-1.93+1.32 (760-P)+0.4157_??_Wdt+4.362G cos (160°-ψ).
Many attempts have been made on the atmospheric oscillation but, because the vertical current are neglected in them, they have a failure that a maximum fo the pressure occurs at the time of the highest temperature and the amplitude calculated does not fit to the observations satisfactorily. In former papers the author treated this problem with respect to cartessian coordinates taking into account the vertical current. In the present paper the author solved the problem regarding the earth as a rotating sphere. In the diurnal case, from a physical point of view, the solutions of the equation of oscillation are separately treated, one of which is the forced oscillation caused directly by the temperature change and the other is a free oscillation. Then it is explained that, if the temperature changes as to the height making a logarithmic spiral in the amplitude and phase-diagram as W. Schmidt or D. Brunt shows, the forced change of the pressure also makes a logarithmic spiral but its phase delays 15 hours, and the free oscillation has the phase which is constant to the height though 3 hours later than that of the temperature and the amplitude which is almost constant to the height but a little smaller than that of the forced at the ground. Thus the resultant of this forced and free oscillations may well agree with the observed diurnal pressure change. As for the semi-diurnal oscillations, there is of course the forced one as the result of the temperature change of the half day period, though the main part is the free oscillation which has a cause in the diurnal case, and its phase and amplitude almost do not vary vertically. Nextly the author calculates the pressure changes using only the temperature observations at Lin lenberg and compares them with the pressure observations of Potsdam. Lastly the formula which shows the amplitudes of the free oscillations has some inadequateness for the lower latitudes but this defect can be removed if the surface resistance of the earth be taken into account, but it can not be assured numerically because the coefficient of resistance is an uncertained constant.
The equation of the exchange of the water vapour in the atmosphere is given approximately by the following equation, assuming that the direction of the x-axis is normal to the equi-vapour-tention line, where δm is the condensed mass per unit volume per unit time, φ the mass of water vapour per unit volume, u the horizontal velocity, Kn and Ka the coefficients of the horizontal and vertical diffusivity. The right hand side of this equation was calculated numerically and it is found that the most important exchange of the water vapour is done by the convection and the horizontal eddy diffusion. The condensation at each height of the atmosphere was calculated by this equation.
It is believed by the Central-European School that “Glashauswirkung” of water vapour is the primary cause of pressure gradient prevailing in the atmosphere. Such a problem is numerically treated here, the main attention being devoted to the maintenance of energy of anticyclones in horse latitudes. The deepening of the anticyclone due to “Glashausabkülung” of water vapour is at most of the order 0.48×106 ergs sec-1m-2 in energy while the energy dissipation amounts to 0.76×107 ergs sec-1m-2 at least. Thus the theo_??_y by the above school is partly rejected here.
The effect of various meteorological elements on the diurnal variation of air temperature were calculated and it is found that, in winter the greatest effect is due to the eddy conduction from the surface both before and after sunset while in summer it is due to the latent heat of the evaporation from the surface before sunset and the eddy conduction of the air after sunset.
It is proved that the diffusion in the atmosphere is not isotropic and the horizontal diffusion coefficient is much larger than the vertical. In this paper, the horizontal eddy diffusion coefficient was calculated from hourly observation of wind velocity and found that it is about 109.