Thermal excitation as the cause of semidiurnal atmospheric oscillations is discussed, special emphasis being put upon the formation of node around 30 km. It is theoretically shown that if somewhere in the mesosphere above around 30 km is thermally excited only one node is formed around 30 km. The height of the node is dependent very little upon the temperature of the atmosphere which is assumed isothermal. No node is formed by tropospheric thermal excitations. Provided that the existence of 30 km node is true, the place of thermal excitation responsible for the semidiurnal oscillation seems to be in the mesosphere
Thermal excitation as the cause of semidiurnal atmospheric oscillations is discussed in continuation of Part I (Sawada, 1962), special emphasis being put upon the effects of the temperature profile on the formation of node around 30 km. An atmospheric temperature profile is designed so that mathematical treatment becomes very simple, but it approximates the real profile. The result shows that the temperature profile does not affect the conclusion obtained for an isothermal atmosphere discussed in Part I, and a mesospheric thermal excitation produces a nodal surface around 30 km for pressure and wind oscillations, but no node for the vertical motion.
The vertical distribution of wind velocity in the shallow layer over the ground is represented empirically by the power law u∞zP for obtaining design wind pressure to the structure, where u is wind velocity at height z above the ground. The power exponent p is specified only by the roughness of the ground under the strong wind. However, under moderate and slow wind, the power exponent also depends on the vertical difference of temperature of the air layer. The observations of wind profile have been made in the air layer up to 48 m above the ground at Kokubunji in the outskirts of Tokyo, and dependencies of the power exponent on time, wind velocity, temperature difference and stability of the air layer were obtained. It seems to be rather advisable to use logarithmic wind profile than power wind profile for the study of transport and diffusion in the atmosphere. In this case, linear term of height is added as the correction for the stability. The profile, expressed as u=u∗k-1(ln z/zo+βL-1 •(z-z0)), where u*, zo and L are friction velocity, roughness length and stability length respectively, was derived from similarity hypothesis, and has been already assured to fit well results of observations in the lowest ten meters above the level grassy land. Following the logarithmic plus linear wind profile, our results have been analyzed and the profile seems to represent the observed wind velocity in the same degree as power profile does. Parameters of wind profile, u∗, zo and L, were calculated from the data and the comparision with the results of shallower layer was made.
In the Far East we have two pronounced rainy seasons, one in the early summer and the other in the early autumn. During the early autumn, the surface map is characterized by the circumstance that there are two somewhat persistent anticyclones over Siberia and over the Pacific ocean, as is presented in Fig. 2b. In between these two anticyclones we find a trough stretching from north to south, which might be favourable for the rainfall near Japan. In Figs. 2a, 2b and 2c, the establishment of the Siberian high seems to be connected with contemporaneous occurrence of nonadiabatic cooling over Siberia. A comparison between Fig. 3 and Fig. 4 reveals that the retreat of the Pacific high keeps pace with the change of nonadiabatic effect from cooling to heating. As a common belief since the discussions by Jacobs, the western Pacific ocean acts as a cooling source in summer, and as a heating source in winter. The rainy season comes to an end when the Siberian high covers the whole Japanese islands, as can be observed in Fig. 3c. We also note in Figs. 3a, 3b and 3c that the so-called monsoon low, known to locate to the north of India during the monsoon season, weakens and shifts towards south as the season advances. In the upper atmosphere we encounter the marked blocking situation over the Far East, i.e., the splitting of jet stream into two branches, as can be found in Figs. 5a, 5b, 5c and. Figs. 6a and 6b. Glancing at Fig. 6c, such blocking situation is observed to break down at the same time with the end of the rainy season. Yin (1949) has suggested that the establishment of the subtropical jet stream along the southern periphery of, the Tibetan Plateau in the beginning of October occurs simultaneously with the retreat of the southwesterly monsoon over India. We observe the same kind of sequence of events in Fig. 8, giving the time isopleth of zonal wind velocity along the longitude 85°E. Accordingly it may be concluded that the retreat of the southwesterly monsoon from India almost coincides with the end of the rainy season near Japan. Most important vapor transport for the rainfall near Japan is accomplished by the southeasterly monsoon which flows anticyclonically around the Pacific high. In Fig. 10 we find that vapor transfer accompanied by the southwesterly monsoon seems to be im portant only over the southeast Asia. As for the rainy season in the early summer, how ever, the latter is as important as the former for the rainfall near Japan, which was discussed by Murakami (1959). Fig. 12 gives the horizontal distribution of F computed by eq. (12). A positive sign indicates that evaporation exceeds precipitation and a negative sign, precipitation exceeds evaporation. One of the interesting things in the figure is the existence of a broad band of positive value stretching from east to west across the Eurasian continent. An inspection of Fig. 3 and Fig. 12 shows that this positive band agrees fairly with the axis of the surface anticyclonec belt. Speaking of the zonal mean north-south vapor transport [qv], the greater part of it is accomplished by the mean meridional current and only small fraction of it, by the disturbances. Fig. 16a presents the vertical and latitudinal dependence of vapor transport due to the mean meridional current. The transport attains its maximum value at the ground surface and rapidly decreases its magnitude with increasing height. Moreover the trans port is northward in the higher latitudes and southward in the lower latitudes. The vapor transport due to disturbances presented in Fig. 16b is the maximum around 850 mb level, and is northward at almost all latitudes with thee largest value around 60°N.
The latitudinal distribution of high cloud and middle cloud is computed statistically by use of the eye observation data along 14°E over Japan. A schematic representation of the cloud system is proposed based on the relation of the distribution of cloud to the fronts and the jet streams in the upper air.