No abstract

-AbstractThe monthly mean vertical cross sections drawn throughout a year show seasonal variations of the tropical, middle latitude and polar tropopauses in the middle latitudes. Mean geostrophic jet streams which are found between two tropopauses are associated with relatively dry areas under them.

This is the final report on our analysis during winter months. The model of the large-scale atmospheric disturbance which was suggested in former reports is made more complete through wind field analysis. It becomes clear that there is a rather close relation between the trough and the divergence field, and that the middle layer divergence in front of the trough is not only an important factor on the cyclone development, but also acts as frontogenesis resulting in the so-called polar front. The vertical motion caused by divergence field explains the characteristic temperature field accompanied by a trough. This seems also to be applicable to the disturbance of a larger scale, such as revealed by 5-day mean.

The amount of precipitation is examined in relation to the upper wind field and temperature field in the pre-summer rainy season in Japan,, June and July, for the years 1950 and 1951. It may be well explained as a result of vertical motions of baroclinic unstable waves in the westerlies. In the atmospheric structure during this season, there can be found a similarity to the structure of baroclinic unstable wave.It is found from the data of 1946, 1950 and 1951 that the horizontal distribution of the mean zonal velocity also has some relations with the mechanism of precipitation, which suggests that a more complete,three-dimensionaltheory of baroclinic instability is needed.Some brief remarkes are given on the relations between the basic flow and the disturbances.

The geostrophic angular momentum transfer vector on the 500 mb level is calculated over the Far East. The angular momentum convergence in a sector of 10 degrees in latitude and 30 degrees in longitude is compared with the local zonal-index tendency in the same area. Good relation is found between them on the variation with a period of several days, but only 30% of the horizontal convergence of angular momentum appears in the local change. Also, 45-day mean angular momentum transfer vector is calculated. Its horizontal convergence shows a remarkable locality having such a value as suffices to cause the zonal velocity change 6 m •sec-I in a day. The eddy transfer has northward component almost everywhere in our region, but is smaller than the transfer due to mean wind.