Momentum budget analysis is made o ver western Japan in the Baiu season in 1968and 1969. Ensemble means of results obtained and other variables are compared between two cases, i. e., with an area-mean precipitation of 5 mm/12 hr or more (Case MH) and one of less than 5 min/12 hr (Case NL). Heat budget analysis shows that convective activity is high throughout the troposphere in Case MH and confined to the lower troposphere in Case NL.
Although some systematic error is contained in the estimated residual terms of the momentum equation, especially in its component normal to the wind, it is shown from comparison between results for Cases MH and NL that in Case MH the convective vertical transport of westerly momentum is downward with a maximum value of 2-4 dyn/cm²in the mid-troposphere. The estimated downward transport of southerly momentum does not exceed 2 dyn/cm².
Total kine t i c energy dissipation for the layer 1000-100 mb is about 20 w/m² in Case MH, one order of magnitude greater than in Case NL. Dissipation rate remains low in the layer 800-600 mb even in Case MH in spite of the great total dissipation. These facts suggest that in Case MEI there exist energy transfers from the large- to subgridscale motion, especially in the upper troposphere, and from the upper to the lower troposphere through the downward momentum transport by the subgrid-scale or convective motion. The difference in the interaction between the large- and subgrid-scale motions may be responsible for the difference between the observed wind profiles in Cases MH and NL.

View full abstract

Special observation to make clear t he mechanism of the medium scale disturbance generated on the sea to the south of Shikoku island was performed under the co-operations of the marine department of the Meterological Agency, the Osaka district observatory and the Kobe marine Observatory, in October, in 1974 and 1975.
The generating process of the medium scale disturbance that was formed about 9 J. S. T.,12th, October,1974, was described using the numerical analysis of the model of 5-layer quasi-geostrophic approximation in connection with radar and synoptic analysis. By the analysis of 12-hr thicknesses changes of the lower, middle and upper layers, the difference between the thicknesses observed at 24 hrs and 12 hrs before the formation of the depression shows a remarkable general trend to increase in the neighborhood of Japan, while the thickness change of the upper layer of 300-500 mb over the area where the formation of a medium scale disturbance is expected, shows a trend to decrease.
A deep medium scale disturbance appears in the middle and upper lay e rs of the above-mentioned area, but not yet near the surface. At the time that the depression is formed near the surface, the entire layer of the troposphere over the medium scale disturbance is warmed remarkably by temperature advection and condensation heat. With regard to the generating process of the medium scale disturbance, the roles of vorticity and temperature advection, vertical motion, condensation heat and heat transfer from the sea surface are analysed using the model of 5-layer quasi-geostrophic approximation. The effect of density of the observation network on numerical analysis is discussed.

View full abstract

A new pressure distribution formula in typhoon fields is proposed which represents accurately the shape of typhoon pressure fields in their various stages, and also will be capable of being adopted for practical usage. This new formula is derived from an improvement on several well-known pressure formulas.
The above pressure functions have been p roposed by several authors, Schloemer's (1), Takahashi's (2), Bjerknes' (3) and Fujita's (4). These functions are shown in Fig.1 in normalized scale. Those functions contain only one parameter γ₀, which is used as a non-dimensional parameter on distance r and is related to the position of the maximum pressure gradient, γ_max. The characteristics of the curves, P(γ₀) and P(γmax) are given in Table 1. It is a weak point of them that the property of these pressure curve is not altered by a change of γ₀ which is estimated from a distribution of the observed data. Then a satisfactory goodness of hitting by use of them can not be expected especially when the pressure fields of a typhoon change with the stages of its life cycle.
An improvement on the pressure function was planed by Fujita introducing another parameter “a” of Eq. (5) to avoid the weak point mentioned above in his Eq. (4). It was too complicate to hit the curve (5) for the observed data.
We will try to improve these pressure functions by introducing some parameters which add freedom to them. Firstly, these functions are analyzed mathematically to show that they can be derived from a few well-known fundamental differential equations describing the so-called “growth process”. Secondly, as the mathematical background for the formulation of these curves is made clear, some methods for their improvement are examined under conditions where the assumption from which these functions are derived is maintained.
Though the linear assumption in (7) gives Schloemer's formula (1), a more general rule for the formulation of (1) is given by the differential Eq. (9). This can be called the pressure gradient equation, which expresses that the pressure gradient is in proportion to the deviation of pressure from the stationary state Pc and that the proportional constant gives the value of -γ₀. The subsequent discussion is developed by use of the pressure gradient equation. The solution of Eq. (9) is shown schematically in Fig.2. As the integration constant in (9) is fully utilized to formulate the equation (1), there is no room for any improvement on (1).
When the transformation (12) is i ntroduced, Takahashi's (2) and Bjerknes' (3) are expressed by the straiget lines in Fig.3. The pressure gradient equations for them are given by (15), which represents the pressure gradient proportion to the product between the decrease from the Pand the increase from the P. The proportional constants in (15) are related to the power of the main term, (γ/γ₀) in (2) and (3). This equation is analogous to the second order of reaction in the field of chemistry. The solution of (15)is shown schematically in Fig.4, which is called a (symmetrical) logistic curve. The pressure functions (2) and (3) are derived from Eq. (15) under the restriction that the parameter γ₀ must be chosen to make the integral constant C=0 in (15), and that n in (15) is determined a priori as n=1 for Takahashi's and as n=2 for Bjerknes'.

View full abstract