Some of the characteristic features of the winter monsoon circulations are investigated using 200mb wind data operationally determined by the National Meteorological Center, Washington, D.C., during three winters in 1970-71, 1971-72, and 1972-73. Three-year winter mean 200mb divergence was large over the East China Sea, the Philippines, the South China Sea, and Indonesia. A pronounced velocity potential outflow center was located near Indonesia with a prominent N-S overturning, vertical circulation along 125°E, while a distinct 200mb anticyclone lay centered to the east of the Philippines. Energy equations, similar to that proposed by Saltzman, were applied to three specific latitudinal regions to investigate the atmospheric energetics in the wavenumber domain during the three winters; region 1 (14.8°N-30.8°N), region 2 (0°-14.8°N), and region 3 (14.8°S-0°). In region 1, wave-zonal mean flow interactions are substantially more pronounced than wave-wave nonlinear interactions. Regions 2 and 3 are characterized by the predominance of wave-wave interaction processes. A computational model to partition kinetic energy exchanges into standing (winter mean) and transient wave motions, has been proposed. Most transient waves, over all regions, receive kinetic energy from standing waves via “standing to transient” wave interactions. All transient waves furnish kinetic energy to zonally averaged winter mean flows with large contributions from wavenumbers 3 to 8.
A diagnostic method including cumulus updrafts and downdrafts is applied to GATE A/B-scale upper-air observations for Phase III to study the interaction of the cumulus convection with large-scale motions. Average results for Phase III show that the mass fluxes for deep and shallow clouds are larger than those for clouds with tops in the middle troposphere and downward mass fluxes equivalent to 40%-50% of updrafts convective mass fluxes are accomplished by cumulus downdrafts. Correlation coefficients between deep, middle and shallow cloud mass fluxes and the large-scale vertical velocity are computed to examine the response of cumulus clouds to the large-scale forcing. The results show that deep clouds are highly correlated with the largescale vertical velocity at all levels, middle clouds have little correlations and shallow clouds are correlated only with the vertical velocity near cloud base. Maximum activity of deep clouds occurs nearly at the same time as maximum large-scale upward motion at higher levels but about 6 h after maximum low-level convergence. Strongest downdraft mass fluxes occur corresponding to maximum deep cloud activity. Shallow cloud mass fluxes have negative correlations with the low-level convergence indicating that shallow cloud activity is weak when the low-level convergence is large. A compositing method is used to clarify relationships between the cumulus convection and the African wave disturbance observed during GATE. Maximum deep cloud mass flux is found in the region near the wave trough where the large-scale vertical velocity has a peak. This is also the region where largest downdraft cloud mass fluxes exist. Shallow cloud activity is weak near the wave trough but strong near the wave ridge. Weak subsidence with amplitude of a few mb h-1 is present nearly everywhere in the environment of the clouds in the wave disturbance, but slightly larger subsidence occurs in the regions where shallow cloud mass fluxes dominate.
During the period from June 16 to 23, 1969, oscillations in height, temperature and wind with a period of about 2.5 days were detected in the upper troposphere over the Japan Islands. Detailed analysis shows that these periodic variations were associated with a weak disturbance with a wavelength of about 4, 000km around the jet stream in the Baiu frontal zone. Structure and energetics of the disturbance will be described based on the periodic variations which were picked up by fitting a sinusoidal curve to 6-hourly observed values. Energetic consideration indicates that the disturbance is a baroclinic wave. The kinetic energy of the disturbance is produced from potential energy below the jet core and from kinetic energy of the mean flow just north of the jet axis. It is, however, inferred that the essential energy source is potential energy.
On June 27, 1972, torrential rains fell over the southwestern Kyushu, the western part of Japan, when a weak frontal depression passed to the north of Kyushu. The large-scale characteristics of heavy rainfalls in the Asian subtropical-humid region are represented through the case study on this heavy rainfall. Synoptic and subsynoptic-scale analyses indicate that 1) the heavy rainfalls occurred within the subtropical air mass to the south of the subtropical jet stream's axis; 2) destabilization of the stratification on the heavy rainfall area was not caused by the cold air intrusion aloft but mainly by the intrusion of the tropical air mass in the lower layer; 3) the extremely moist tropical air mass was transported toward the depression from the south along the western periphery of the Pacific subtropical anticyclone. We also give our attention on the differences between the large-scale features of the heavy rainfalls in the Asian subtropical-humid region and these of severe storms in the subtropical-humid region of U.S.A. In essence, the heavy rainfalls are the convective phenomena within the subtropical air mass occurred under the condition of the intrusion of tropical air, while the severe storms are those within the subtropical air mass occurred under the condition of strong cold advection aloft.
Detailed features of heavy rainfalls in three medium-scale disturbances developed in Baiu front over the southwestern Japan Islands have been studied basing on time series of 10-min rainfall amounts and 3-radar PPI composite photographs. For rainfall analysis, the rainfalls were classified into two categories, continuous and convective rains, according to intensity fluctuation and intensity threshold (2mm/10min). Results revealed that the wave-shaped echo system associated with one medium-scale disturbance consists of thin weak widespread echo in the front and in the north of the disturbance and of convective echo cells distributed around the crest or the trailing portion of the wave. The convective echo cells are mostly organized into mesoscale echo clusters arranged with 100-150km spacing. Results also revealed that the wave-shaped medium-scale rainfall system consists of two specified areas. The one characterized by continuous rain spreads to north and east of the wave crest, corresponding to thin weak widespread echo area. The other characterized by convective rain is the heavy rainfall zone in the crest or the trailing portion of the wave, corresponding to the areas of convective echoes organized into mesoscale cluster. Applying spectral analysis method to time series of 10-min rainfall amounts, it isconfirmed objectively that the rainfalls in the heavy rainfall zone are due to mesoscale pulsations of convective rain. The stratification over the convective rain area is evidently different from that over the continuous rain area, i.e., the lower layer is unstable in the convective rain area (southern convective echo area), while stable in the continuous rain area (northern thin weak widespread echo area).
In this paper a non-linear model is derived to represent mesoscale cellular convection in the earth's atmosphere. The moisture content within the layer is allowed to vary and a comparison is made between the results of the numerical integrations and the observed characteristics of convection cells that occur, primarily over warm ocean currents.
Micro-physical processes around melting layer in precipitating clouds were studied on the basis of quantitative observation by the use of a vertically pointing radar to the northern side of “Baiu” front on June 28, 1975, in Nagoya. Following three periods which are different in fine structure of radar echo and micro-physical processes were found out in travelling radar-echo system from the detailed analyses of vertical gradients of radar reflectivity factor. In period A, when a low laid on the front was about 1, 000km west-south-westward from Nagoya, it is not apparent whether there was breakup of snowflakes in the melting layer or not. In the region above melting layer, ice particles would have evaporated in some parts of this period. In the region below melting layer, the evaporation of raindrops would have been prevalent. In period B, when the low was about 700km south-westward, from Nagoya, it is inferred in the melting layer that the discrepancy in the type of ice particles above 0°C level would have resulted in the difference in aggregation activity of particles (snowflake formation) in spite of similar size distribution of them. In the region above melting layer, the accretion of supercooled droplets and/or the aggregation among ice particles would have occurred. In the region below melting layer, raindrops would have falled without remarkable growth or evaporation. In period C, when the low was about 500km south-south-westward from Nagoya, breakup of snowflakes would have occurred in the melting layer. In this period, the discrepancy of size distribution of particles above 0°C level is inferred to have caused the difference in aggregation activity in contrast with period B. In the region above melting layer, it is suggested that the accretion of supercooled droplets and/or the aggregation among ice particles would have occurred. In the region below melting layer, raindrops would have coalesced among them or captured cloud droplets.
Empirical relation between the turbulent velocity υ and the scale of phenomenon L in the atmosphere is obtained as υ-L1/6 instead of -L1/3 for the wide range including the inertial subrange, which was expected as well as in the ocean. Then Okubo's experimental diffusion relation, K-L1.19 for the wide range, instead of L4/3, is explained by the above relation and the new enercv spectrum and auto-correlation function are proposed.