The well-known Prandtl-Batchelor theorem for high Reynolds number flows is generalized and applied to quasi-geostrophic flows with or without meso-scale eddies in a closed geostrophic contour. Laminar quasi-geostrophic flows without any external forcings are shown to be stagnant in a closed streamline owing to the Ekman friction. As for the turbulent case, two markedly different mean states are suggested in the limit of both weak eddies and weak dissipation. One state corresponds to the laminar result in the limit and the other is the state of uniform potential vorticity over a domain surrounded by closed geosrophic contours. The latter state is not inconsistent with Rhines and Young (1981).
Space-time cross spectra are experimentally estimated from given sinusoidal waves by use of the multivariate maximum entropy method. This method gives not only power spectra but also cospectra, phase difference and coherence with fine frequency resolutions from short time records. As an example of its application, a space-time spectral analysis is made of external Rossby waves simulated by a GFDL spectral general circulation model.
An objective analysis scheme based on an extension of the correction method was applied to the unprecedented volume of FGGE/WMONEX Level IIa and IIb wind data at 850 and 200mb over the greater WMONEX region (45°N-30°S, 40°E-100°W) for the 1978-79 winter (1 December 1978 to 28 February 1979). Some of the characteristic features of the 1978-79 winter circulation were investigated by examining the monthly and season mean zonal and meridional winds, and streamfunction (vorticity) and velocity potential (divergence) fields. The season mean 850mb wind field revealed a well-defined zone of equatorial westerlies which extended between the equator and 10°S, from the Indian Ocean (80°E) to the central South Pacific (170°E). These westerlies probably denote the extent of the Southern Hemisphere summer monsoon region during this period. In addition, the eastern end of the nearequatorial trough at around 10°S, 170°E may be a major heat source region for the Southern Hemisphere monsoon, as it was dominated by a pronounced low-(upper-) level convergent (divergent) center. Over the Indonesian Seas-western South Pacific area between 5° and 15°S, monthly mean near-equatorial westerlies at 850mb exhibited a substantial increase from December to January. However, this intensification of the Southern Hemisphere near-equatorial westerlies was not due to 850mb northerly surges over the Sumatra-Borneo-New Guinea region, as these cross-equatorial flows from the Northern Hemisphere were of the same general intensity throughout the season. In contrast, significant month-to-month changes were noted in the 850mb cross-equatorial flows near the dateline. At 200mb, strong upper-level westerlies over the Australian continent in December were replaced by easterlies as a weak anticyclonic cell developed over the region in January. Interestingly, this upper-level anticyclonic cell (20°S, 130°E) was located apporximately 40° to the west of a major 200mb divergent center (°S, 170°E) in January.
Some of the characteristic features of large-scale disturbances over the greater WMONEX region were investigated by using 12-30 days filtered wind, vorticity and divergence data at 850 and 200mb for the 1978-79 winter (1 December 1978 to 28 February 1979). Composites of these meteorological variables at each grid point and level were made with respect to changes in the first eigenvector (empirical orthogonal function) coefficients for 12-30 days filtered vorticity data at 200mb. Over South and Southeast Asia, the 200mb circulation undergoes massive changes in association with the eastward passage of upper-tropospheric troughs and ridges. The mechanisms that are responsible for the development, slow eastward propagation (-4ms-1), and dissipation of these disturbances are identical to those described by Murakami (1981c). Pronounced long-period (-20 days) disturbances at 200mb can also be traced from Japan, across the central North Pacific, to the equatorial Pacific. These upper-tropospheric disturbances are primarily maintained through barotropic nonlinear interactions with the winter mean flow and disturbances of all period ranges. Over the western South Pacific, 200mb long-period disturbances are most intense off the east coast of Australia. In this area, vertical vorticity transports due to convective motions appear to be significant. At 850mb, northerly (anomaly) surges are predominant over the Afghanistan-Pakistan, Japan-Philippines and central North Pacific regions. Along the western periphery of the Tibetan Plateau, low-level southwesterlies ahead of an eastward moving trough flow up the extremely steep topographic gradient. This induces strong low-level convergnece, which results in increased westerly divergent winds near the trough and contributes to the enhancement of the local northerlies. Orographic enhancement is also responsible for the acceleration of the northerlies over the Japan-Philippines region. Converesly, over the central North Pacific along about 170°W, intensification of the northerlies is largely associated with transient disturbance activity poleward of 30°N. Along the equator, zonal wind and divergence fields at 850mb exhibit marked intraseasonal (-20 day) changes. Here, anomalous westerlies (easterlies) and convergence (divergence) over the central Pacific coincide with easterlies (westerlies) and divergence (convergence) over the Indian Ocean. The east-west pressure gradient appears to be the primary factor for initiating changes in the equatorial 850mb zonal winds.
hree dimensional structure of a monsoon depression developed over the Bay of Bengal during summer MONEX is examined by the use of aircraft dropwindsonde and conventional upper-air observations. Wind, temperature and relative humidity data are interpolated objectively at 1° latitude/longitude grid points over the area of 11°-24°N and 80°-93°E and at 25mb pressure levels from the surface to the 500mb level for the period 3-8 July 1979. The depression formed over the Bay of Bengal at about 19°N and 90°E on July 6 and moved westward with about 2° day-1 speed and reached a coast line of India on July 8. The developed depression has the maximum vorticity of about 1.5×10-4s-1 and a warm core slightly to the east of the depression center. Horizontal convergence and rising motion occur to the west of the depression center where active convective clouds exist. The horizontal axis of the depression is inclined from southwest to northeast. The depression transports heat and momentum northward and gains its kinetic energy from the mean tonal flow. The analysis during the pre-formation period shows that there exists large positive vorticity of about 5×10-5 s-1 in the area over the bay which is close to the area of depression formation. This large positive vorticity is caused by both the existence of a weak trough and the meridional gradient of the zonal flow. The latitude-height distribution of the zonal wind averaged over the Bay of Bengal for the pre-formation period shows that the necessary condition for instability of the zonal flow is satisfied. A stability analysis of the zonal flow averaged in the lower troposphere below 500mb indicates that the flow is barotropically unstable with the maximum, growth rate at about 3, 500km. The unstable wave has several similar characteristic features to those observed.
Some of the characteristic features of synoptic-scale disturbances over and around the Rocky Mountains were investigated by using 3.5-7.0 day filtered u, υ, T and Φ data at eight levels (the surface, . 850, 700, 500, 300, 200, 100, and 50mb) for the 1978-79 winter (1 December 1978 to 28 February 1979). Composites of each meteorological variable at every level and grid point were made with respect to changes in the first eigenvector (empirical orthogonal function) coefficients for 3.5-7.0 day filtered meridional wind data at 300mb. The composite charts of 300mb winds and geopotential heights reveal that troughs and ridges in this region propagate eastward systematically, with phase speeds of approximately 7(12)ms-1 to the west (east) of the Rocky Mountains. These high mountains exert a strong orographic influence on the movement of low-level disturbances. For example, when anticyclonic (cylonic) cells approach the California coast, low-level northerlies (southerlies) flow down (up) the southern slopes of the American Rockies and enhance the local, low-level divergence (convergence) field. Consequently, the anticyclonic (cyclonic) perturbations tend to move along the southern (not northern) periphery of the American Rockies, where they become most intense. However, the intensity of these low-level disturbances decreases sharply as the enhanced divergence (convergence) field weakens away from the high mountains. At 850mb, strong cyclo- and anticyclogenesis occurs near the eastern periphery of the Rockies. The orographically induced, small-scale, anticyclonic (cyclonic) cells then propagate southward and merge with major anticyclonic (cyclonic) systems at the southeastern border of the American Rockies. When this merging occurs, anomalous low-level, northerly (southerly) surges originate from cold central Canada (the warm Gulf of Mexico) and flow across Texas. Pronounced northward sensible heat fluxes over the central U.S. are associated with this surge activity and indicate that there is a large increase in eddy available potential energy in the area. This region is also characterized by a strong upward geopotential flux into the stratosphere. In contrast, the vertical geopotential flux over the Rockies is downward and contributes to the maintenance of low-level, eddy kinetic energy against frictional dissipation over these mountainous regions.
Forecast experiments on a medium-scale disturbance associated with a long-lived cumulonimbus cluster in the Asian subtropical humid region are made wing a 6-level 77km-mesh primitive equation model. The purposes of the present study are to ascertain whether the medium-scale disturbance is predicted by using the present fine-mesh primitive model or not, and to inspect thermodynamic influence of cumulus convections on the development of the disturbance. We make five experiments (Exp. 1 with standard moisture data, Exp. 2 with moist moisture bogus data as indicated by satellite IR cloud imagery, Exp. 3 with low-level moist moisture bogus data, Exp. 4 with artificial dry moisture bogus data and Exp. 5 without moist convective adjustment scheme) for each of the four development stages of the disturbance: formation, Cb cluster, Cb cluster-depression and depression stage. We also compare the results with prediction by 6L FLM (6-level 150km-mesh primitive model of JMA). In the formation stage, the disturbance is located over the China Continent and the Cb cluster is apparently found on satellite IR cloud imagery. However, the amplitude of the disturbance is very weak, and the baroclinicity in the lower layers around the disturbance is very weak. For this period, all experiments can not simulate formation nor development of the disturbance. For Cb cluster stage, Exp. 2 and 3 simulate the development of the disturbance while Exp. 1, 4, 5 and 6L FLM can not. For Cb ciuster-depression and depression stages, the predictions of all experiments are accurate to same extent. However, the precipitation in Exp. 4 and 6L FLM are much smaller. The present study indicates the significant influence of cumulus convections on the development of the disturbance in Cb cluster stage. The accuracy of precipitation prediction in Cb cluster, Cb cluster-depression and depression stage is evidently improved by the increase of the grid resolution and the improve of the initial analysis of the moisture field.
A three-dimensional numerical model of the land and sea breezes allowing mountain effect was developed. The system of governing equations is so-callde Boussinesq hydrostatic one, using z*-coordinate system as a vertical coordinate which refers the lower and upper boundaries to the ground surface and the upper surface of the model, respectively. The model atmosphere with a vertical scale of 2800m and a horizontal scale of 217.5km is divided vertically into 12 layers and horizontally into 30×30 grids with 7.5km mesh. It is assumed that the surface temperature on sea is constant but that on land changes sinusoidally with a constant diurnal range around the mean temperature at each level. Temperature contrast induced by the above mechanism causes the land and sea breezes and the mountain and valley winds. By using this model, some experiments were carried out for studying the influence of mountains on the land and sea breezes in the case of the Kanto district. In mountain-free model, the most prominent sea breeze circulation was restricted within about 40km from sea shore. The local circulations simulated by the model including mountain effects are in good agreement with observation in the Kanto district. The experimental results show that the mountain and valley winds appear earlier than the appearance of the land and sea breeze circulation. But this phenomenon is not yet confirmed by observation.
The observed thermodynamic structure of the winter monsoon ABL during the air-mass transformation experiment is studied with a simple steady-state inversion model, taking into account the effects of large-scale subsidence, turbulent mixing, advection and radiative cooling. The model output provides the heights of cloud base and inversion, the profiles of static energy and mixing ratio and also the vertical profiles of the turbulent fluxes. The results of this steady-state model are in accordance with observations and other computations, and indicate, that the main processes are included into the model. The strong dependence of the results on some of the input parameters is shown by a sensitivity test.
The mountain-valley wind system in a large valley in Khumbu Himal, East Nepal was investigated on the basis of winds and other meteorological observations carried out mainly at 4, 420m above m.s.l. during about a whole year in 1975. The characteristics of this wind system revealed by the observations are as follows. The daytime valley wind is observed at a frequency of more than 90% in all seasons which means that this wind system is very stable throughout the year. The wind system in the dry season (Jan.-May, Oct.-Dec.) showed the typical mountain-valley wind circulations, while the valley wind in the wet summer monsoon season continued until 04:00 LST throughout the night. This nighttime valley wind is thought to be due to latent heat release from cloud formation and preicpitation. The annual trends of the daytime average wind speed showed maximum in May-June at 8.5m/sec and minimum in Oct. at 5.0m/sec. During the observation period, there were 45 days when the valley wind was abnormal. They were classified into three types of abnormality. The causes for them were the existence of snow cover, middle and high level clouds and the invasion of strong upper wind into valley. Most remarkable change was seen in the effect of snow cover. After the snowfall, the value of daytime wind speed decreased to 30 to 90% of the value before the snowfall. This decrease in wind speed was much remarkable when the air temperature was above 0°C.
The aerosol concentration in the atmosphere and the number of aerosol particles attached to falling snow crystals were simultaneously observed in Sapporo and at the summit of Mt.Teine (1, 024m) in March 1978. As a result, it was found that the number of aerosol particles attached to snow crystals in Sapporo was about one order greater than at the summit of Mt. Teine. This shows that the snow crystals collected the airborne aerosol particles over Sapporo during their fall. The collection efficiency of aerosol particles by snow crystals was calculated on the basis of the observational data obtained in Sapporo. The collection efficiency by rimed snow crystals was about one order greater than that by the unrimed snow crystals. The aerodynamic effect and rain-out effect may be attributed to this large difference. As for larger aerosol particles from 0.4 to 10μm in diameter, it was found that the collection efficiency by the unrimed snow crystals increased with the increasing aerosol size, and ranged from 1×10-3 to 6×10-2. These values are almost in agreement with those experimentally obtained by other workers. As for aerosol particles smaller than 0.1μm in diameter, the collection efficiency was roughly estimated to be 1×10-2.
Observations of ice crystals at Amundsen-Scott South Pole Station, Antarctica were made during January 1975 and November 1978. Results of analysis indicate that the maximum precipitation intensity varied between 0.01 and 0.2 mm•hr-1 and the size distribution of ice crystals could be expressed by the exponential equation, ND=N0 exp(-ΛD), similar to other types of solid precipitation. The Z-R relation, Z=10 R1.0, was introduced for ice crystals. The relation between axial ratio (c/a) and length of the c-axis for bullet and column types of ice crystals coincided with results previously reported.