Numerical experiments are carried out by using the Yamagata-Flierl equation for eddies of a scale intermediate between the Rossby deformation radius and the beta-scale. These experiments are performed for two types of zonal flows, one has a linear shear and the other is sinusoidal velocity profile along meridian. Through these experiments, we tried to investigate the motion and the life time of the intermediate scale eddies in the cases of linear shear zonal flow, and attempted to compare the results with the development of the South Tropical Disturbance (STrD) observed in 1979-81 on Jupiter for the cases of sinusoidal velocity profile. From the experiments for the case of linear shear flow, we found the following facts. (1) If the eddy and zonal flow have the same sign of vorticity, the eddy can survive stably. If the eddy and zonal flow have the opposite sign of vorticity, the eddy disappears or is deformed remarkably. (2) In the case of uniform zonal flow, a cyclonic eddy can survive stably in a high speed westward flow, whereas an anticyclonic eddy can survive in a low speed westward flow. When the westward zonal flow has the non-dispersive wave velocity, both cyclonic and anticyclonic eddies can survive stably. (3) In a low speed westward flow, a cyclonic eddy shifts poleward and an anticyclonic one equatorward. In a high speed westward flow, a cyclonic eddy shifts equatorward and an anticyclonic one poleward. (4) Anticyclonic eddies move westward with higher speed than that of non-dispersive wave. On the other hand, many of cyclonic eddies move westward with lower speed. In one of our experiments for the sinusoidal velocity profile, a cyclonic eddy, whose radius is comparable to the latitudinal width of the westward jet, can go poleward through this jet, and is elongated in northeast and southwest direction in the northern hemisphere before it disappears. The STrD seems to be this kind of cyclonic eddy, because the STrD also had cyclonic vorticity and developed poleward, and because the STrD also was elongated in the same direction as our results. The life time and longitudinal motion of this eddy is not consistent with those of the STrD, but through collisions with small vortices the STrD may get a long life time.
Cold surges around the Tibetan Plateau in a numerical prediction model have been analysed. Cold surges trapped in the periphery of the mountain and in the lower troposphere are well simulated in the model. From the comparison of forecasts of models with different mountain heights, the enhancement of northerly wind along the eastern boundary of the mountain, which is similar in structure to that of coastal Kelvin waves, is demonstrated. The estimates of a width and a phase speed due to the linear theory are two or three times smaller than those due to the forecast results, which may indicate the limitation of the linear theory. To investigate the mechanism of cold surges, momentum budget analyses at the 850mb level have been performed on two cases showing enhancements of northerly winds. The first enhancement of northerly associated with the southward passage of a cold front is noticed, where the acceleration effect due to the ageostrophic component is dominent. However, this northerly has gradually weakened at the southeastern corner of the Tibetan Plateau. Main enhancement of the northerly occurs in the cold air behind the cold front and the northerly is further propagated southward. Although the effect due to the ageostrophic component plays a dominant role in the early stage of propagation, the effect due to the non-linear advection term becomes important in the late stage around 20°N. By this effect the northerly leaves the mountain and travels further southward. Although basic features are well simulated in the model, there remains some errors; One is the intrusion of the cold air into the Indo-China peninsula instead of the South China Sea. This error may be due to the insufficient treatment of the orography over the IndoChina peninsula. Slight improvement has been noticed in an experiment with the envelope mountain. Another error is the overprediction of cold surges around the mountain. This suggests the insufficient treatment of the effect of the mountain, probably the neglect of the effect due to the sub-grid scale undulations. Extra eddy fluxes due to mountain waves excited by the sub-grid scale undulations can be expected. In fact, an experiment including a parameterization of this effect showed a remarkable improvement of the forecast fields.
A new second-order finite-difference form of three-dimensional momentum equations in the anelastic system is proposed. This conserves total kinetic energy in the three-dimensional motion as well as total enstrophy in the two-dimensional motion. It can be applied to the equations expressed by the curved orthogonal coordinate and to the variable-grid model in the Cartesian coordinate system.
Data collected by the NCAR Electra aircraft during AMTEX '75 have been analyzed to determine the spectral characteristics of four boundary layer types based on air-sea temperature differences (ΔT) : (a) Strongly convective (ΔT_??_5°C) with mesoscale cellular convection (MCC), (b) Strongly convective (ΔT_??_5°C) without MCC, (c) Weakly convective (2°C_??_ΔT<5°C) and (d) Non-convective (ΔT<2°C) . Comparisons between the spectra of the four types suggest that a boundary layer containing MCC does show some distinguishing characteristics which are not observed to be present in other boundary layer groups. The specific humidity (q) spectra for flight segments containing MCC were characterized by a prominent peak in the wavelength band corresponding to the observed MCC cell size. Also, the high wavenumber end of the q-spectra at the lower level was observed to contain characteristically larger amounts of energy. The most significant new finding of this study was some evidence supporting the presence of a basic convective mode (BCM) with an aspect ratio comparable to those of Benard-Rayleigh convection cells. It is suggested that an organizational process from the BCM to the MCC mode occurs, with energy transferred up-scale from high wavenumbers to low wavenumbers. The strongly convective boundary layers without MCC were influenced by other synoptic forcing which apparently prohibited the development of MCC. Two of these cases also showed correspondingly larger amounts of energy at the high wavenumber end of the spectrum than in the weakly convective and non-convective cases. Another case near a pattern of MCC cells showed the presence of a strong MCC mode although the MCC cloud patterns were not present in satellite imagery. The Weakly Convective cases were distinguishable in the small amount of energy present throughout the spectrum with little to no evidence of the MCC or BCM modes. The non-convective group also showed little energy present in the spectra at lower levels. Two flight segments at the upper level, however, showed evidence of forced convection due to the presence of a nearby cold front. This was evident in the q-spectra as a prominent peak at low wavenumbers.
The spatial structures of atmospheric mixed and transition layers were investigated using the image data of aerosol distribution observed by a scanning Mie lidar. The mixed layer height, the transition layer thickness and the horizontal scale of convective cell structures in the transition layer as well as their time variation were analyzed. The main results are : the mixed layer height reaches its maximum at about 13h and its rising speed is 100-300m/h in the period of 8-12h. The transition layer thickness increases in the morning and takes a constant value thereafter. The aspect ratio of the convective cell structure is estimated about 2-4 in the early morning and reaches about unity as the mixed layer develops.
Observational study is made on the morning and evening surface boundary layer and on the nocturnal quasi-steady state boundary layer. It is found two type breaking processes of surface inversion layer. For moderate wind cases, a stable surface layer is heated up with nearly uniform heating rate without accompaniment of an unstable layer near the ground surface. For weak wind cases, however, after the unstable layer is formed first near the ground surface, heating process propagates upwards. In these two processes the countergradient heat flux (C. G. H.) is observed. Possible mechanisms of the counter-gradient heat flux are proposed; one is an exchange process between neighboring hot and cold bubbles, the other is the penetration of the hot bubbles into the stable layer. In quasi-steady night, scaling temperature θ*, is depends on the wind speed and surface cooling rate, and the ratio of sensible heat flux to net radiation is proportional to the surface friction velocity.
The aerosol optical thickness of the atmosphere was measured using a spectropyrheliometer at Tsukuba for three winter seasons from October 1980 to March 1983. The optical thickness was abnormally large in the ('82-'83) winter, because of the enhancement of the stratospheric aerosols due to the volcanic eruptions of El Chichon. Size distributions of aerosols in the vertical air column were inferred, which show characteristic features of bimodal distribution for the pre-E1 Chichon season and power law distribution for the post-E1 Chichon season. Optical properties of the El Chichon volcanic aerosols have been estimated; the stratospheric optical thickness was about 0.1 on the wavelength average in the ('82-'83) winter, and aerosol size distributions were relatively monodispersive with a mode radius around 0.3μm. Possible effects of the enhanced stratospheric aerosols on the solar radiation budget were estimated and compared with available observations. The El Chichon induced aerosols in the ('82-'83) winter could reduce the total solar radiation on the ground by as much as 3-4% on cloudless days, and enhance the spherical albedo by about 10%.
According to the lidar measurements of stratospheric aerosol for 2 years after the large eruption of El Chichon (Mexico, March and April 1982), the vertically integrated backscattering coefficient showed large fluctuations for about 5 months after the eruption and began to decrease at the end of November 1982. The e-folding decay time of the integrated backscattering coefficient was about 5.6 months for the period from November 1982 to May 1983 and it was about 12.4 months for the period after May 1983. The peak height of the backscattering coefficient and the centroid height of the integrated backscattering coefficient became lower in the earlier decay period, suggesting sedimentation of particles. The time variation of the peak height of the aerosol layer implies a decrease in the mean radius of particles around the peak of the aerosol layer. It is speculated that sedimentation of aerosols played a significant role in the removal of aerosols from the stratosphere in the period from about 6 months to one year after the eruption.
The vertical distribution of ammonia mixing ratio is reexamined by using a one-dimensional numerical model. As ammonia molecules are removed effectively from the atmosphere by absorption and wet deposition and as its low concentration is often observed on the ground, its atmospheric concentration is suggested to be lower than those calculated by previous researches.