The linear stability of a front whose lower layer has a uniform potential vorticity is investigated. The results are compared with the unstable modes of a frontal model which consists of two homogeneous flows with an interacting interface. Unstable modes with a phase speed close to the velocity of the basic flow in the lower layer, which exist in the frontal model of two homogeneous flows, are stabilized in the present model. This feature is explained by the absence of Rossby waves in the reduced one-layer problem of the lower layer resulting from the uniformity of its potential vorticity. The results show that the stability of a front is strongly affected by the potential vorticity distribution. Careful attention is needed for the application of the frontal model with uniform potential vorticity, which is a critical situation.
The seasonal variation of medium-scale waves (whose wavelengths are 1000-3000km) in the upper troposphere is examined using regional climate model data provided by the Meteorological Research Institute. Medium-scale waves in the upper troposphere are observed at mid-latitudes over East Asia and the western Pacific throughout one year. The waves are most active in the spring season. The phases of the waves begin to propagate eastward at about 100°E. The latitude where the medium-scale waves are dominant follows the meridional shift of the subtropical westerly jet. By using a lag-correlation analysis the eastward phase velocities are estimated at about 24ms-1 in the winter and about 17ms-1 in the summer, which are always larger than those of the synoptic-scale waves. The difference of the phase velocities of the two waves is consistent with the results shown by Sato et al. (1993) for the spring of 1990. The vertical structure of the medium-scale waves is also examined. Their variance has two maxima in the vertical: one is located at 300hPa corresponding to the tropopause, and the other is located near the surface. The phases of the lower waves propagate more slowly than those of the upper waves, in contrast to the synoptic-scale waves having phase velocities almost independent of height. By a composite analysis it is shown that the amplitude of the upper waves is maximized at the 300hPa level and is confined near this level. Vertical tilting of the phase line of the upper wave is not observed. These results indicate that the appearance of the upper tropospheric medium-scale waves is strictly controlled by the jet both in vertical and horizontal directions.
During the intensive observing period (IOP) of TOGA-COARE, a significant westward-moving cloud cluster was observed in the equatorial Pacific from 10 to 13 November 1992. Although cloud activity in this region was less intense in November than that of other periods of the IOP, an eastward-propagating super cluster was still observed. The westward-moving cloud cluster was considered to be a hierarchic component of the super cluster. The cloud cluster moved westward against the prevailing westerly caused by a westward-propagating large-scale vortex and split into two clusters. One part of the cloud cluster continued to move westward and the other part was almost stationary in the zonal direction. In order to clarify mechanisms of the westward movement and splitting of the cloud cluster and its relationship with large-scale disturbances, a numerical simulation experiment and sensitivity tests were performed. The experiment simulated the westward movement and the splitting of the cloud cluster. The large-scale vortex intensified the westerly in the equatorial region. The cloud cluster developed where latent heat flux from the sea was enhanced by the strong westerly. Consequently, the westward-moving part of the cloud cluster moved with the westward-propagating vortex. The remaining part of the cloud cluster was associated with stationary twin cyclones and was, therefore, stationary in the zonal direction. On the basis of the data analyses and results of the numerical modelings, we conclude that the development and westward movement of the cloud cluster of the present paper were caused by the westward-propagating large-scale vortex and that the splitting of the cloud cluster is attributed to the different phase velocities of the westward-propagating vortex and the stationary twin cyclones.
This paper presents a study of the physical processes responsible for the summer monsoon onset over South China, which is defined using the rainfall data over Hong Kong. Four sets of meteorological parameters are examined: 1) mean-sea-level pressure; 2) surface and upper-level winds; 3) surface and the 850hPa temperatures; and 4) blackbody temperature (Tbb). The first three sets are from the gridded analyses of the European Centre for Medium-Range Weather Forecasts (ECMWF) for the period 1985-1990 while the Tbb data are from the Geostationary Meteorological Satellite-4 (GMS-4). Daily values of each parameter for the six years are then composited each day for the period ten days before to after the onset day. The results of the composite study suggest the following sequence of events. Ten days before the onset, a north-south temperature gradient between South China (SC) and northern Australia is first established as the temperature over the latter decreases. This temperature gradient is further enhanced as time goes on when the temperatures over SC and South China Sea (SCS) increase. Development of a surface low over Beibu Wan (around 107°E 18°N) is also observed four days before onset. This surface low deepens throughout the onset period. The enhanced temperature gradient and the deepening of the low apparently lead to an eastward shift of the cross-equatorial flow originally present around 60°-80°E. Together with another branch of this flow within 90° to 110°E, the low-level southerly winds over the SCS are therefore enhanced.This results in an increase in low-level convergence in the region of 100°-120°E, 15°-25°N which leads to enhanced rising motion. This is indicated by the enlargement of the 200-hPa anticyclone on Day (0). The northerlies on the eastern flank of this anticyclone then form a prominent return flow towards the equator. A monsoon trough is thus established along the SC coast and convection activity increases, as indicated by the Tbb data. This represents the onset of the summer monsoon over South China.
Split window measurements aboard the NOAA polar orbiting satellites have been used to study cloud cover change between day and night over the western Pacific. The split window technique can discriminate optically thin cirrus type clouds (ice clouds) from optically thick cumulus type clouds. In this study, cirrus and cumulus type clouds are each divided into two classes depending on cloud brightness temperatures (TBB). Cirrus are classified as either warm or cold type, while cumulus type are divided into cumulonimbus type and low-level cumulus/stratocumulus type. From the comparison with ISCCP analysis, mean optical thickness of warm cirrus, cold cirrus, cumulus/stratocumulus and cumulonimbus type clouds were found to be 2.2, 7.4, 15.3 and 33.7, respectively. The diurnal change in cloud cover of the above cloud types is studied for typhoon cases as individual convective systems and for an area of 20 degrees latitude by 30 degrees longitude over the western tropical Pacific. Cumulonimbus type clouds, warm cirrus type clouds and low-level cumulus/stratocumulus type clouds show tendencies toward higher cloud cover at night (about 2:30 local time). Cold cirrus type clouds show a tendency toward more cloud cover during the day (about 14:30 local time).
The size, motion, and structure of the rainbands in Typhoon 8913 are presented from observations using a long-range conventional radar and two Döppler radars. These rainbands were classified into wide slow-moving and narrow fast-moving. Wide (50-150km) rainbands have a small angular velocity (≤10°h-1) and a long lifetime (6-10h or more). Narrow (25-50km) rainbands are generated in a region more upwind and inward than wide rainbands every 3-4h. They rotate cyclonically around the typhoon center at a large angular velocity (30°h-1) during their short lifetime (1-4h), and catch up and merge with wide rainbands. One wide slow-moving rainband is accompanied with a disturbance (convergence, updraft, and high vorticity) inclining outward with a large slope (1/10) on its inside. One narrow fast-moving rainband is associated with a nearly upright disturbance. The structure of the disturbance differs, however, from the inertia-buoyancy waves proposed in some previous studies as the mechanism behind fast-moving rainbands. Both types of rainbands are characterized by stratiform precipitation in which low-level shallow convective cells are embedded, and make no cold pools in the boundary layer.
Polar cloud climatology is an urgent issue to be solved for the study of global climate from satellite data. However, detection of clouds in the polar regions involves many difficulties on account of the high albedo and low temperature of the snow and ice covered ground surface. Discrimination of clouds was done using AVHRR split window channel data. Brightness temperature differences of 11μm (ch 4) and 12μm (ch 5) were one of the indices of thin clouds; the correlation of the brightness temperature difference and the brightness temperature itself was used. Cloud amounts thus derived were well explained by the all-sky camera data measured at the ground. The cloud analysis was done from daily NOAA-9 data for thirteen months from January 1987 to January 1988, received and processed at Syowa Station, Antarctica. Spatial and temporal distribution characteristics of clouds over the East Antarctic continent are discussed. Annual mean cloud fraction ranged from 50% to less than 10% according to the region, being larger near the coast and smaller over the high interior plateau. Within the interior, cloud amounts were liable to be higher over the western slope facing the Weddell Sea compared to the eastern slope. Semiannual variations were seen in most of the area related to the behavior of the disturbances. In July to September, cloud amounts in the interior increased in some regions, and in November to January, increased in most regions; however, not much variation was seen in the coastal area. The analyzed area was divided into three regions of different variation characteristics, related to the topography. An oscillation of about 7 or 15 days in the time variation of clouds was noticeable in most regions. Comparing the brightness temperature for clear and cloudy sky, the radiative effect of clouds at the top of the atmosphere was found to be negative (cooling) in winter in the interior, and small positive (heating) in the longwave in summer months over the whole area.
The Communications Research Laboratory (CRL) has developed a simultaneous reception Ku-band dual-polarization radar. Simultaneous reception can give high accuracy of the differential reflectivity (ZDR) measurement from fewer averaged samples because of a good correlation between simultaneously sampled horizontal and vertical reflectivity. This paper pursues the possibility of estimating the rainfall rate using ZDR alone by comparison with a conventional polarimetric radar algorithm using ZDR and ZH on the assumption of the data handling of a CRL's dual-polarization radar. A computer simulation based on a physical raindrop model is used to pursue this possibility. The results show that estimating the rainfall rate using ZDR alone is not a good method.
Chaotic Lagrangian motion in a steady baroclinic wave, which is highly relevant to meteorological and oceanographical problems, was investigated by Sugata and Yoden (1994) by tracing a marked fluid particle for a long time in a numerical solution of their model. To test their result, we have conducted experiments on steady baroclinic waves in a differentially-heated rotating fluid annulus by tracking a 3-D trajectory of one neutrally buoyant tracer particle suspended in the fluid for a long time. We show here four trajectories that have been analyzed up to now. In the wavenumber-5 wave, the observed trajectories include the preferred routes of region transitions expected from the numerical investigation. This supports the Lagrangian view of the heat transport presented by Sugata and Yoden: The jet absorbs a large number of hot fluid particles from the outer boundary layer and releases them in the inner boundary layer, while the fluid particles nearly conserve their temperature in the meandering jet. Furthermore, it is of great interest that there was found one event of region transtion which does not take place in the numerical simulation. In the wavenumber-4 wave, however, the tracer particle was observed to remain trapped within the jet for a long time. This leads to the orthodox view of the heat transport: The jet absorbs a large amount of heat from the outer boundary and releases it into the inner boundary so that heating and cooling of the fluid particles take place during every cycle of the meander of the jet.