The nonlinear behaviors of Eady-type baroclinic waves forced quasi-resonantly by topography are investigated. The complex amplitude equation which is a Landau equation with forcing terms is derived by using the perturbation method; the effect of viscosity is taken into account in the leading order as the Ekman layers at the top and the bottom of the fluid layer. For a given wave number and a basic zonal flow with a constant vertical shear and with zero vertical mean, the resonance due to the topographical forcing occurs just at the critical thermal Rossby number (β=β0) of baroclinic instability. However, when the wave self-interaction is taken into account, the maximum response occurs at a value smaller than β0. Three equilibrium solutions, say A, B, and C, are found near the maximum response; the largest amplitude state, A, is stable, while B, the smallest amplitude, and C with amplitude almost equal to A are both unstable. It is shown that B is unstable for amplitude and phase perturbation, while C is unstable for phase perturbation only. It is noted that C is identical with A except phase difference 180° and stable if the bottom is flat. In addition, by plotting flow vectors composed of the time change of amplitude and phase on amplitude-phase plane, it is suggested that the potential surface characterizing the stabilities would have a minimum around state A, a maximum around B and a saddle point around C. At a slightly off-resonant condition of the basic zonal flow, too, we can have three equilibrium solutions for a certain narrow range of β. The stabilities of these solutions are similar to those obtained under the resonant condition. In addition, it is shown by the numerical integration of the complex amplitude equation that baroclinic waves propagate, interacting with the forced waves when the height of topography is small, and hence the amplitude and phase velocity oscillate with time. When the height is large, the phasee velocity becomes small, and if it exceeds some critical value, the baroclinic waves are trapped by the topography.
The properties of convective clouds and CISK (instability of the large-scale disturbance interacting with the clouds) under vertical shear flow are studied by numerical experiments as an extension of previous studies (Yamasaki, 1975, 79, 83), which have used a fine resolution model with explicit calculations of convective clouds. The CISK treated in this study is such that the downdraft and cooling due to the evaporation of falling rain play an essential role. The primary objectives are to understand how this type of CISK is modified by vertical shear flow and to show how it can account for some observed phenomena in the tropical easterlies. According to numerical experiments, when the intensity of an easterly flow decreases with increasing height (quasi-linear flow with westerly shear), convection in the eastern portion of the convective area is more preferred than that in the western portion. When the easterly flow is strongest at some height such as 3km (jet-type flow), convection in the western portion is preferred. These results indicate that vertical shear in the lowest few kilometers is important and that convection is more enhanced when low-level inflow into the convection is augmented by vertical shear flow. Individual convective clouds in the absence of vertical shear are organized into a larger convective system (referred to as 'mesoscale convection') with a time scale of about 3 hours (Yamasaki, 1983). It is shown in this study that this time scale is not much modified by the quasi-linear flow (L) and that it is modified to about 8 hours for the jet-type flow (J). In the cases of no shear and (J), convective clouds grow successively from a low-level cloud which persists continuously, whereas in the case of (L), new clouds tend to form at some distance from existing clouds, constituting mesoscale convection in all cases. The properties of a large-scale disturbance obtained for the flow (L) are similar to those of the easterly wave of the type studied by Riehl (1954) and Yanai (1961) ; that is, a coldcored structure at low levels, an eastward tilt of a trough, major convective activity located to the east of the trough and an upper-level ridge located further to the east are well simulated. It is also shown that vertical shear is not indispensable to the growth of the disturbance, thought it is very important in explaining observed features of the structure. The western convection which is preferred in the flow (J) exhibits several important features in observed squall-lines studied by Zipser (1977) and others. The simulated structure is also similar to that obtained in numerical experiments for an isolated cloud (e. g., Moncrieff and Miller, 1976). However, in the present model, convective clouds form successively from persistent low-level clouds. Such successively formed clouds constitute mesoscale convection, the lifetime of which is roughly in agreement with observations.
The sea and land breezes in the northern region of Kyushu Island in Japan under the condition of no synoptic wind in typical summer season are simulated by a three dimensional numerical model. This study provides an evaluation of the diversification of the local circulation owing to the topography, e. g. irregular seashore lines and mountainous topographies. In order to investigate the influence of mountainous topography on the local circulation in Kyushu, we compared the case of the realistic ground elevation with that of the flat in the northwestern region of Kyushu which is characterized by the irregular coastal lines, the comparative flat topography and dispersive mountains. Further we also compared them with the case of the realistic topography in the north-eastern region of Kyushu which has the monotonus seashores and the spread mountain areas. In the northern region of Kyushu, the irregular coastal lines (esp. topography of peninsula) produce characteristic local circulations and variation of winds in connection with mountainous topographies. The mountainous topographical effects are summarized; (1) the upward vertical winds concentrate above the mountain area in the day time (down slope wind in the night), (2) the horizontal winds are oriented by the valley (e. g. in the sea area surrounded by a mountainous topography) and (3) the winds are enforced in the area of converged mountainous topography. The sea breezes in the small islands near the Kyushu are strongly affected by that in Kyushu. Further the penetration of sea breeze to inland or other area is effective factor on the diurnal wind variation at any place (esp, in the western region).
The nature of the Baiu frontal zone (a stationary precipitation zone appearing in June ∼July around the Japan Islands) is examined in comparison with polar frontal zone in the northern hemisphere for 19∼26 June 1975. The polar frontal zone defined as the maximum zone of time variation of potential temperature and vorticity and the maximum zone of potential temperature gradient is located in 50∼60°N. Such characteristics of the polar frontal zone are not found for the Baiu frontal zone located along ∼30°N in 110∼150°E. The Baiu frontal zone is characterized by a narrow steady precipitation zone, strong gradient of equivalent potential temperature, thick moist neutral layer, steady generation of convective instability. It is concluded that the Baiu frontal zone is not a polar frontal zone but a significant subtropical front in the East Asia. It is infered that the southwestward extension of the Pacific anticyclone into the region of very moist tropical maritime air-mass is the primal condition for the formation of the Baiu front, because the frontogenesis in equivalent potential temperature field, generation of convective instability and the strong moisture convergence in the Baiu frontal zone are owing to the SW wind in the tropical maritime airmass in the western periphery zone of the westward protruding Pacific anticyclone.
The nature of disturbance activity during the summer monsoon over East Asia was investigated using FGGE Level IIIb data from 1 May to 31 July 1979. This 92-day period was divided into two 46-day subperiods, i. e., Period I(1 May-15 June), and Period II (16 June31 July). To facilitate further study, rainfall variations over the Yangtze River basin were taken as a reference. The designation day "0" was used for days when rainfall peaks occurred over the Yangtze River basin. By this definition it is only natural to find a prominent cyclone located over central China on day 0, which is referred to as the "Period I China cyclone" and the "Period II China cyclone", for each subperiod, respectively. The northeastern edge of the Tibetan Plateau is the birthplace of Period I China cyclone. A small-scale low-level cyclone trapped below 700mb is induced by the mechanical (not thermal) effect of the Tibetan Plateau on day -2. This edge cyclone then propagates south- ward along the eastern periphery, reaching the upper reaches of the Yangtze River by day 0. Between days +1 and +2, the low-level edge cyclone merges with an eastward propagating upper trough. Associated with this merged cyclone is widespread rainfall over central China and Japan. The central Tibetan Plateau is the birthplace of the Period II China cyclone. At the initial stage on day -2, its character resembles the soalled "Tibetan vortex" with a vertical depth of 1km or less. It propagates eastward across the Tibetan Highland, eventually merging with a major mid-latitude cyclone in the vicinity of the Yellow River basin between days 0 and +1. At this time, the merged cyclone becomes coherent in the vertical and extends as high as 250mb. Thus, the Period II China cyclone experiences a dramatic change in its vertical structure from the incipient stage over the Tibetan Highland to the mature stage over the Yellow River basin.
Dominant SST anomaly patterns in the North Pacific were elucidated by the EOF analysis. Derived first spatial pattern corresponds to the Pacific/North American (PNA), while the second coincides with the Western Pacific (WP) teleconnection pattern in Northern Winter. The first dominant spatial pattern persists for several months, or shifts to the second component three months later. Negative time coefficients of the the second component in February are connected with the warm Hokkaido and cold West Japan. The low or high index in the middle latitude circulation is necessary to explain shifts in the SST anomaly pattern.
Surface Observation of sea fog was carried out in and around Kushiro City on the south-eastern coast of Hokkaido from July 14 to July 20, 1981 and from July 26 to August 4, 1982. The relation between water content and visibility at the observation points in the urban area and in the coastal suburbs was examined with the data gathered from the continuous measurement of water content and visibility. The followings were clarified by analysing the equation of the relation between water content and visibility. The first, the visibility in the urban area was higher than that in the coastal suburbs at a certain value of water content. And the second, the increaseing rate of visibility accompanied by a decrease in water content in the urban area was larger than that in the coastal suburbs. These characteristics are interpreated as follows: 1) Small droplets of sea fog are dissipated in the urban area and 2) Mean radius of the distribution of the size of the fog droplets varied with the water content in the coastal suburbs.
Synoptic data from progressive tornado outbreak days are examined using an Empirical Orthogonical Function (EOF) expansion. "Composite" tornado outbreak charts are created. A methodology for using EOF correlations for statistical forecasting is presented.