A close relationship between the variation of the quasi-stationary wave pattern and the transition of the early summer rainy season in Japan (the so-called Baiu) is described by using deviations of 5-day mean 500mb heights from zonal average in 1979 and nephanalyses based on GMS image. During the period from May to August the quasi-stationary troughs over East Asia and the western North Pacific intermittently moved westwards. Major cloud systems in the nephanalyses formed most frequently in the east side of the quasistationary trough, and the region of high frequency of appearance of the major cloud system moved west with the shift of the trough. The seasonal transition of the Baiu goes through three stages; the Baiu in the Ryukyus, the Baiu in the Japanese Main Islands and the end of Baiu. From the view point of the quasi-stationary upper flows, the transition between the stages is connected with the westward shifts of the quasi-stationary trough as follows: During the first stage the trough stayed along 140°E, and the rainy zone at the Ryukyus could not extend north beyond 30°N. The northward advance of the Baiu (transition from the first to the second stage) was a result of the westward shift of the zone of baroclinic activity associated with the movement of the quasi-stationary trough to 120°E. The cloud systems in the Baiu of the Japanese Main Islands were organized mainly by baroclinic disturbances in the east side of the trough at 120°E. In the final stage, the trough moved further west and weakened. The decay of the trough caused a diminution of cyclonic activity over Japan and brought about fair summer weather.
The features of the time variation of cloud amount with time scale longer than 15 days, in the western Pacific region (50°N to 46°S and 90°E to 172°W) were studied using 5-day averaged cloud amount data of Geostationary Meteorological Satellite (GMS I or II) and one-day averaged cloud amount data which were made from infrared images of GMS microfilms. Following conclusions were attained. Besides the seasonal variation which is dominant north of 40°N, in the zones of 20°N to 10°N and 5°S to 20°S, the time variation of cloud amount with period of 20 to 40 days takes place almost all over the western Pacific. Though this 30-day period variation does not continue to be dominant throughout the whole year, it is predominant for several months. The domain where cloud amount varies simultaneously with about 30-day period has the size of about 2000km in the E-W direction and 1000 to 1500km in the N-S direction. Neither the regular propagation of 30-day period variation in the E-W direction nor that in the N-S direction the propagation of it is clear. The variance or amplitude of 30-day period variation is the largest in the zone on the north or south of the equator where mean cloud amount is small, and these zones move in the N-S direction with season. Over the ocean south of Japan the maximum anomaly of 30-day period variation appears as the continuation of cloud amount near 10 for several days or the appearance of maximum cloud amount at intervals of 20 to 40 days.
The global-scale east-west circulation in the tropics associated with the Southern Oscillation (SO) was investigated by using two different data sets (FSU data and NMC data), which contain about 20 years of tropical wind field observations. As a predominant mode of the interannural time scale, the eastward-propagating anomalous divergent circulation (with a phase speed of 6°-10°longitude month -1) was found to be on the time scale of the SO (40-60 month period) with a zonal wavenumber-one structure. The eastward propagation is most prominent over the Indian Ocean toward the eastern Pacific. This mode shows a large amplification over the Indonesian maritime continent through the eastern Pacific, which may correspond with the modulation of the local Walker circulation. Another east-west oscillation with a similar zonal structure but a faster eastward propagation (12°-15°longitude month -1) was also noted with the quasi-biennial (QBO) time scale (20-30 month period). In addition, these two modes show a coupling (decoupling) of the anomalies which seems to be associated with the occurrence of the major (minor) El Nino events over the central through the eastern equatorial Pacific.
A multi-variables model of quasi-geostrophic barotropic flow over sinusoidal topography is constructed to investigate the bifurcation properties of the nonlinear system and to assess the availability of the low-order models by Charney and DeVore (1979) and Yoden (1985) on the same problem. When the horizontal diffusion by small-scale eddies is incorporated into the model (the Reynolds number is 103), multiplicity of stable steady solutions is obtained for a wide range of external parameters contrary to the cases of high Reynolds numbers (104 and ∞) . Periodic solutions are also obtained which branch off a steady solution by the Hopf bifurcation. These results support qualitatively some of those obtained in the low-order models, although some spurious properties due to the severe truncation are also recognized. Multiple flow equilibria first pointed out by Charney and DeVore (1979) could be demonstrated by laboratory experiments of rotating fluids.
Mesoscale mountain wave effects induced by the catastrophic breakdown in the lower stratosphere on the general circulation were testified by using Lindzen's (1981) parameterization scheme and Holton's (1982) quasi-one-dimensional model. The basic data obtained by Lilly and Kennedy (1973) were applied for the momentum flux radiated by the mountains. The results demonstrate that the mountain wave breakdown produce drag forces enough to predict the realistic weak wind profiles in the lower winter stratosphere. In addition, stronger meridional flows than expected so far was obtained within thin turbulence layers less than a few kilometers. More detailed simulations for the general circulation must be desired to know what is the main driving force to maintain the BrewerDobson circulation.
An analysis was made of the structure and behaviour of small-scale motions in the stratosphere and lower mesosphere with the aid of meteorological rocket observations over the period of six years from 1977 to 1982, covering the wide range of latitudes. By applying a filter to observed wind data with respect to height, wind fluctuations with characteristic vertical scales close to 10km are separated from large-scale components such as the mean field, planetary waves and tides. From the hodograph analysis it is found that at northern hemisphere stations most of horizontal wind vectors show the clockwise rotation with increasing height while they rotate anti-clockwise in the southern hemisphere. This strongly suggests that the wind fluctuations are due mainly to upward propagating inertia-gravity waves. On the basis of a simplified theory of inertia-gravity waves, the wave-frequency distribution is estimated statistically from the degree of elliptic polarization of holographs, and it is shown that the most predominant values of f/ω (f; the Coriolis parameter, ω; the intrinsic wave frequency) fall into a range of 0.2-0.4. Namely, the typical time scale of these waves is of the order of several hours in middle and high latitudes and of a day in low latitudes. Further discussions are made of the vertical profile of the wave energy density, and it is suggested, from the uniform decay of the wave amplitude with height that the wide spectra of horizontal phase velocities should be taken into account.
Two methods for deriving space-time spectra from asynoptic satellite data, the frequency transform method (Hayashi, 1983a) and the asynoptic space-time transform method (Salby, 1982a) are shown to be equivalent. That is they give identical spectra and have identical aliasing characteristics. When asynoptic sampling conditions are satisfied, this linear transformation is unique and recovers the correct spectrum over wavenumber and frequency for a satellite observed field. Moreover the synoptic mapping of asynoptic data, as implied by this operation, is the only linear transformation capable of extracting the true synoptic behavior from asynoptic measurements. Other procedures used to derive synoptic behavior must reduce to this transformation if they are to consistently recover the correct structure and evolution.
The intensity distribution of diffuse solar radiation with the zenith angle under overcast sky with stratified clouds (hereafter we call merely angular distribution) were measured at wavelengths of 400, 500, 600, 700 and 750nm. The detailed analysis of their features was also carried out with the aid of numerical computations. The main results obtained are as follows. The angular distribution in the case of overcast sky has a maximum at the zenith and decreases from the zenith to the horizon in contrast with the case of clear sky in which the angular distribution increases from the zenith to the horizon except for the surroundings of the sun. And this feature of the angular distribution is almost constant beyond a critical thickness of cloud (about 150m thick) and does not almost depend on the cloud thickness, the cloud species, the wavelength, the solar zenith angle, the cloud base height and so on but depends much on the albedo of ground surface. This suggests the possibility of obtaining the surface albedo by measurement of the angular distribution under overcast sky. From the measurements and the numerical analysis of the angular distribution in this study, it becomes clear that the conventional empirical equation expresses the angular distribution well.
Spectral distributions of diffuse solar radiation were measured at wavelengths of 400 to 800nm under both clear and overcast skies. Under clear sky the form of the spectral distribution varies with the atmospheric turbidity. On the other hand, under overcast sky, it is almost uniform. The relative spectral distributions to the solar spectral irradiance at the top of the atmosphere are also analyzed and it was found that the relative spectral distribution under the overcast sky is almost constant for the wavelengths of 400 to 800nm except for the spectral regions of absorption bands due to gases such as water vapor and oxygen molecules. The relationships between absorptivity of the absorption band and cloud thickness are obtained by the measurements, and they are confirmed by theoretical calculations. Those are; for 762nm oxygen absorption band, the results of the calculation have a good coincidence with the measurement ones so that the amplitude of absorption band have the linear relation to the cloud thickness; for 720-730nm water vapor absorption band, the amplitude depends much on cloud thickness, cloud height, atmospheric temperature profile and humidity profile.
Strong attenuation of millimeter wave due to precipitation particles is the most serious problem when we use K-band (0.86cm wavelength) radar for the observation of the precipitation. In this paper, we calculated the attenuation of K-, X- (3.2cm) and C-band (5.6cm) waves based on a model of precipitation, for three precipitation intensities, i, e., heavy (32mm/hr), ordinate (4mm/hr) and weak (0.5mm/hr) precipitations. And, we surveyed any possibility of the observation with weak affection of the attenuation using the bands. The observation of the precipitation was executed in Part II. The results show that the attenuation coefficient is larger in the melting layer by one order or two orders than in the lower raindrop region. The results also indicate that the affection of the size distribution to the attenuation abruptly increases as the attenuation is enhanced. The attenuations of X- and C-band waves are small enough to be neglected except for the heavy precipitation. In the case of heavy precipitation, K-band wave is severely attenuated, especially in the melting layer, so that the signals cannot be detected by the radar receiver. In the cases of ordinary and weak precipitations, however, we can obtain true radar reflectivity factor of K-band wave by corrections of the attenuation. In addition, the correction method using the radar reflectivity factor of the X-band radar was investigated. We can mention conclusively that after the correction the precipitations weaker than about 4-7mm/hr can be safely observed within an error of 2dB by the twowavelength radar technique using K- and X-band radars in the raindrop region. In the melting layer, the correction method can be utilized without large error when the precipitation is weaker than about 2-4mm/hr.
Whether the size distribution change due to coalescence or breakup of melting snowflakes and raindrops occur significantly in the lower-half region of the melting layer and the raindrop region is investigated by the two wavelength radar observation together with model calculation. This observational technique is based on a fact that response of radar reflectivity factor to size distribution transition between two altitudes is dependent on theradar wavelength. We use two radars with wavelength of 3.2cm (X-band) and 0.86cm (K-band) since the two wavelengths give a significant difference of the radar reflectivity factor within a size range of observing particles. The comparison of the observed and calculated results reveals a strong evidence of particle size distribution transition equivalent to coalescence in the lower-half region of the melting layer for stably uniform precipitations, i.e., typical stratiform precipitation. The activity of the distribution transition does not strongly depend on the precipitation intensity. For the same type of the precipitation, when the precipitation rate is relatively large, increase of medium-sized particles (0.1-0.25cm in diameter) is likely to be predominant in the raindrop region just below the melting layer. Any distinguishable evidences cannot be found for less-uniform precipitations with weak cellular structures.
The collection efficiency of charged natural snow crystals (carrying an average charge of 0.4 e. s. u. cm-2) for aerosol particles of 0.1μm to 6μm in diameter was experimentally examined under conditions where phoretic forces and external electric fields were negligibly small. From a comparison between the collection efficiency of charged snow crystals (the present experiments) and that of uncharged snow crystals (Murakami et al., 1985a), the effects of snow crystal charge on the collection efficiency were estimated. The enhancement of collection efficiency due to electrostatic charges on snow crystals increased with the decreasing aerosol size, and became maximum at around 0.2μm in diameter. Coulombic force was considered to be the dominant electrostatic force under the conditions of the present experiments. On the basis of the present experimental results and the previous results obtained by Murakami et al. (1985a, b), the semi-empirical expression for the collection efficiency of plane-type snow crystals was derived.
Surface ozone concentrations measured for nine months in the Harima district were analysed to find a relationship with some meteorological factors. Emphasis is put on the high concentrations of ozone in the afternoon in late spring to summer. In general, the daily maximum ozone concentration is proportional to the integrated insolation. The day-to-day fluctuation in the daily ozone maxima is mostly caused by the various patterns of the winds. Dependence of the particularly high ozone values on the location of the sites can be explained by the effect of advection due to land and sea breezes.
Climatic noise is an important concept, because it is a measure of statistical sampling error in time mean and of potential predictability of climate. However, reliable method of its estimate has not yet been established mainly because of difficulties in separation of climatic signals from noise. Noticing year-to-year fluctuations of intra-month variances, an estimate of the noise is attempted in the present paper, under the following three assumptions: i) anomaly of daily weather can be expressed by sum of anomalies of changes due to internal dynamics (n) and those due to external or boundary forcing (s) ; ii) intra-month variance may be equal to sum of squares of deviations of (n) and (s), iii) effect of signal on noise estimate may vanish when intra-month variance is the smallest in 30-year time series. Year-to-year fluctuations of intra-month variances are confirmed by analysis of 30-year time series of daily mean surface air temperature at 19 stations in Hokkaido. And no appreciable influence of the signal on autocorrelation could be found in the data analysis. Estimate of the noise is made by utilizing autocorrelation of autoregressive model of the second order and the smallest intra-month variance during 30 years. The noise thus estimated is 0.3-0.8°C, and clearly less than noise by method of other authors. Comparison with the observational error shows that the noise is greater than observational error, and implies that consideration of the noise alone is sufficient for ambiguity of time mean. It is noted that the noise of spatial mean over an area (ca.400km×400km) has no appreciable difference from that of individual stations, and for this fact an interpretation is given.