The impact of satellite wind data on the analysis and forecast is examined by applying the normal mode theory. It is demonstrated that the impact of satellite wind data on the analysis appears largely in the external Rossby modes with low meridional index, especially 5-day wave. Significant impact on the forecast is found in the mid-latitude in the northern hemisphere, part of which is the propagation of the initial impact on the Rossby mode. This impact may be classified as that of the planetary scale and that of the synoptic scale. The former is related to the tropical data effect (Daley et al., 1981), and can be enhanced by applying the appropriate broad vertical prediction error correlation. The latter, on the other hand, may be related to the baroclinic instability, and be dependent on the vertical shear of basic current. The shear is the difference of the wind between upper and lower level, and can be modified by the insertion of the satellite wind data. Finally, it is concluded that, in order to make the maximum use of satellite wind data in the data assimilation cycle, the estimation of the vertical prediction error correlation of the model and the reliability of satellite wind data is essential.
A numerical prediction experiment using an operational Northern Hemispheric prediction Model (8L-NHM) at Japan Meteorological Agency was performed on an upper level cyclone observed in the upper troposphere around 20°N of western North Pacific late July in 1979. Low and high level cloud motion vectors derived from GMS (Geostationary Meteorological Satellite) pictures were the major data sources for the initial analyses. Four forecast experiments have been performed using 00Z analyses of July 20 to 23 1979. In most cases, the westward motion of the upper level cyclone was fairly well predicted, but its intensity was underestimated. One of the reasons of the poor prediction of the intensities was considered to be due to the defect of the formulation of the diabatic heating process in the model. Despite the deficiency, the prediction seems to be useful for the operational forecast of the upper level cyclone in the Western Pacific. In order to investigate the mechanism of the westward movement of the cyclone, a method to diagnose the phase velocity of a vortex from the vorticity budget equation was developed. The application of the method revealed that the westward movement could be mostly explained by the advection of vorticity by mean flow. The divergence effect may become more important when the cyclone is in the stage of development or decay.
Hitherto, many theoretical and statistical studies on the movement of typhoons have been proposed and it has been recognized that the typhoon moves with a steering current. Recently some quantitative prediction methods for the movement of the storm have been devised and the position of typhoon has been determined accurately by the data from the meteorological satellite. In this study, the movement of the storms is investigated statistically over the Far East area from the viewpoint of synoptic meteorology. Some interesting results are obtained as follows: (1) Though the storms move with the steering current on the whole, the decaying storms deflect to the higher pressure side from the basic flow and the developing storms to the lower pressure side. (2) The vertical wind shear of the basic flow changes from the lower to upper troposphere for the developing cases, but there is no change of the vertical wind shear for the decaying cases.
In the winter season of 1976-77, the second coldest temperature of -40.8°C in Japan was recorded in a basin in Hokkaido island in that season. Three examples of extremely low temperature in Hokkaido were analyzed in mesoscale. As a result it was found that the pattern of horizontal temperature distribution was common to all the cases except cloudy region, although the absolute values were different. Taking into account of cloud amount, altitude and distance from the seashore, the horizontal distribution of extremely low temperature which observed in Hokkaido was well simulated theoretically by calculating the vertical heat transfer near the snow surface. However the simulation for small land basins or narrow valleys was not good. This may be because of unexact estimation of eddy diffusivity which depends on a local wind speed. It was shown that the lower the wind speed the*lower the air temperature immediately near the snow surface, but the temperature at the height of regular screens is the coldest when the wind speed at 9m height is around 1m/s.
The modification process of convective snow-clouds in landing the Japan Sea coastal region was observed quantitatively by a RHI radar which was set at Wajima in the Noto Peninsula. The analysis of observational data-was made mainly for snow-clouds in two cases of A and D-in case A the outbreak of cold air became intensive with time after the passage of a low pressure and in case D it was predominant persistently. In both cases most of convective radar-echoes which travelled from the sea showed two stages of variation of echo structure in their landing. In the first stage the echo becomes intensive in its center before landing. Its most intensive part descends several kilometers offshore and the other part found out in its frontal side extends upward with time. The latter part descends inland rapidly after having crossed over the seashore. In the second stage seen after landing the intensity of the echo increases gradually again as a diffused broad echo. In the first stage graupels were observed on the ground predominantly and in the second stage snowflakes were observed as predominant particles. It was suggested from the comparison of echo intensity over the sea with that over the land that the amount of snowfall seen in the seaside of the peninsula was larger than that in the land side in case D, while it was smaller than that in the land side in case A.
The effect of the size distribution and chemical composition of cloud condensation nuclei (CCN) on the size distribution of cloud droplets was studied by a numerical model of an adiabatically ascending air-parcel. Specially the formation of a broad droplet size distribution was discussed. Results of computation show that the supersaturation realized in the air-parcel is reduced, a broader size distribution of cloud droplets is produced. If the number concentration of CCN is higher in one air-parcel than that in the other air-parcel, the broader size distribution of droplets is produced in the former air-parcel. Both the addition of anthropogenic CCN in the range of Aitken size and the slower ascent of the air-parcel also produce the broader size distribution of droplets. The growth rate of a droplet formed on a small nucleus is sensitive to the change in supersaturation because of small molarity of the solution droplet, while a droplet formed on a large nucleus grows insensitively to the change in supersaturation because of large molarity. Therefore the droplet size distribution in the air-parcel of lower supersaturation is broader than that in the air-parcel of higher supersaturation. The size distribution of cloud droplets formed on nuclei in which water-insoluble matter is internally mixed is narrower than that of cloud droplets on non-mixed nuclei, if the size distribution of mixed nuclei is the same as that of non-mixed nuclei. But the size distribution of droplets formed on mixed nuclei is hardly different from that associated with non-mixed nuclei, if the mass distribution of soluble matter of mixed nuclei is the same as that of non-mixed nuclei.
The paper describes a laboratory study to investigate charge deposition by propagating positive streamers on ice particles. A variety of ice samples similar to those present in natural thundercloud was used in this study. It was found that the deposited charge increased with the ambient electric field showing an enhancement in the streamer tip charge. Some possible implications of the results to thundercloud are discussed.