A linear perturbation analysis for the study of large-scale disturbances observed in the conditionally unstable atmosphere of low latitudes is attempted by use of a two-dimensional model in which latitudinal variations of wave properties are ignored. The effect of the conditionally unstable stratification is incorporated into the model with the assumption that the amount of released latent heat is proportional to the vertical velocity at the top of the surface boundary layer, as OOYAMA (1964) proposed in the research on tropical cyclones. Linear analysis shows that three kinds of unstable waves other than the tropical cyclone mode can exist in the conditionally unstable atmosphere in low latitudes. Two of them are unstable waves induced by the combined effect of heat and vertical wind shear of the zonal flow under the existence of surface friction. This type of unstable waves may be classified into two types with and without a steering level. The growth rate of the former increases with increasing vertical wind shear and the latitude, while that of the latter is maximum for moderate intensity of the shear and decreases with increasing latitude. The preferred horizontal scale of these waves is 2,000 to 4,000 km. Its structure is discussed in comparison with that of the easterly wave in the tropical troposphere. The preferred scale of the third wave is 2,000 to 12,000 km, depending upon the amount of heat released in the upper troposphere. This unstable wave can be obtained also in case of no vertical wind shear. In contrast to the other unstable waves this wave has large amplitudes in the lower stratosphere. Its phase lines have a large westward inclination with height in the stratosphere and an eastward inclination in the troposphere. This wave resembles the large-scale wave in the equatorial lower stratosphere discovered by YANAI and MARUYAMA (1966) when the amount of heat released in the upper troposphere is not so small. It is also shown that the vertical distribution of heat sources is of primary importance for the occurrence of these unsta ble waves.
Elevated-source dispersion is compared with simultaneous surfacesource dispersion. The analysis starts from the finding that the tracer cloud height generally varies with the downwind distance. Over a complicated terrain such as Kainan City the Pasquill stability of elevated-source dispersion is high and that of the corresponding surface-source dispersion is somewhat lower. Over a relatively flat terrain as Oita City the Pasquill stability for an elevated source is in general low or very low under weather conditions not much different from those at the Kainan City experiment. While the Pasquill stability for a surface source is definitely higher, which is reverse to the dispersion over Kainan City. The height of tracer cloud generally increases with the downwind distance, irrespectively of the elevated or the surface source. The rise of the tracer is more remarkable over a sloped terrain.
It is first examined whether the experimental results in a wind tunnel on the saltation of sand which have been published hitherto by several research workers will hold or not in field observation. Secondly, simple and practical methods to estimate a threshold fric- tion velocity for saltation in field observation are suggested.