The performance of the July simulation conducted with a five-level atmospheric general circulation model of the Meteorological Research Institute (MRI·GCM-I) is described. The model is the same as was described in Tokioka et al. (1985). The model simulates with reasonable accuracy most of the observed features of the large-scale distributions of sea-level pressure, circulation, divergent wind system and precipitation, although there are some systematic differences in relatively high latitudes in the Northern Hemisphere. Both evaporation and precipitation over the northern Eurasian Continent and Canada are excessive over the observed ones, and a false low pressure belt is found around 70°N. The followings are consistent with this deficiency: wet ground condition, persisting moisture flow from adjacent oceans, preferred condition to generate shallow cumulus convection and less stable stratification than the observed one. In the Southern Hemisphere the model successfully simulates subtropical highs, the circumantarctic low-pressure zone in sea-level pressure and the double-jet structure in the mid-tropospheric westerly wind. The velocity potential and the resultant divergent wind fields clearly demonstrate two major circulation systems in the tropics — the east-west circulation between the western Pacific Ocean and the eastern Pacific Ocean, and the northeast-southwest oriented circulation over the Indian Ocean. Characteristic features of the model-generated diabatic heating field are also discussed.
A two-dimensional numerical model which contains mountain and sea surface has been developed mainly to study the influence of cumulus convection and radiation on the local circulation. Numerical experiments were carried out for four cases. Case 1 is the basic one which contains only the computation of net radiation on the ground surface to predict ground surface temperature. Cooling or warming due to radiation in the atmosphere and the effects of cloud are not taken into consideration. Net radiative fluxes at each level including the influence of cloud are estimated in Case 2, but the process of cumulus convection is not included in this case. Case 3 is the same as Case l but the cumulus convection can be treated. Net radiation at each level containing cloud effects and cumulus convection are taken into consideration in Case 4. The experimental results show that clouds exert great influence on ground surface temperatures and the local circulation itself through the processes of condensation and radiation. The computation of net radiation in the atmosphere is very important in simulating the surface inversion layer and the thin downslope wind on the slope at nighttime. The formation of clouds in the daytime suppresses the circulation due to rapid depression of net solar radiation on the ground. However, it enhances vertical velocity over the mountain through the cumulus convection. In this case, the anti-slope wind seems to be weakened. Namely clouds have effects which oppose each other on the circulation. The high temperature in a cloud as compared with its environs and lower one at the upper adjacent layer can be simulated if both effects are taken into consideration. In addition, clouds delay the onset of downslope wind if they remain after sunset.
A method of deriving the sea surface temperature (SST) from space is described by using the infrared channels of NOAA-AVHRR radiometer with reference to a model atmosphere-ocean system. It was found that when free of stratospheric aerosols the combined use of channels 3 (3.7 μm), 4 (11 μm) and 5 (12 μm) is effective for the SST derivation for a moderate amount of precipitable water. However, in the case of a large amount of water vapor, its vertical profile has to be simultaneously determined to correct the water vapor effect. Furthermore, to evaluate the effect of the stratospheric aerosols on SST, the visible channels are also utilized for the atmospheric correction, where the radiation from the atmosphere is affected more by the presence of stratospheric aerosols.