Abstract
The GISS global general circulation model has been used to simulate July conditions, in a manner analogous to the previously described January simulation. Sea surface temperatures, ice cover, snow line and soil moisture were assigned values based on climatological data for July, and the integration was started from real data for 18 June 1973. Because of the realistic initial condition, the model rapidly approached a quasi-steady state. Mean statistics were computed for the simulated calendar month of July, and compared with climatological data, mainly for the Northern Hemisphere troposphere. Qualitatively, the model-generated energy cycle, distributions of winds, temperature, humidity and pressure, dynamical transports, diabatic heating, evaporation, precipitation and cloud cover are all realistic. Quantitatively, the July simulation, like the January simulation, tends to underestimate the strength of the mean meridional circulations, the eddy activity and some of the associated transports. The July simulation of zonal mean temperature and zonal wind fields is superior to the January simulation in the Northern Hemisphere because of the absence of the polar night jet, and the decreased importance of large-scale dynamical heating and cooling in summer.
In order to assess the model's ability to simulate seasonal differences, the July and January simulations were compared with each other and with climatological data on seasonal changes. The model simulates accurately the northward displacement of the mid-latitude jets, the low-latitude Hadley cells, the tropical rain belt, the trade winds, and the ITCZ in July compared to January, the reversal of the Indian monsoon, and the weakening of the zonal meridional circulations and the decline of eddy activity in the summer. The simulated seasonal differences in the Southern Hemisphere are much less pronounced than in the Northern Hemisphere as expected.
From a climatological point of view, there are three particular aspects of the model's simulations that need to be improved: 1) arctic regions in January are as much as 10°K too cold, because of the model's underestimate of the dynamical transports of heat into high latitudes; 2) the simulation of the climatological fields in the vicinity of the Himalayas and Southeast Asia is noticeably poorer than in other areas—for example, in Southeast Asia in the July simulation the rainfall is half the observed amount; and 3) the global albedo in July is too high when compared to satellite-derived values (0.35 vs 0.26), at least partially because the model-simulated deep, penetrating cumulus clouds occur too frequently in July.
