A numerical simulation study of the influence of surface albedo on July circulation in semi-arid regions, using the general circulation model (GCM) of the Goddard Laboratory of Atmospheric Sciences (GLAS) is presented. The results are based on two 47-day integrations. In the first integration, called the control run, the surface albedo was normally prescribed, whereas in the second integration, called the anomaly run, the surface albedo was modified in four regions: the Sahel in Africa, the western Great Plains in the United States, the Thar Desert border in the Indian subcontinent, and northeast Brazil in South America. This experiment is similar to that of Charney et al. (1977); however, it was performed with the GLAS model with vastly different boundary forcings and several changes in the physical parameterizations.
Each run was started from observed initial conditions for 15 June 1979, based on NMC analysis. An analysis of the two simulations shows that in the Sahel, and the Thar Desert border regions the current results again show reduced precipitation with increased surface albedo in accordance with Charney et al. (1977) and Charney (1975). The semi-arid northeast Brazil region, which had a winter circulation, also conforms with Charney's (1975) hypothesis. However, the Great Plains region was an exception. There the total precipitation in two simulations was unchanged, as was the total cloudiness. In this region, the convective precipitation was reduced somewhat, while the large-scale precipitation compensated this decrease by approximately the same amount. The lack of albedo impact on precipitation in the Great Plains region could be due to the influence of the Rocky Mountains in generating variations in the large-scale flow.
The purpose of repeating Charney's albedo experiment was to test the sensitivity of his earlier results with the current GLAS GCM which has substantially modified physical parameterizations, particularly the planetary boundary layer (PBL), crucially important for such impact studies. Taking into account that both the GCM and the initial conditions of the atmosphere were different, it is worth noting that this experiment still provides support for Charney's hypothesis (1975) regarding the influence of surface albedo on mean-monthly climatology in the subtropical desert margin regions.
Another important influence noted was the variability of mean monthly simulation in areas far away from the albedo anomaly regions. Because the differences occur in regions of high observed climatological variability, it is necessary to separate the contribution of albedo anomalies as opposed to the model variability. The influence of surface albedo changes to produce large changes in the mean montly circulation elsewhere is very intriguing and needs further investigation.