Recent studies have documented a large decadal-scale downward trend in the International Satellite Cloud Climatology Project (ISCCP) global mean cloud amount data. This trend would have far reaching impacts on the climate for several reasons, including changes in the top-of-atmosphere radiation budget and variations in the tropical Walker and Hadley circulations. Furthermore, given that the coupling between clouds and the general circulation of the atmosphere is an important part of any attempt to model the global climate, it is important for us to understand this trend and be able to reproduce it. However, there is some debate on whether or not this trend is physical. Furthermore, previous studies have not demonstrated mechanisms that might drive such change in the cloud amount data, either on the local or global scales. For these reasons, this talk will present evidence that the trend in the cloud amount data is both robust and real, especially in the tropics.

Two global cloud amount datasets are used to create maps of the mean cloud amount trend, both on a 2.5° x 2.5° grid and in the zonal mean. Then, having removed both the linear trend and the seasonal cycle from the data, the time series of the variability of global cloud amount are projected onto maps of the variability of reanalysis fields such as 200 mb zonal wind, 500 mb geopotential height, and 200 mb divergence. From these projections, areas of significant statistical correlation provide corroborating evidence of the physical bases for the global and regional cloud amount trends. Using multiple sources of data to repeat the analyses, each analysis consistently produces similar spatial patterns of correlation. Thus, the cloud amount trends can be accepted with some confidence. Furthermore, using the patterns of correlation with the various dynamical fields, we speculate as to how the dynamical fields might be coupled with the clouds, which could allow us to better model these decadal-scale changes in the climate system.