This figure shows the distribution of remote correlation coefficient between the forcing index of various levels of clouds and sea surface temperature (SST) averages over an area in the eastern equatorial Pacific. From top to bottom are for total cloud, high cirrus, middle cloud, and low cloud. The cloud indexes are derived from ISCCP data and SST is from blended AVHRR data. The top figure is almost identical with the correlation between SST and surface solar flux, showing you that surface solar flux is very much governed by the total clouds produced by ISCCP. The correlation between high cloud and SST are, to some extent, opposite to the correlation between low cloud and SST. The correlation between total cloud and SST appears to be a sum of the effects of various type of clouds. A good example is in the eastern Pacific just north the equator. In this area, SST anomalies has a strong negative correlation with the high cloud and a strong positive correlation with low clouds; the result is a weak correlation with the total cloud forcing. Detailed discussion is given by Fu, Liu, and Dickinson (J. Climate, 9, 616-634, 1996).