Spatial and temporal dependence of clouds and their radiative impacts on the large-scale vertical velocity profile

[1] The structure of tropical large-scale vertical velocity from the European Centre for Medium-Range Weather Forecasts Re-Analysis is compared with simultaneous satellite measurements of precipitation, top of atmosphere radiation, and clouds from the Tropical Rainfall Measuring Mission (TRMM) on timescales ranging from hours to months. The first two empirical orthogonal functions of the vertical velocity profile represent the traditional deep circulation (PC1) and a shallower circulation (PC2) associated with middle-level divergence. Together they explain 90% of total variance and can distinguish two types of upward and downward motion: ‘‘top heavy’’ and ‘‘bottom heavy.’’ Cloud and radiation budget quantities measured on TRMM have coherent relationships to PC1 and PC2 on all timescales from simultaneous to long-term means. The relative importance of PC2 is greater on short temporal and small spatial scales. ‘‘Top heavy’’ ascent is associated with deep cloud systems, more intense precipitation, lower outgoing long-wave radiation, stronger cloud long-wave forcing, and extensive anvils. Cloud short-wave forcing depends primarily on PC1, while the net cloud forcing depends more on PC2. High-thin clouds are less correlated with short-term variations of the vertical velocity. Shallow precipitation measured by TRMM precipitation radar is associated with ‘‘bottom heavy’’ upward motion. Temporal compositing with respect to intense precipitating events shows that strong upward motions tend to develop in the lower atmosphere first and then change to the more ‘‘top heavy’’ type of upward motion. The associated cloud systems show consistent temporal changes in which high-thick clouds develop first and extensive anvil clouds develop later. These results suggest that the elevated latent heating from stratiform precipitation and the development of ‘‘top heavy’’ upward motion profiles in the tropics are related to each other. The coherent relationships shown here between large-scale vertical velocity and independently measured cloud and precipitation data can be used to test the performance of climate models.