We have performed one dimensional radiative transfer calculations to evaluate the impact of cirrus clouds on the tropical radiation budget. We investigate the sensitivity of solar and infrared fluxes to cloud optical depth, particle size distributions, and cloud height. If the observed solar cloud forcing in excess of 100 W-m-2 is to be attributed to cirrus anvils alone, then the optical depth of these anvils must be at least 5 (assuming 50% cloud cover and an ice crystal effective radius of 15 &mgr;m). The net radiative forcing of cirrus near the tropical tropopause is positive (heating) for cloud optical depths less than about 16 and negative (cooling) for larger optical depths. If cirrus clouds alone are responsible for the equal and opposite shortwave and longwave cloud forcing in excess of 100 W-m-2 observed by ERBE, then the cirrus must typically take the form of deep, optically thick clouds with relatively small particles (radii of 10--20 &mgr;m) and cloud-tops well below the tropopause.

The maintenance of this balance on monthly time scales can be attributed to a variety of correlations: The cloud cover of optically thick cirrus or thin cirrus overlying low-level stratus clouds could vary; or cirrus anvil height could increase along with a decrease in the ice crystal effective radius and an increase in optical depth. It would be of great interest to determine observationally which of these correlations is responsible for the observed lack of variation in cloud forcing. ¿ American Geophysical Union 1994