SEASONAL-VARIATION OF SURFACE AND ATMOSPHERIC CLOUD RADIATIVE FORCINGOVER THE GLOBE DERIVED FROM SATELLITE DATA

Global distributions of surface and atmospheric cloud radiative forcin g parameters have been derived using parameterized radiation models wi th satellite meteorological data from the International Satellite Clou d Climatology Project, and directly measured top-of-atmosphere radiati ve fluxes from the Earth Radiation Budget Experiment. Specifically, sh ortwave, longwave, and total cloud forcing at the surface, and column- averaged values of longwave cloud forcing of the atmosphere were deriv ed for the midseasonal months of April, July, and October 1985 and Jan uary 1986, covering a complete annual cycle. Seasonal variability is i llustrated by comparing the results for July 1985 and January 1986, wh ich represent the seasonal extremes. Surface shortwave cloud forcing i s always negative, representing a cooling of the surface, with stronge st cooling (-120 to -180 W m-2) occurring over midlatitude storm track s of the summer hemisphere. Surface longwave cloud forcing is always p ositive, representing a warming of the surface, with strongest warming (60 to 75 W m-2) occurring over storm tracks of the winter hemisphere . Zonal averages show the entire summer hemisphere dominated by shortw ave cooling, the middle and high latitudes of the winter hemisphere do minated by longwave warming, and a broad zone of transition in between . The globally averaged total cloud forcing amounts to a cooling throu ghout the year, ranging from a low of about -12 W m-2 for July 1985 to a high of about -25 W m -2 for January 1986. The longwave cloud forci ng of the atmosphere shows a strong warming over deep convective regio ns in the tropics and a moderate cooling outside the tropics, amountin g to a weak cooling (-2 to -5 W m-2)in the global average. Comparisons of the results with general circulation model simulations show broad qualitative agreement regarding the locations of prominent warming and cooling regions. Quantitative comparisons, on the other hand, show si gnificant differences between the magnitudes of warming and cooling in these regions. Most of the larger differences can be attributed to kn own deficiences of the general circulation model simulations. Comparis ons with satellite-derived results of other investigators show much be tter agreement.