We use global, multiyear observations of the properties of clouds, the atmosphere, and the surface to calculate global shortwave (SW) and longwave (LW) fluxes at the top of the atmosphere and at the surface at a resolution of 280 km and 3 hours for every third month from April 1985 to January 1989. Our validation studies suggest that the specification of cloud effects is no longer the dominant uncertainty in reconstructing the radiative fluxes at the top of atmosphere and at the surface. Rather cloud property uncertainties are now roughly equal contributors to the flux uncertainty, along with surface and atmospheric properties. Overall, the estimated uncertainties in regional, monthly mean fluxes, based on comparisons to more direct measurements, are as follows: for S↑t, 11 W/m2 bias with a rms scatter of 7 W/m2 (the bias may be as little as 5 W/m2); for L↑t, -1 W/m2 bias with a rms scatter of 4 W/m2; for S↓s, 10--20 W/m2 bias with a rms scatter of 10--15 W/m2 (the bias may only be ≈10 W/m2 caused by aerosols); for S↑s, ≈3 W/m2 bias with a rms scatter of 10 W/m2; for L↓s, probably <15 W/m2 bias (but with variable sign) with a rms scatter of 15 W/m2 (bias and scatter depend most on atmospheric temperature uncertainties); and for L↑s, ≲12 W/m2 and ≲24 W/m2 for combined bias and scatter over oceans and land, respectively (very sensitive to surface temperature determinations).

The resulting SW and LW flux data sets suggest the following conclusions: (1) The net SW heating of Earth appears predominantly at the surface, whereas the net LW cooling arises predominantly from the atmosphere. The net cooling effect of clouds on top of atmospheric radiation appears primarily at the surface rather than in the atmosphere. (2) Clouds have almost no net effect on the global mean radiation balance of the atmosphere, but they enhance the latitudinal gradient in the LW cooling and reinforce the radiative forcing for the mean atmospheric circulation. Clouds act to mute seasonal contrasts however. (3) Clouds enhance the land-ocean contrasts of the atmospheric cooling, reinforcing the growth of standing eddy motions; but reduce land-ocean contrasts of the surface heating. ¿ American Geophysical Union 1995