Top-of-atmosphere radiation bias from neglecting three-dimensional cloud morphology

Clouds cover on average nearly 70% of the Earth's surface and play a dominant role in setting the global shortwave albedo. The details of the shortwave reflection depend on the location of the clouds, their optical properties, and their 3D structure. Earth system models are unable to resolve the 3D morphology of clouds and thus neglect its effect on radiative transfer. Here we show how the resulting radiative energy flux bias depends on cloud type, morphology, and solar zenith angle. Using large-eddy simulations to produce realistic 3D cloud fields, a Monte Carlo code for 3D radiative transfer, and ISCCP observations of cloud climatology, we estimate the effect of this flux bias on the simulated global climate in models. The radiative flux bias is largest at small zenith angles and for deeper clouds, while the albedo bias is largest (and negative) for very large zenith angles. The estimated radiative flux bias is therefore largest at the equator where zenith angles are smaller on average and deep clouds are often present. Globally, the radiative flux bias is on the order of 1.5 W m², comparable to the magnitude of anthropogenic greenhouse gas forcing, and locally can be on the order of 10 W m².