The diurnal cycle of radiative fluxes using 3D stratocumulus cloud fields from Large-Eddy simulations

In operational GCMs shortwave radiative fluxes are commonly calculated with 1D plane-parallel radiative transfer approximations. The generally scarce cloud information in GCMs does not require per-se to use more sophisticated radiative transfer methods. However, it is desirable to include unresolved optical property fluctuations in GCM radiative transfer calculations to account, e.g., for the well-known effect of cloud inhomogeneities on radiative fluxes. In order to assess the variability of optical parameters on GCM subgrid scales we analyze the diurnal cycle of stratocumulus cloud fields by means of a Large-Eddy Simulation (LES) model. The model is initialized on the basis of rawinsonde vertical profiles measured off the coast of California during the FIRE I experiment. In agreement with the microwave retrievals the modeled diurnal cycle of the mean liquid water path (LWP) evolution is characterized by an increase during the night, and thinning during day-time as a result of cloud warming by shortwave radiative absorption. The LES results are also used to determine the evolution of the probability distribution function of the LWP. We have performed additional LES runs where we changed the magnitude of the humidity inversion jump, aiming to assess its role on the generation of spatial fluctuations in the LWP. Preliminary results show that a drier inversion causes a smaller mean LWP and reduced cloud fractions. These LES cloud fields are used in 1D, IPA, and full 3D radiative transfer simulations to intercompare radiative fluxes and heating rates. A key question is how the LWP evolution relates to the variations in the plane parallel albedo-bias.