Remote sensing of surface solar irradiance: corrections for 3-D cloud effects

It has been shown that simple 1-D radiative transfer models can be used with narrow-band satellite radiance observations to estimate the surface radiation budget with a very good overall accuracy. However, as the temporal and spatial resolution increase, large errors may occur in the presence of non-homogeneous cloud fields as comparisons with surface observations indicate. This suggests that 3-D cloud effects become increasingly important for the radiance field observed by the satellite sensor, and that simple 1-D retrieval schemes are not adequate under general cloudy conditions. We used a 3-D Monte Carlo model with simulated cloud liquid water and ice fields to compute radiances at the top of the atmosphere and surface radiation fluxes. Stratiform and deep convective cloud systems are simulated using the PSU/NCAR Mesoscale Model (MM5). The radiance data base thus simulated is analyzed to determine the impact of horizontal resolution and cloud geometry on the relationship between 3-D cloud properties, top of the atmosphere radiance and surface irradiance. The goal is to determine the 3-D cloud characteristics that are responsible for the deterioration of the 1-D surface radiation retrieval scheme and find possible corrections for the cloud 3-D effects.