Cloud Horizontal Inhomogeneity and the Plane-Parallel Albedo Bias—Contrasting High and Coarse Resolution Atmospheric Models

Despite significant research effort, the realistic representation of low-level clouds in atmospheric models remains an unsolved challenge, due to the complexity and multi-scale nature of the relevant cloud processes. As the radiation budget is strongly modulated by low-level clouds, this introduces significant uncertainties in climate predictions and in our understanding of cloud-radiation interactions and cloud feedbacks. A well-known issue is the plane-parallel albedo bias, which refers to a reduction of cloud albedo or reflectance for clouds exhibiting realistic variability below that of an otherwise identical homogeneous cloud. To correct for this bias and to account for horizontal inhomogeneities, current general circulation models often employ simple correction schemes, such as scaling the liquid water path by a constant reduction factor to obtain a radiatively effective amount. Such corrections however can mask deficiencies in the model representation of cloud properties, if they are determined empirically. In our investigation, we contrast simulations conducted with the atmospheric model ICON-LEM at a resolution of 156m with coarse-resolution model runs with the ECHAM6 model at 80km resolution to determine the magnitude and variability of the plane-parallel albedo bias of low-level clouds. Besides variability in liquid water path, the influence of horizontal and vertical variations in cloud droplet size is investigated, which is available from the two-moment cloud microphysics of ICON-LEM. The scale dependence of variability in cloud properties is studied, in particular to enable a meaningful comparison of the model cloud properties at different resolutions and with satellite data. Initial ideas are proposed toward a cloud regime-dependent and scale-aware correction of the plane-parallel albedo bias for coarse-resolution atmospheric models.