Quantifying biases in aerosol-cloud-radiation interactions associated with spatiotemporal averaging

Abstract:

Aerosol-cloud interactions represent one of the largest sources of error in projections of future climate states. Climate model resolutions are too coarse to resolve clouds, aerosol, and their interaction at important small scales; At the same time, observationally based aerosol-cloud metrics are highly uncertain, and their magnitude a matter of debate. Here we consider the uncertainty in aerosol-cloud interaction metrics associated with the spatiotemporal averaging of aerosol and cloud fields, and the attendant biases in the strength of albedo response to aerosol perturbations. Through the use of ensembles of large eddy simulations, as well as satellite-based observations, we quantify this bias for a range of cloud and aerosol conditions. While cloud albedo biases are fairly well constrained, drop concentration biases are more uncertain. These compound to produce much more significant biases in the albedo susceptibility that are contingent on the heterogeneity of the cloud and aerosol fields, and the correlation between the two. The results caution against using highly aggregated aerosol-cloud response functions to represent cloud-scale albedo responses to aerosol perturbations.