Evaluating Global Model Parameterizations of Aerosol Optical Properties Using In Situ Data from ARCTAS

We evaluate the externally-mixed, fixed size distribution aerosol optical property parameterizations used in the NASA Global Modeling Initiative (GMI) model and the GEOS-Chem model with in situ data on aerosol scattering and absorption (at 3 wavelengths) gathered during the Arctic Research of the Composition of the Troposphere from Aircraft and Satellites (ARCTAS) campaign. We also compare the ARCTAS data with the size-resolved sectional Aerosol Simulation Program (ASP), and use the ASP results to evaluate the assumptions about the size distribution, hygroscopic growth, and the refractive indices of organic aerosols (OA) in the global parameterizations, as well as several possible mixing rules for black carbon. The models were intialized with in situ observations of aerosol mass and composition (and size distribution for ASP).

Our preliminary analysis suggests that the GMI and GEOS-Chem aerosol parameterizations show reasonable correlations with the observed scattering and absorption coefficients at all three wavelengths (r²=0.5-0.75). Both models have slight positive mean biases in scattering and absorption coefficients, with over 70% of the model-observations differences within a factor of 2 of this mean bias. The GMI and GEOS-Chem model predictions of single scattering albedo (SSA) and asymmetry parameter (g) show little mean bias; however, there is little correlation between the modeled and observed SSA and g.