Articles | Volume 8, issue 12
https://doi.org/10.5194/acp-8-3149-2008
https://doi.org/10.5194/acp-8-3149-2008
24 Jun 2008
 | 24 Jun 2008

Tropospheric aerosol microphysics simulation with assimilated meteorology: model description and intermodel comparison

W. Trivitayanurak, P. J. Adams, D. V. Spracklen, and K. S. Carslaw

Abstract. We implement the TwO-Moment Aerosol Sectional (TOMAS) microphysics module into GEOS-CHEM, a CTM driven by assimilated meteorology. TOMAS has 30 size sections covering 0.01–10 μm diameter with conservation equations for both aerosol mass and number. The implementation enables GEOS-CHEM to simulate aerosol microphysics, size distributions, mass and number concentrations. The model system is developed for sulfate and sea-salt aerosols, a year-long simulation has been performed, and results are compared to observations. Additionally model intercomparison was carried out involving global models with sectional microphysics: GISS GCM-II' and GLOMAP. Comparison with marine boundary layer observations of CN10 and CCN(0.2%) shows that all models perform well with average errors of 30–50%. However, all models underpredict CN10 by up to 42% between 15° S and 45° S while overpredicting CN10 up to 52% between 45° N and 60° N, which could be due to the sea-salt emission parameterization and the assumed size distribution of primary sulfate emission, in each case respectively. Model intercomparison at the surface shows that GISS GCM-II' and GLOMAP, each compared against GEOS-CHEM, both predict 40% higher CN10 and predict 20% and 30% higher CCN(0.2%) on average, respectively. Major discrepancies are due to different emission inventories and transport. Budget comparison shows GEOS-CHEM predicts the lowest global CCN(0.2%) due to microphysical growth being a factor of 2 lower than other models because of lower SO2 availability. These findings stress the need for accurate meteorological inputs, updated emission inventories, and realistic clouds and oxidant fields when evaluating global aerosol microphysics models.

Download
Altmetrics
Final-revised paper
Preprint