Simulated Dust Aerosol Lifecycle in the NASA GEOS-5 Atmospheric Transport Model and Sensitivity to Source and Sink Mechanisms
Nowottnick, Edward Paul
Colarco, Peter R
MetadataShow full item record
Understanding interactions of mineral dust aerosols with the Earth system remains a key uncertainty in assessing global climate change. A significant portion of this uncertainty arises due to an incomplete understanding of the source, transport, and loss processes that control the dust aerosol lifecycle. Global aerosol transport models compliment traditional observational platforms to serve as useful tools for exploring aerosol-Earth system interactions. However, global models are limited by scale, requiring parameterizations to represent the lifecycle of dust. Here, the simulated dust lifecycle is explored in versions 4 and 5 of the NASA Goddard Earth Observing System (GEOS-4/5) model. Different treatments of the mobilizing physics are first explored by considering two mobilization schemes in GEOS-4. Both schemes produced similar distributions of aerosol optical thickness (AOT) and extinction that become more comparable with observations downwind of the source region. Despite similarities in the optical comparisons, the schemes differ in mass loadings owing to differences in emitted particle size distributions, suggesting that emission scheme choice is significant for mass budgets and particle size distributions. The effect of spatial resolution on source processes was investigated in GEOS-5. Model spatial resolution had a significant impact on simulated dust distributions, as increased model spatial resolution resolves higher wind speeds used to parameterize dust emissions. Model spatial resolution had regional implications, as simulated dust distribution exhibited the greatest sensitivity over the Asian source region. The incorporation of sub-grid wind variability in a coarse resolution simulation led to improved agreement with observed AOT magnitude, but did not improve the timing of simulated dust events over the Asian source region.GEOS-5 was used to investigate the cause of an observed barrier to dust transport across Central America into the Pacific. The baseline simulation did not develop as strong of a barrier when compared to observations. Better agreement was obtained when the parameterization for wet removal was treated as other hydrophilic aerosols. Analysis of the dust transport dynamics and loss processes suggest that while both mechanisms play a role in defining the barrier, loss processes by wet removal are about twice as important as transport.