ALBERT

Description: In this study, five different marine primary organic aerosol (POA) emission schemes have been evaluated using the GEOS-Chem chemical transport model in order to provide guidance for their implementation in air quality and climate models. These emission schemes, categorized in two groups based on varying dependences of chlorophyll a concentration ([chl a]) and 10 m wind speed (U10), have large differences in their magnitude, spatial distribution, and seasonality. Model comparison with weekly and monthly mean values of the organic aerosol mass concentration at two coastal sites shows that the source function exclusively related to [chl a] does a better job replicating surface observations. Sensitivity simulations of the sea spray-based parameterizations show that improved predictions of the seasonality of the marine POA concentrations can be achieved by varying the U10 and [chl a] dependence of the organic mass fraction of sea spray aerosol. A top-down estimate of submicron marine POA emissions based on the parameterization that compares best to the observed weekly and monthly mean values of marine organic aerosol surface concentrations has a global average emission rate of 6.3 Tg yr−1. Evaluation of existing marine POA source functions against a case study during which marine POA contributed the major fraction of submicron aerosol mass shows that none of the existing parameterizations are able to reproduce the hourly-averaged observations. Our calculations suggest that in order to capture episodic events in submicron marine POA concentration over the ocean, new source functions need to be developed that are grounded in the physical processes unique to the organic fraction of sea spray aerosol.