A comparison of sea salt emission parameterizations in northwestern Europe using a chemistry transport model setup

Abstract

Atmospheric sea salt particles affect chemical and physical processes in the atmosphere. These particles provide surface area for condensation and reaction of nitrogen, sulfur, and organic species and are a vehicle for the transport of these species. Additionally, HCl is released from sea salt. Hence, sea salt has a relevant impact on air quality, particularly in coastal regions with high anthropogenic emissions, such as the North Sea region. Therefore, the integration of sea salt emissions in modeling studies in these regions is necessary. However, it was found that sea salt concentrations are not represented with the necessary accuracy in some situations. In this study, three sea salt emission parameterizations depending on different combinations of wind speed, salinity, sea surface temperature, and wave data were implemented and compared: GO03 (Gong, 2003), SP13 (Spada et al., 2013), and OV14 (Ovadnevaite et al., 2014). The aim was to identify the parameterization that most accurately predicts the sea salt mass concentrations at different distances to the source regions. For this purpose, modeled particle sodium concentrations, sodium wet deposition, and aerosol optical depth were evaluated against measurements of these parameters. Each 2-month period in winter and summer 2008 were considered for this purpose. The shortness of these periods limits generalizability of the conclusions on other years. While the GO03 emissions yielded overestimations in the PM10 concentrations at coastal stations and underestimations of those at inland stations, OV14 emissions conversely led to underestimations at coastal stations and overestimations at inland stations. Because of the differently shaped particle size distributions of the GO03 and OV14 emission cases, the deposition velocity of the coarse particles differed between both cases which yielded this distinct behavior at inland and coastal stations. The PM10 concentrations produced by the SP13 emissions generally overestimated the measured concentrations. The sodium wet deposition was generally underestimated by the model simulations but the SP13 cases yielded the least underestimations. Because the model tends to underestimate wet deposition, this result needs to be considered critically. Measurements of the aerosol optical depth (AOD) were underestimated by all model cases in the summer and partly in winter. None of the model cases clearly improved the modeled AODs. Overall, GO03 and OV14 produced the most accurate results, but both parameterizations revealed weaknesses in some situations.