Global modelling of direct and indirect effects of sea spray aerosol using a source function encapsulating wave state

Recently developed parameterizations for the sea spray aerosol source flux, encapsulating wave state, and its organic fraction were incorporated into the aerosol-climate model ECHAM-HAMMOZ to investigate the direct and indirect radiative effects of sea spray aerosol particles. Our simulated global sea salt emission of 805 Tgyr(-1) (uncertainty range 378-1233 Tgyr(-1)) was much lower than typically found in previous studies. Modelled sea salt and sodium ion concentrations agreed relatively well with measurements in the smaller size ranges at Mace Head (annual normalized mean model bias -13% for particles with vacuum aerodynamic diameter D-va < 1 mu m), Point Reyes (-29% for particles with aerodynamic diameter D-a < 2.5 mu m) and Amsterdam Island (-52% for particles with D-a < 1 mu m) but the larger sizes were overestimated (899% for particles with 2.5 mu m < D-a < 10 mu m) at Amsterdam Island. This suggests that at least the high end of the previous estimates of sea spray mass emissions is unrealistic. On the other hand, the model clearly underestimated the observed concentrations of organic or total carbonaceous aerosol at Mace Head (-82%) and Amsterdam Island (-68%). The large overestimation (212%) of organic matter at Point Reyes was due to the contribution of continental sources. At the remote Amsterdam Island site, the organic concentration was underestimated especially in the biologically active months, suggesting a need to improve the parameterization of the organic sea spray fraction. Globally, the satellite-retrieved AOD over the oceans, using PARASOL data, was underestimated by the model (means over ocean 0.16 and 0.10, respectively); however, in the pristine region around Amsterdam Island the measured AOD fell well within the simulated uncertainty range. The simulated sea spray aerosol contribution to the indirect radiative effect was positive (0.3 W m(-2)), in contrast to previous studies. This positive effect was ascribed to the tendency of sea salt aerosol to suppress both the in-cloud supersaturation and the formation of cloud condensation nuclei from sulfate. These effects can be accounted for only in models with sufficiently detailed aerosol microphysics and physics-based parameterizations of cloud activation. However, due to a strong negative direct effect, the simulated effective radiative forcing (total radiative) effect was -0.2 W m(-2). The simulated radiative effects of the primary marine organic emissions were small, with a direct effect of 0.03 W m(-2) and an indirect effect of -0.07 W m(-2)