1JULY 2008 AYASH ET AL. 3207 Direct and Indirect Shortwave Radiative Effects of Sea Salt Aerosols TAREK AYASH Department of Chemical Engineering and Applied Chemistry, University of Toronto, Toronto, Ontario, Canada SUNLING GONG Air Quality Research Branch, Meteorological Service of Canada, Toronto, Ontario, Canada CHARLES Q. JIA Department of Chemical Engineering and Applied Chemistry, University of Toronto, Toronto, Ontario, Canada (Manuscript received 29 May 2007, in final form 13 November 2007) ABSTRACT Sea salt aerosols play a dual role in affecting the atmospheric radiative balance. Directly, sea salt particles scatter the incoming solar radiation and absorb the outgoing terrestrial radiation. By acting as cloud condensation nuclei, sea salt aerosols indirectly modulate the atmospheric radiative budget through their effective contribution to cloud formation. Using the Canadian Aerosol Module (CAM)–Canadian Centre for Climate Modelling and Analysis (CCCma) GCM, version 3 (GCM3) framework, the direct as well as the indirect shortwave (SW) radiative effects of sea salt aerosols are simulated. The model results herein suggest that sea salt aerosols exert a significant direct radiative effect over oceanic regions, with seasonal means in the range from Ϫ2toϪ3WmϪ2 over the Southern Ocean. Globally, sea salt’s SW indirect effect (annual mean Ϫ0.38 W mϪ2) is found to be less than its direct effect (annual mean Ϫ0.65 W mϪ2). However, sea salt’s indirect effect is found to be far stronger over the Southern Hemisphere than over the Northern Hemisphere, especially over the Southern Ocean with seasonal means around Ϫ4WmϪ2, which exceed its direct effect. The model results herein suggest that sea salt aerosols significantly modulate the atmospheric radiation budget over oceanic regions and need to be accounted for in global climate models. 1. Introduction high solubility, sea salt particles grow hygroscopically to large sizes and, thus, are quite efficiently removed by Climate on earth is known as a highly dynamic and both dry and wet deposition. This results in sea salt complex system, and atmospheric aerosols have been particles having typically the shortest residence times increasingly recognized as principal agents in that sys- and dominating atmospheric aerosol mass burdens near tem. In this regard, the role of aerosols of anthropo- their high source regions. genic origin, mainly sulfate and carbonaceous aerosols, Sea salt aerosols play a dual role in affecting the within climate change has been of primary interest. atmospheric radiative balance. Directly, sea salt par- However, of equal importance is understanding the ticles interact with the incoming solar radiation and the role of natural aerosols, dominated by sea salt and soil outgoing terrestrial radiation. Unlike the more hydro- dust, in modulating the earth–atmosphere radiative phobic soil dust aerosol, sea salt particles uptake water balance and, consequently, the dynamics of the climate readily and, hence, are highly scattered at shortwave system. Most global models show that, among all aero- (SW; solar) wavelengths with virtually no absorption sol types, mass emissions are dominated by sea salt (e.g., Takemura et al. 2002). However, similar to soil aerosol (Textor et al. 2006). However, owing to their dust aerosol, because of their large sizes, sea salt par- ticles are also efficient absorbers at longwave (LW, ter- restrial) wavelengths (Satheesh and Lubin 2003). The Corresponding author address: Tarek Ayash, Dept. of Chemical Engineering and Applied Chemistry, University of Toronto, 200 highly hygroscopic nature of sea salt particles renders College Street, Toronto, ON M5S 3E5, Canada. them very efficient cloud condensation nuclei (CCN). E-mail: [email protected] Over the remote oceanic regions where the presence of DOI: 10.1175/2007JCLI2063.1 © 2008 American Meteorological Society Unauthenticated | Downloaded 09/30/21 11:10 PM UTC JCLI2063 3208 JOURNAL OF CLIMATE VOLUME 21 other aerosol species is relatively weak, sea salt aero- observational data provide evidence of the mechanisms sols are the dominant contributor to CCN numbers underlying the indirect effect of sea salt aerosols. For (Quinn et al. 2000). Even in the presence of the effi- example, satellite data revealed plumes of reduced ciently condensing sulfate aerosol, sea salt particles ac- cloud particle size and suppressed precipitation origi- tivate to form cloud droplets more readily than sulfate nating from major urban and industrial areas; however, aerosol (O’Dowd et al. 1999) because of their larger precipitation from similarly polluted clouds over oceans size and the lower supersaturation threshold that is re- appeared to be much less affected (Rosenfeld 2000). quired for their activation. However, because CCN One possible explanation is that the giant sea salt nuclei concentration depends on supersaturation, the smaller override the precipitation suppression effect of a large but more abundant sulfate particles often dominate sea number of small pollution nuclei (Rosenfeld et al. salt as CCN at supersaturations exceeding 1%. By act- 2002). This is in accordance with the findings that the ing as CCN, therefore, sea salt aerosols indirectly more hygroscopic a CCN, the earlier it will become a modulate the atmospheric radiative balance through droplet as water saturation is achieved, and that, at a their effective contribution to cloud formation. given composition, larger CCNs will acquire water first Several model studies provide estimates of the direct (Kondratyev et al. 2006). radiative effects of sea salt aerosol. General circulation The Canadian Aerosol Module (CAM) was devel- model (GCM) simulations (Dobbie et al. 2003; Grini et oped for simulating the atmospheric cycling of aerosols, al. 2002; Jacobson 2001; Takemura et al. 2002) provide with the Canadian Centre for Climate Modelling and a wide range of estimates of the SW and net (shortwave Analysis (CCCma) GCM, version 3 (GCM3) as its cli- and longwave) direct radiative effects, which is between matological driver model. The global budget and dis- Ϫ0.15 and Ϫ2.2WmϪ2 for the clear-sky global annual tribution of sea salt were simulated by CAM and mean and between Ϫ0.31 and Ϫ1.1WmϪ2 for the compared to observations (Gong et al. 2002). Sea salt’s whole-sky global annual mean. Such a wide range of impact on the global budget and distribution of non–sea estimates mainly reflects the large uncertainty underly- salt (nss) sulfate aerosols was also simulated (Gong and ing sea salt’s emission modeling, which is mainly be- Barrie 2003). Using this model framework, here we aim cause sea salt’s generation mechanisms are highly at quantifying the direct and indirect SW radiative ef- determined by wind speed (e.g., Gong et al. 2002). This fects of sea salt aerosols. The model framework and the is well documented by Haywood et al. (1999), who pro- simulation methodology are described in section 2. Re- vide “low” (Ϫ1.51 W mϪ2) and “high” (Ϫ5.03 W mϪ2) sults of model simulations are presented and discussed GCM estimates of the SW sea salt direct effect (annual in section 3, ending with conclusions in section 4. mean, ocean averaged) by using two different empirical approaches that encompass the considerable range in 2. Model framework and simulation methodology the effect of wind speed on sea salt aerosol concentra- a. CAM–CCCma atmospheric general circulation tion. However, Haywood et al. show that the difference model, version 3 framework between their GCM-simulated and observed clear-sky irradiance over the oceans is smallest if a high burden The atmospheric general circulation model, version 3 of sea salt is assumed. In addition to GCM-based esti- (AGCM3) shares many basic features with the AGCM, mates, several estimates based on a combination of version 2 (AGCM2; McFarlane et al. 1992). As in models and observational data suggest significant SW AGCM2, the spectral transform method in AGCM3 is and LW direct radiative effects of sea salt at local and used to represent the horizontal spatial structure of the regional scales (Lubin and Vogelmann 2004; Podgorny main prognostic variables, while the rectangular finite and Ramanathan 2003; Satheesh and Lubin 2003). elements defined for a hybrid vertical coordinate rep- Regarding sea salt’s indirect radiative effect (IRE), resent the vertical spatial structure (LaPrise and Girard no GCM-based estimates have yet been made avail- 1990). A semi-Lagrangian transport algorithm also ex- able. By deriving empirical relationships based on ists in AGCM3 as an option for transport of moisture shipborne and island-based observational data, Vinoj and other trace constituents. The spectral representa- and Satheesh (2004) estimate that the IRE of sea salt tion currently used in AGCM3 corresponds to a higher aerosols over the Arabian Sea may be as large as horizontal resolution than that used in AGCM2, com- Ϫ7 Ϯ 4WmϪ2 compared to Ϫ2 Ϯ 1WmϪ2 for their posed of a 47-wave triangularly truncated (T47) spheri- direct radiative effect (DRE) estimate. Vinoj and cal harmonic expansion. The vertical domain of Satheesh’s findings, therefore, suggest that the indirect AGCM3 is deeper than in AGCM2, extending from the radiative effect resulting from sea salt aerosols may be surface to the stratopause region (1 hPa), and the ver- more prominent than their direct radiative effect. Some tical resolution is also higher (32 layers). The model Unauthenticated | Downloaded 09/30/21 11:10 PM UTC 1JULY 2008 AYASH ET AL. 3209 time step is