Non-sea salt sulfate (ng m )
FIGURE 14.45 Log-log plot of CCN or cloud droplet concentration measured in a number of field studies as a function of non-sea salt sulfate (nss). Data are from (□) Cape Grim, Tasmania, CCN measured at 0.46% supersaturation (Gras, 1989; Ayers and Gillett, 1989); (a) Puerto Rico, CCN measured at 0.5% supersaturation (Novakov and Penner, 1993); (O) western Washington, CNN measured at 0.3% supersaturation (Berresheim et al., 1993); (O) the Azores (Hegg et al., 1993); (x) Whiteface Mountain, New York, cloud droplets measured (Pueschel et al., 1986); ( *) central Ontario, Canada, cloud droplets measured (Leaitch et al., 1992); and (•) North Atlantic, cloud-processed accumulation mode particles measured (Van Dingenen et al., 1995) (adapted from Van Dingenen et al., 1995).
at different supersaturations versus cloud droplet concentrations, the large range in sulfate concentrations, etc.), such correlations (r2 = 0.42) suggest that it is possible to relate particle mass to CCN and cloud droplet number concentration, and ultimately to changes in cloud albedo, albeit within a large uncertainty.
However, some caution is needed in applying such correlations, as might be expected, particularly given our current relatively rudimentary understanding of the indirect effects of aerosol particles on clouds. For example, the data in Fig. 14.45 include both CCN and cloud droplet number concentrations, which both appear to be correlated to nss. However, this is not always the case. While CCN and non-sea salt mass concentrations were observed to be highly correlated at a marine site in Puerto Rico, cumulus cloud droplet number concentrations were not, and stratocumulus cloud droplet number concentrations showed a very low sensitivity to nss (Novakov et al., 1994). Similarly, Anderson and co-workers (1994) did not observe a convincing relationship between the cloud droplet number concentration and the aerosol number or volume at a coastal mountain site in the state of Washington. Entrainment and mixing processes in the clouds may have played major roles in these (and, of course, many other) studies. Finally, some studies suggest that particularly at very small sulfate concentrations, a wide range of CCN can be observed (e.g., Hegg, 1994), which may be related to non-sulfate species acting as CCN.
indeed, while the initial focus on the indirect effects of anthropogenic aerosols has been on sulfate, there is increasing evidence that other species may also not only contribute significantly to CCN but actually dominate it under many circumstances. For example, although CCN at 1% supersaturation were correlated with sulfate in air masses over the northeast Pacific and the northeast Atlantic, the slope of the curve relating the two was much higher for the relatively clean northeast Pacific (Hegg et al., 1993). The authors suggest that this is consistent with DMS as the major source of sulfate over the Pacific. Over the Atlantic, however, the slope of CCN versus sulfate was smaller and there was a significant intercept, suggesting that much of the CCN was not formed from sulfate. Observations at a coastal site in the state of Washington also suggested that components other than sulfate may be important in CCN formation at 0.9% supersaturation (Berresheim et al., 1993).
Novakov and Penner (1993) measured the mass size distributions of sulfur, organic carbon, and chlorine (characteristic of sea salt) as well as the CCN concentration (at 0.5% supersaturation), nss, and Aitken nuclei concentrations at a mountain peak in Puerto Rico.
They concluded that about 63% of the CCN at this site was due to organic aerosol particles, possibly due to some unspecified anthropogenic sources. Similar measurements at Point Reyes, California, gave variable contributions of organics and sulfate to CCN, ranging from 4 to 78% for organic particles, from f 9 to 64% for sulfate, and from 9 to 31% for NaCl in sea salt particles (Rivera-Carpio et al., 1996). While Andrews et al. (1997) suggest that the organic aerosol particles in the Puerto Rican studies may have originated from the rain forest below the mountain peak sampling site, subsequent studies at the mountain site and at a site in the Atlantic Ocean suggest that a large fraction of this organic aerosol may originate from the ocean (Novakov et al., f997a). A similar oceanic source of CCN measured in Antarctica was suggested by Saxena (1996). ft is particularly interesting that most of the organic aerosol particles in the Puerto Rican study were water soluble; in addition, their average mass concentration (390 ng m~3) was larger than that of sulfate (270 ng m~3). This combination of water solubility and relatively high mass fraction suggests that the organic particles may be particularly effective as CCN (Novakov et al., f997a).
Recent studies have provided additional evidence for the contribution of organics to CCN. For example, Matsumoto et al. (1998) measured CCN at 0.5 and 1% supersaturations, along with the aerosol particle composition and size distribution at the Ogasawara Islands in the northwest Pacific Ocean, in agreement with earlier studies, air masses affected by continental emissions had CCN concentrations of ~ f50-f000 cm"3 (at f% supersaturation) compared to 30-150 cm-3 for clean air masses. Sulfate, nitrate, and ammonium in the particles were correlated with 222 Rn, which is a tracer of continental air masses. Oxalate was also found in the particles, primarily in the accumulation mode (<1.1 yu,m), and was highly correlated with 222 Rn, indicating an anthropogenic source. However, formate and ac-
Was this article helpful?