Volatility measurements by Twomey (1971) and Hudson and Da (1996) showed that the vast majority of CCN over the ocean cannot be NaCl.  However, other reports indicate that NaCl is a major CCN component.  Here we contrast detailed cloud condensation nuclei (CCN) spectral volatility (thermal fractionation) measurements from three aircraft field projects to provide insight into the relative contribution of sea salt. The most remote location, the Pacific Aerosol Sulfur Experiment (PASE) (mid-Pacific), had the highest average CCN concentrations (N_CCN) probably because it was the least cloudy whereas the less remote, but more cloudy, Rain in Cumulus over the Ocean (RICO) project (Caribbean) had the lowest average N_CCN (Hudson and Noble 2009).  In RICO particle concentrations in all size ranges larger than 0.3 micrometers were well correlated with wind speed (R ~ 0.87) but uncorrelated with N_CCN (Fig. 1A).  Smaller particles in RICO were correlated with N_CCN but uncorrelated with wind speed.  In PASE only particles larger than 10 micrometers were correlated with wind speed and concentrations in these size ranges were uncorrelated with N_CCN.  Particles smaller than 10 micrometers in PASE were uncorrelated with wind speed but well correlated with N_CCN.  In both projects particle concentrations smaller than these respective sizes were highly correlated with N_CCN, at all S in PASE but mainly with N_CCN at high S in RICO.  CCN volatility measurements showed high correlations between refractory (non-volatile) N_CCN and wind speed (Fig. 2A and 2B), especially for low supersaturation (S) N_CCN, and no correlation of volatile N_CCN at all S with wind speed (i.e., Fig. 2C for RICO).  In PASE there was only a weak positive correlation between refractory N_CCN and wind and also no correlation between volatile N_CCN and wind speed.  These results indicate that in clean maritime air the wind originated component of N_CCN can be substantial (i.e., > 30% for wind speed > 14 m/s) but that in maritime air with higher N_CCN the wind derived CCN component is probably less than 10%.  We also find that the small concentrations of giant nuclei (> 3 micrometers) can affect drizzle concentrations in RICO (Hudson et al. 2011).

                In each project N_CCN at all S was uncorrelated with all ambient particle concentrations larger than these same respective sizes (Fig. 1).  N_CCN at all S was also uncorrelated with U in both projects.  The contrast in cloudiness between the two projects was responsible for many of the differences noted between the two projects.  A third project (Physics of Stratocumulus Tops—POST) showed appropriately intermediate results to those of RICO and PASE.  The results indicate that the effects of clouds on N_CCN play a major role in the relative influence of sea salt on N_CCN and ultimately on climate.   

Hudson, J.G. and S. Noble, 2009: Geophys. Res. Let., 36, L13812,

Hudson, J.G. and X. Da, 1996:  J. Geophys. Res., 101, 4435-4442.

Hudson, J. G., V. Jha, and S. Noble, 2011: Geophys. Res. Lett., 38, L05808.

Twomey, S.:  1971:  J. Atmos. Sci., 28, 377-381.