Stratus cloud supersaturations

Conventional wisdom is that stratus cloud supersaturations [S] are < 0.3% (100.3% RH) (Hegg et al. 2009) and that higher CCN concentrations (N_CCN) restrict cloud S (i.e., Twomey 1959). The Physics of Stratocumulus Tops (POST) field experiment in July-August, 2008 off the central California coast provided a wide enough range of N_CCN and droplet concentrations (N_c) to test these concepts. Condensation nuclei (CN) were measured with a TSI 3010, CCN were measured by a Desert Research Institute CCN spectrometer (Hudson 1989), cloud droplets were measured with a Cloud Aerosol Spectrometer (CAS) probe (diameter 0.58-51 µm), W was measured with a GPS corrected C-MIGITS III all aboard the CIRPAS Twin Otter. N_c are averages of droplets larger than 0.58 µm diameter for those one second periods where CAS liquid water content (LWC) exceeded 0.1 gm^-3. N_CCN are measured below the clouds.

Row 1, Table 1 shows the high correlations (R) between N_1% (N_CCN at 1% S) and N_c. Although below-cloud CCN provide the greatest influence on cloud microphysics the often-observed high N_CCN immediately above California coastal stratus (e.g., Hudson and Frisbie 1991) might also play a role in cloud microphysics by entrainment. In an attempt to ascertain this influence, the 69 vertical cloud penetrations are divided according to vertical aerosol profiles of CN concentrations (N_CN). These are used because the faster response of CN makes for easier discrimination from within-cloud measurements. Columns 3 and 4 of row 1, Table 1 indicate high and similar R of N_c-N_1% for the 40 cases with higher N_CN immediately above the stratus than below cloud N_CN (designated HA) and the 29 cases with just above cloud N_CN lower or similar to below cloud N_CN (designated LSA). These facts indicate that there was minimal influence of the above cloud CCN on N_c. This is also indicated by a division of each cloud penetration into altitude quartiles, which revealed the highest N_c-N_1% R for the highest cloud altitude quartile.

Table 2 uses the entire CCN spectra between 1.5 and 0.04% S to estimate effective cloud S (S_eff) by determining the S of the various N_CCN that match mean N_c of each of the corresponding 69 vertical cloud penetrations (i.e., Hudson 1983). The last column of Table 1 is an attempt to minimize the effects of entrainment and droplet coalescence that reduce N_c and thus obscure N_c-N_CCN relationships that usually establish N_c at cloud base. Such reductions of N_c could be independent of N_CCN , which would perturb the N_c-N_CCN relationship and thus also the relevance of S_eff, which is the most pertinent when it reflects the influence of N_CCN on N_c at cloud base. Therefore, penetrations with linear LWC altitude profiles that are characteristic of adiabatic unmixed clouds should show higher N_c-N_CCN R. Since they do not this indicates that entrainment does not seem to disrupt the N_c-N_CCN relationship_.

Table 2 shows mean S_eff within various N_1% intervals. This is consistent with row 5, Table 1 in showing that S_eff is lower for higher N_CCN. R for S_eff-N_c is also negative (with one small exception), but with smaller absolute values (Row 6, Table 1) than R for S_eff-N_1%. R for S_eff-N_c are less negative than R for S_eff-N_1% because both S_eff and N_c are functions of N_CCN.

W also determines N_c; i.e., higher W produces higher S and N_c. R for N_c-W is positive in row 3, Table 1 (except last column) but is much lower than R for N_c-N_1% (row 2). Variations of W thus contribute to the scatter of N_c-N_1% because W variations ought to be independent of N_1%. But N_1% and W are positively correlated (row 4; except last column). But since the effects of N_CCN and W on S_eff are opposite (higher N_CCN reduces S_eff whereas higher W increases S_eff) this positive albeit small coupling of N_CCN with W is probably the reason that S_eff does not display the expected positive correlation with W (last row, Table 1). Apparently this coupling allowed the greater effect of N_CCN variations on S_eff to overwhelm the effect of the W variations on S_eff. This means that the decrease of S_eff with N_1% shown in Table 2 could have been even greater had W been constant or independent of N_CCN.

Consistent with this is the fact that for the 17 adiabatic clouds W and N_1% were not correlated (R = -0.28; last column of row 4) and this probably contributed to the greater negative R of S_eff-N_1% for the adiabatic clouds compared to all 69 vertical clouds (0.84 versus 0.74) in spite of the more limited N_1% range of the adiabatic clouds, which thus prevented observations of the greatest reductions of S_eff caused by the highest N_1%.

Table 2 shows the decrease of cloud S with N_CCN and higher stratus cloud S than conventional wisdom especially since these are underestimates of cloud S. This means that much smaller particles can nucleate stratus clouds (e.g., 30 nm rather than 60 nm ammonium sulfate particles). This has important implications for the largest climate uncertainty—the indirect aerosol effect (Alley et al., 2007) and for geoengineering schemes to brighten marine stratus clouds. Smaller particles are easier to produce either deliberately or inadvertently.

Alley, R.B., and Coauthors, 2007: Cambridge University Press, 1-18.

Hegg, D.A., D.S. Covert, H.H. Jonsson, and R. Woods, 2009: Tellus, 61B, 669-676.

Hudson, J.G., 1983: J. Atmos. Sci., 40, 480-486.

Hudson, J.G., 1989:  J. Atmos. & Ocean. Techn., 6,  1055-1065.

Hudson, J.G. and P.R. Frisbie, 1991: J. of Geophys. Res., 96, D11, 20,795-20,808.

Twomey, S., 1959: Pure. Appl., 43, 243-249.

Table 1. Correlation coefficients (R) between various parameters. First row-number of clouds.

	all	HA	LSA	adiabatic
number	69	40	29	17
Nc-N1%	0.85	0.89	0.84	0.79
Nc-W	0.45	0.39	0.51	-0.23
N1%-W	0.39	0.30	0.53	-0.28
Seff-N1%	-0.74	-0.86	-0.35	-0.84
Seff-Nc	-0.35	-0.62	0.13	-0.41
Seff-W	-0.12	-0.06	-0.18	-0.57

Table 2. Mean S_eff within various average N_1% ranges.

N1% (cm-3)	< 100	100-200	200-300	300-400	> 400
mean Seff	0.87	0.51	0.46	0.43	0.11