CCN measurements during ACE-2 and their relationship to cloud microphysical properties

Measurements of cloud condensation nuclei (CCN) concentration at 0.1% super saturation were made onboard the CIRPAS Pelican over the northeast Atlantic during June and July, 1997, in the vicinity of Tenerife, Spain, as part of the second Aerosol Characterization Experiment (ACE-2). The average CCN co ncentration (N-ccn) in the marine boundary layer for clean air masses was 2 7 +/- 8 and 42 +/- 14 cm(-3) for cloudy and clear conditions, respectively, consistent with measurements made near the British Isles and close to Tasm ania, Australia, during ACE-1 for similar conditions. A local CCN closure e xperiment was conducted. Measured N-ccn is compared with predictions based on aerosol number size distributions and size-resolved chemical composition profiles determined from measurements and the literature. A sublinear rela tionship between measured and predicted N-ccn, N-ccn similar to N-ccn,predi cted(0.51), was found. This result is consistent with some previous studies , but others have obtained results much closer to the expected 1:1 relation ship between measured and predicted N-ccn. A large variability between meas ured and predicted N-ccn was also observed, leading to the conclusion that, for 95% of the data, the predictions agree with measurements to within a f actor of 11. Relationships between below-cloud N-ccn and aerosol accumulati on mode concentration, and in-cloud cloud droplet number concentration, mea sured onboard the Pelican and the Meteo-France Merlin-IV, respectively, are calculated For periods while the 2 aircraft were in close proximity at app roximately the same time. Measured relationships are reproduced by an adiab atic parcel model, and are also consistent with some previous studies. Howe ver, the shape of the CCN spectrum, or the aerosol size distribution, and t he updraft velocity are predicted by the model to affect these relationship s to a significant extent. Therefore, parameterizations of cloud microphysi cal properties need to include these variables to accurately predict cloud droplet number concentration. A relationship between N-ccn and cloud drople t effective diameter is also calculated and shown to be consistent both wit h the literature and with the parameterization of effective diameter propos ed by Martin et al.