Intercomparison of airborne and surfacebased measurements of condensation nuclei in the remote marine troposphere during ACE 1

Intercomparisons of aerosol particle number concentrations measured with various condensation particle counters (CPCs) during the first Aerosol Characterization Experiment (ACE 1) are made to assess the accuracy of the airborne measurements. When no ambient 3-10 nm diameter particles (nanoparticles) were present, median concentrations from four CPCs on the National Center for Atmospheric Research (NCAR) C-130 aircraft agreed to within-6% and were highly correlated (r > 0.9). These instruments sampled from several different inlets and used various arrangements (e.g., tubing size and length, flow rates) to transport sampled air to the detectors. When the ambient aerosol contained significant numbers of nanopartieles, agreement between the CPCs deteriorated, likely from differences in nanoparticle transmission a d detection efficiencies. During these periods, average total number concentrations measured by two ultrafine CPCs varied on average by 60% with a correlation coefficient of 0.85. Intercomparisons of airborne and surfacebased measurements were made during low-altitude fiybys of surface measurement sites. During fiybys, few nanopartieles were detected, and measured total condensation nuclei (CN) concentrations differed by roughly +5 to -20% (CNsurfa•/CNairbom•' 1) suggesting that the airborne measurements of fine aerosols agreed with ambient surface values to within 20%. Overall, we found that ambient fine particles (-20-100 nm diameter) are fairly insensitive to airborne sampling techniques and thus are likely to be measured accurately. Sampling losses of smaller particles, however, may lead to a significant undermeasurement of ambient CN concentrations in rare instances when the number spectra re dominated by the smallest particles. This intercomparison involves measurements made in regions absent of liquid water to avoid artifact particles produced by fragmenting water droplets.