Interactions between Aerosol and Marine Stratocumuli Over the South East Pacific

Abstract

During the VOCALS Regional Experiment in 2008 fourteen research flights were conducted over the South East Pacific (SEP) to study the interaction between aerosol and cloud covering a wide range of conditions from a more polluted coastal troposphere to the remote marine environment. Size-resolved aerosol physiochemistry was measured above and below stratocumulus (Sc) clouds along with in-cloud observations of cloud droplet size spectra. Because the nature and variability of aerosol effective as cloud condensation nuclei (CCN) impact cloud microphysical properties through cloud droplet number concentration (Nd), size-resolved aerosol measurements were utilized to establish air mass characteristics and their CCN activity over the SEP. The aerosol number size distribution measurements were clustered into groups using a simple k-means technique that recognizes similar pattern over the aerosol size range assessed. This cluster technique yields four distinct aerosol size distributions for the free troposphere (FT). These are attributed to two local coastal pollution sources and long-range transport of aerosol from the South Pacific and Australia based upon back trajectory analysis and investigation of aerosol physiochemistry. Marine boundary layer (MBL) observations reveal six distinct clusters associated with different stages of aging and processing of coastal combustion sources, clean South Pacific and heavy drizzling air masses. All air masses show CCN activity is strongly dependent on aerosol number concentration and size distribution shape, while aerosol hygroscopicity plays a smaller role. This confirms earlier studies although observed MBL hygroscopicity considerably exceeds the previously recommended value of 0.7. Derived MBL CCN also reveal a 1:1 relationship to Nd over the range of air mass characteristics observed once droplet concentrations are corrected for instrumental artifacts that tend to undercount Nd in polluted clouds. This is contrary to some previous estimates of aerosol-cloud interaction and could result in changes in local cloud radiative forcing of -3 to -10 W m−2. In-cloud measurements from VOCALS campaign also show a robust dependency of Nd with the empirical correction factor k∗ utilized in cloud radiative transfer codes in climate models to account for droplet spectral properties. This relationship can be traced to aerosol size distributions below the clouds. Measurements for both clean marine and pollution influenced aerosol populations indicate that as they undergo cloud processing, reducing the number of CCN and Nd, their droplet mean radius (rµ) increases while spectrum width (rσ ) is unaffected. The associated k∗ increases, as it is roughly proportional to rµ / rσ . If this dependency is not accounted for, local forcing could be overestimated by 3 to 6 W m−2 in polluted clouds close to the Chilean coastline. In-cloud measurements also showed that the highest drizzle rates occur in the absence of typical pollution indicators as carbon monoxide and black carbon.