August 31, 2017 Journal article Open Access

What do we learn from long-term cloud condensation nuclei number concentration, particle number size distribution, and chemical composition measurements at regionally representative observatories?

Bas Henzing; Roman Fröhlich; Ulrich Pöschl; Pasi Aalto; Minsu Park; Joel Brito; Joel Brito; Urs Baltensperger; Hartmut Herrmann; Erik Herrmann; Mikael Ehn; Arnoud Frumau; Rupert Holzinger; Meintrat O. Andreae; Martin Gysel; Atsushi Matsuki; Nikolaos Mihalopoulos; Nikos Kalivitis; Tuukka Petäjä; Erik Swietlicki; Adam Kristensson; Laurent Poulain; Göran Frank; John A. Ogren; Alfred Wiedensohler; Frank Stratmann; Samara Carbone; David Picard; Gerard Kos; Mira L. Pöhlker; André S. H. Prévôt; Julia Schmale; Paulo Artaxo; Aikaterini Bougiatioti; Christopher Pöhlker; Mikhail Paramonov; Mikhail Paramonov; Helmi Keskinen; Colin D. O'Dowd; Nicolas Bukowiecki; Seong Soo Yum; Jurgita Ovadnevaite; Iasonas Stavroulas; Markku Kulmala; Anne Jefferson; Karine Sellegri; Mikko Äijälä; Stefano Decesari; Silvia Henning; Athanasios Nenes; Athanasios Nenes; Yoko Iwamoto; Yoko Iwamoto; Patrick Schlag; Patrick Schlag

Abstract. Aerosol-cloud interactions (ACI) constitute the single largest uncertainty in anthropogenic radiative forcing. To reduce the uncertainties and gain more confidence in the simulation of ACI, models need to be evaluated against observations, in particular against measurements of cloud condensation nuclei (CCN). Numerous observations of CCN number concentration exist, and many closure studies have been performed to predict CCN number concentrations based on particle number size distributions, chemical composition, and the κ-Köhler theory. Most of these studies provide details for short time periods or focus on special environmental conditions. These observations, however, cannot address questions of large-scale temporal and spatial CCN variability. Here we analyze long-term observations of CCN number concentrations, particle number size distributions and chemical composition from twelve sites on three continents. Eight of these stations are part of the European Aerosols, Clouds, and Trace gases Research InfraStructure (ACTRIS). We group the observatories into categories according to their official classification: coastal background (Barrow, Alaska; Mace Head, Ireland; Finokalia, Crete; Noto Peninsula, Japan), rural background (Melpitz, Germany; Cabauw, the Netherlands; Vavihill, Sweden), alpine sites (Puy de Dôme, France; Jungfraujoch, Switzerland), remote forest sites (ATTO, Brazil; SMEAR, Finland) and the urban environment (Seoul, South Korea). Expectedly, CCN characteristics are highly variable across regions. However, they also vary within categories, most strongly in the coastal background group, where CCN number concentrations can vary by up to a factor of 30 within one season. In terms of particle activation behavior, most continental stations exhibit very similar relative activation ratios across the range of 0.1 to 1.0 % supersaturation. At the coastal sites the activation ratios spread more widely across the SS spectrum. Several stations show strong seasonal cycles of CCN number concentrations and particle number size distributions, e.g., at Barrow (Arctic Haze in spring), at the alpine stations (stronger influence of polluted boundary layer air masses in summer), the rain forest (wet and dry season), or Finokalia (forest fire influence in fall). The rural background and urban sites exhibit relatively little variability throughout the year while short-term variability can be high especially at the urban site. The average hygroscopicity parameter, κ, calculated from the chemical composition of submicron particles, was highest at the coastal site of Mace Head (0.6) and the lowest at the rain forest station ATTO (0.2–0.3). We performed closure studies to predict CCN number concentrations from the particle number size distribution and chemical composition measurements. The prediction accuracy for the average concentrations is high. The ratio between predicted and measured CCN concentrations is between 0.87 and 1.4. The temporal variability is also well represented, as reflected by Pearson correlation coefficients > 0.87. We also conducted a series of sensitivity studies for the ratio of predicted versus measured CCN concentration, where we varied the hygroscopicity parameter κ, and made simple assumptions for aerosol particle number concentrations and size distributions. Uncertain particle number concentrations and their size distributions significantly impair the accuracy in predicting temporal variability and hence of absolute concentrations, while the effect of uncertain κ values is limited to the predicted CCN number concentration. Information on CCN number concentrations at many locations is important to better characterize ACI and their radiative forcing. Long-term comprehensive aerosol particle characterizations are labor intensive and costly. For observatories where such efforts are out of scope to obtain nevertheless long-term information of CCN number concentrations, we recommend conducting collocated CCN number concentration and particle number size distribution measurements at individual locations throughout one year at least to derive a seasonally resolved hygroscopicity parameter. This way, CCN number concentrations can be calculated based on continued particle number size distribution information only. This approach is a good alternative to deriving kappa from time-resolved chemical composition measurements which are costly and may still not cover the appropriate size range. Additionally, given the variability in observations at sites of the same category, a certain density in spatial coverage of observations is needed, especially along coastlines. We recommend operating "migrating-CCNCs" at priority locations, identified by model evaluation, around the globe where long-term particle number size distribution data are already available.