Here we apply new analysis methods and approaches to existing long-term measurement series that provide additional insights into the atmospheric processes that control black carbon (BC) in the Arctic. Based on clustering size distribution data from Zeppelin Observatory for the years 2002–2010, observations classified as ‘Polluted’ were further investigated based on BC properties. The data were split into two subgroups, and while the microphysical and chemical fingerprints of the two subgroups are very similar, they show larger differences in BC concentration and correlation with the particle size distribution. Therefore, a source–receptor analysis was performed with HYSPLIT 10-days backward trajectories for both subsets. We demonstrate that within this ‘Polluted’ category, the airmasses that contributed to the largest BC signal at the Zeppelin station are not necessarily associated with traditional transport pathways from Eurasia. Instead, the strongest signal is from a region east of the Ural Mountains across the continent to the Kamchatka Peninsula.

The aim of this study was to explore particle size dependent properties by combining long-term observations of equivalent black carbon (eBC) and number size distributions to investigate their correlation as function of particle size. The work was conducted in two main parts. The first part consisted of a short laboratory experiment to compare observed total particle light absorption (σ_abs) with that observed according to particle size by using a combination of a Differential Mobility Analyzer (DMA) and a Particle Soot Absorption Photometer (PSAP). The laboratory study confirmed strong similarities between the observed and derived σ_abs. In the second part the statistical approach using correlation between the σ_abs and the dN of each bin of the number size distribution was tested on long-term data ranging from 2002 to 2010 observed at Zeppelin station, Ny-Ålesund Svalbard. The data was clustered according to the number size distribution and grouped in four major categories: Washout, Nucleation, Intermediate and Polluted. Each category presented different features with respect to the derived eBC mass distributions, the Intermediate category showed similarities to the few available Single Particle Soot Photometer (SP2) observations in the Arctic. Overall, the statistical distribution of eBC, according to particle size, presented a larger dynamical range in the location of the mode(s). To check for consistency, the eBC mass distributions were transformed into number based eBC size distribution and compared to the observed total number size distribution. Whereas the Washout, Nucleation and Intermediate categories presented plausible number distributions, the Polluted category displayed a mode at small sizes (about 50 nm) that was significantly exaggerated.

Black Carbon (BC) aerosols are known to play an important role in the Arctic, yet their exact contribution to thechanging of the Earth’s climate and Arctic amplification remains unclear. To reduce these uncertainties, the life cycle of BCneeds to be accurately described in general circulation models (GCMs). In this study, four GCMs (ECHAM6.3-HAM2.3,ECHAM6.3-HAM2.3-P3, ECHAM6.3-HAM2.3-SALSA2 and UKESM1.0) are compared in terms of their representation ofBC in the Arctic. A new Lagrangian framework is applied to investigate the history of airmasses reaching the Arctic observationalsite Zeppelin on Svalbard, and compared to the corresponding transport simulated by the GCMs, which are allnudged to reanalysis data from ERA-Interim. Aerosol processes along the trajectories are then compared between the models.ECHAM6.3-HAM2.3-P3 simulates the highest and UKESM1.0 the lowest BC loadings both globally and within the Arcticand ECHAM6.3-HAM2.3-SALSA2 is the GCM that reproduces the observations from Zeppelin station most faithfully. The BC concentration in the Arctic is largely controlled by the wet removal processes described in the models, but dry depositionalso plays a role in explaining some of the inter-model diversity. ECHAM6.3-HAM2.3-P3 is less efficient in wet removal thanthe other models, which is likely a result of an adjusted representation of ice processes compared with the other two ECHAMvariants. UKESM1.0 is the most efficient model in removing BC from the atmosphere, in large part due to more efficient dryremoval than with the ECHAM models. The Lagrangian analysis reveals that the BC concentrations at the Zeppelin station are largely determined by concentrations in airmasses older than the length of our back trajectories, i.e. ten days, highlighting theimportance of remote emissions to local BC concentrations in the Arctic. This further suggests a longer BC lifetime within theArctic as compared with the global average. Our results underline the importance of accurate descriptions of cloud and precipitation microphysics, along with realistic dry and wet scavenging schemes for improved descriptions of BC and its climateimpacts in the Arctic within GCMs.