Abstract:

In this work we use cluster analysis of long term particle size distribution data to expand an array of different shorter term atmospheric measurements, thereby gaining insights into longer term patterns and properties of Arctic aerosol. Measurements of aerosol number size distributions (9–915 nm) were conducted at Villum Research Station (VRS), Station Nord in North Greenland during a 5 year record (2012–2016). Alongside this, measurements of aerosol composition, meteorological parameters, gaseous compounds and cloud condensation nuclei (CCN) activity were performed during different shorter occasions. K-means clustering analysis of particle number size distributions on daily basis identified several clusters. Clusters of accumulation mode aerosols (main size modes > 100 nm) accounted for 56% of the total aerosol during the sampling period (89–91% during February–April, 1–3% during June–August). By association to chemical composition, cloud condensation nuclei properties, and meteorological variables, three typical accumulation mode aerosol clusters were identified: Haze (32% of the time), Bimodal (14%) and Aged (6%). In brief: (1) Haze accumulation mode aerosol shows a single mode at 150 nm, peaking in February–April, with highest loadings of sulfate and black carbon concentrations. (2) Accumulation mode Bimodal aerosol shows two modes, at 38 nm and 150 nm, peaking in June–August, with the highest ratio of organics to sulfate concentrations. (3) Aged accumulation mode aerosol shows a single mode at 213 nm, peaking in September–October and is associated with cloudy and humid weather conditions during autumn. The three aerosol clusters were considered alongside CCN concentrations. We suggest that organic compounds, that are likely marine biogenic in nature, greatly influence the Bimodal cluster and contribute significantly to its CCN activity. This stresses the importance of better characterizing the marine ecosystem and the aerosol-mediated climate effects in the Arctic.

Atmospheric Environment 183 (2018) 1–10. doi.org/10.1016/j.atmosenv.2018.03.060