Chemical and physical characterization for source apportionment of multi-year arctic aerosol records

The work was mainly focused on identifying and assessing the relative contribution of the sources and transport processes delivering aerosol to the Arctic as well on the study of aerosol production (processes of New Particle Formation). Such aims were pursued thanks to the availability a large dataset from sampling and direct measurements (size distribution) spanning a relatively large temporal frame (six years), an accurate optimization of the analytical methods applied to the chemical characterization (with special attention to Rare Earth Elements) and devoted and advanced statistical tools (Positive matrix factorization and Potential Source Contribution Function) to apportion the various sources of atmospheric particulate. During six consecutive spring-summer Arctic sampling campaigns, between 2010 and 2015, particle size-distribution measurements and aerosol sampling for chemical analysis (PM10 and size-segregated aerosol) were carried out at Ny Ålesund (Svalbard Islands). PM10 was collected at daily resolution, while a multistage impactor with a 4-day resolution was used to obtain the aerosol size distribution of the chemical characterization. Samples were then analyzed by Ion Chromatography (18 ions) and Inductively Coupled Plasma - sector field Mass Spectrometry (32 metals). The results were discussed to identify and quantify the contribution of the main aerosol sources, in addition to their seasonal trend and source region. Measurements of aerosol number concentration show that the seasonal cycle is linked to the Arctic Haze, with accumulation mode dominating during April and May, and the Aitken mode dominating during the summer months. The high occurrence of NPF events can be considered the result of both photochemical production of nucleating/condensing species and low condensation sink. From the direct observation of the temporal evolution of the main chemical markers concentrations, it was possible to analyze the main sources of the Arctic aerosol, both natural and anthropic. The size distribution of sea salt Sodium (ss-Na+), as a univocal marker of sea spray, shows that sea spray particles are mostly distributed in the 1.0-10μm range. The temporal trend of methanesulfonate (MSA), the univocal biogenic marker, suggested the occurrence of two periods in which the phytoplankton activity was particularly intense, with maximum values between April-July. Sulfate was one of the dominant species in the Ny Ålesund aerosol and it is present mostly in the sub-micrometric range, confirming that secondary sources (via atmospheric oxidation of SO2 and DMS atmospheric oxidation) are dominant with respect to primary emissions from sea spray and crustal scraping. Sea salt aerosol contributes in the sulfate amount for about 20 % in summer, while spring values were lower (12.0 %). The crustal fraction is always low, with a relatively high contribution in summer (5.5 %). Biogenic emissions are the main sulfate source from early June to early August, accounting for 35.0% in the summer samples. The anthropogenic fraction showed a clear seasonal pattern, being dominant in spring up to the end of May (75 % of the total sulfate budget), probably due to Arctic haze transport phenomenon. Anthropogenic sulfate is long-range transported from lower-latitude industrialized areas, and its acidic form is the most probable agent for chloride removal (as HCl) in aged sea-spray particles. The chloride depletion, with respect to seawater composition, was particularly evident in the sub-micrometric fraction and in spring, when H2SO4 is not completely neutralized by ammonia and free acidity is available. In this work, we also demonstrated the applicability of a method based on an ICP-MS system for the quantification of metals in samples of Arctic particulate matter. The coupled desolvation system allows reducing the presence of isobaric interferences, which are mainly due to oxide formation in the plasma, with a significant gain in terms of signal. The performances of this method were studied in particular for the determination of Rare Earth Elements, which show for most of the cases recoveries larger than 80 %. The study of REEs correlations in the samples reveals the partition of the group between light REEs (from La to Sm) and heavy REEs (from Gd to Lu) and an additional weaker partition based on the even or odd atomic number of the elements. The analysis of their temporal pattern shows an increase of the variability during the summer months for both the air concentration and the light/heavy REEs ratio. The Cerium and Europium anomalies appeared to be suitable tools for a preliminary sources identification. Most samples show no Cerium anomaly, negative Europium anomaly and high light/heavy REEs ratio. From spring to summer a decrease in Cerium anomaly and in LREE/HREE ratio was observed; on the contrary, Europium anomaly exhibited a clear increase. Such a temporal pattern suggests a different contribution of the local and long-range transported dust, which affect with a different extent the two seasons. Thanks to the large database of the chemical composition of the samples collected during 2015, two receptor models were applied: the Positive Matrix Factorization (PMF), for the source type identification, and the Potential Source Contribution Function (PSCF), which uses the air mass back-trajectories to recognize the source geographic regions of the particles reaching Ny Ålesund and bringing high concentrations of a chemical species. Seven factors were found through PMF, which apportioned 86% of the total PM10, whereas the chemical markers allowed to recognize the types of sources affecting the Arctic aerosols. Sea salt and soil are the natural sources of the primary Arctic aerosol and together represent the 54.4 % of PM10. The validity of the receptor model results is confirmed by the good correspondence of the average composition of seawater and upper continental crust with the chemical composition in the profile of the two natural factors. Both sources do not exhibit a clear seasonal pattern in the analyzed period. The main sources of marine aerosols are the seas south of the Svalbard Islands; as concerning the origin of the crustal input to PM in Ny Ålesund, it is mainly attributed to central Russia and the dry land facing towards the Svalbard Islands. The biological activity of the phytoplankton is the main natural source of secondary aerosols. It affects only 3.3% of the Arctic aerosol in terms of load but still gives a remarkable contribution to the atmospheric sulfate concentration (together with MSA), especially during spring and summer. Indeed, this aerosol source is present almost exclusively from April to September and, during these months, this type of aerosol reaches the sampling site from three different regions (starting from the Barents Sea to the North Atlantic Ocean), suggesting a different seasonality for the marine biological activity of the Arctic seas. Regarding the anthropic sources, heavy metals drive the primary aerosol fraction, which is 5.5 % of the total PM10 budget. The strong seasonality of this type of aerosol is clear from the recurring presence of high values until April, which concentrate most of its mass in this period (83%), whereas its presence is almost negligible since May. The main influence comes from the major industrial areas of Central and Western Russia. However, the temporal trend and the occurrence of air masses from eastern North America and the Arctic region between western Russia and northern Greenland, confirm how closely this type of aerosol is related to the Arctic haze. The same geographic apportionment result was obtained for the other two anthropic aerosol sources: ammonia-sulfates and nitrate, that account for 16.3 % and 5.8 % respectively. Indeed, both represent secondary aerosol and therefore it was impossible to ascribe a source region. The back-trajectory study can only confirm the permanence or formation of these species in the Arctic atmosphere, coming from long-range transport. Although the presence of nitrates does not have a clear seasonal trend during the sampling period, the ammonia-sulfate source shows a seasonality which is similar to that seen for heavy metals, with high values in spring and early summer, confirming the transport of a large amount of sulfates due to this atmospheric phenomenon. During the summer, there is still a small presence of ammonium and sulfates which, as seen above, are mainly present in the two forms of neutralization. The biomass burning source reaches 14.7 % of the total and has a high variability, linked to forest fire occurrences. A large aerosol event, which happened in July 2015 and lasted for a few days, made possible the characterization of this source, thanks to the connection of extensive forest fires in Alaska with fast air transport processes towards the Svalbard Islands. Oxalate, Potassium and elemental Carbon have proved to be reliable markers of this source, which can have a decisive weight on the Arctic environment, especially due to the large aerosol loads that it can emit into the atmosphere and the direct effect of black carbon on the atmospheric radiative balance.