Aerosol time-series measurements over the tropical Northeast Atlantic Ocean: Dust sources, elemental composition and mineralogy
Patey, M.D.; Achterberg, E.P.; Rijkenberg, M.J.; Pearce, R. (2015). Aerosol time-series measurements over the tropical Northeast Atlantic Ocean: Dust sources, elemental composition and mineralogy. Mar. Chem. 174: 103-119. dx.doi.org/10.1016/j.marchem.2015.06.004
In: Marine Chemistry. Elsevier: Amsterdam. ISSN 0304-4203; e-ISSN 1872-7581, more
|   |
| Author keywords |
Aerosol time-series; Cape Verde; Trace metals; Mineralogy; Dust; Aerosol; REEs; North Africa |
| Authors | | Top |
- Patey, M.D.
- Achterberg, E.P.
- Rijkenberg, M.J., more
- Pearce, R.
|
|
|
| Abstract |
The North Atlantic receives the largest dust loading of any of the world's oceans due to its proximity to North African deserts and prevailing wind patterns. The supply of biologically important trace elements and nutrients via aerosols has an important influence on biogeochemical processes and ecosystems in this ocean region. In this study we continuously sampled aerosols between July 2007 and July 2008 at the Cape Verde Atmospheric Observatory (CVAO), which is situated on an island group close to the North African continent and under the Saharan/Sahelian dust outflow path. The aim of our work was to investigate temporal variations in aerosol concentration, composition and sources in the Cape Verde region over a complete seasonal cycle, and for this purpose we undertook mineralogical and chemical (42 elements) analyses of the aerosol samples and air mass back-trajectory calculations. Aerosol samples were also collected during a research cruise in the (sub-) tropical Northeast Atlantic Ocean in January 2008.The concentration of atmospheric Al, a proxy for mineral aerosol concentration, at CVAO was in the range of 0.01–66.9 µg m- 3 (maximum on 28–30 January 2008) with a geometric mean of 0.76 µg m- 3. It showed distinct seasonal variations, with enhanced Al concentrations in winter (geometric mean 1.3 µg m- 3), and lower concentrations in summer (geometric mean 0.48 µg m- 3). These observations have been attributed to dust transport occurring in higher altitude air layers and mainly north of the Cape Verde during summer, while in winter the dust transport shifts south and occurs in the lower altitude trade winds with consequent greater influence on the Cape Verde region. The elemental composition of the aerosols closely agreed with mean upper crustal abundances, with the exception of elements with pronounced anthropogenic sources (e.g. Zn and Pb) and major constituents of sea water (Na and Mg). Mineral analysis showed that clays were the most abundant mineral fraction throughout the whole sampling period, with an increase in quartz and clays during strong dust events and an associated decrease in calcite. This could have important implications for the estimation of release of for example Fe from mineral dust with clays having a higher Fe solubility than quartz.The elemental composition and mineralogy of aerosol samples collected during the cruise were indistinguishable from those collected at the CVAO during the same period, although mean atmospheric Al was 65% higher at the CVAO than those measured on the ship due to the irregular and uneven nature of dust transport.Air mass back-trajectories showed an important role for southern source regions of the North African deserts during summer, with 92.5% of the samples indicating a contribution from the Sahel. Significantly elevated ratios of V, Ni, Cu, Zn, Cd and Pb with Al were present in samples originating from the Sahel compared with samples with a more northerly origin. This was likely due to enhanced anthropogenic emissions related to the greater population densities in the Sahel compared with the less developed Saharan regions further north.Ratios of other elements and trends in rare earth elements could however not be used to distinguish differences in source regions. Similar source material compositions, the mixing of dust from different regions during transport, and the pooling of samples over a 1–3 day collection period appear to have diluted specific signals from source regions |
|
