Source apportionment for African dust outbreaks over the Western Mediterranean using the HYSPLIT model

M. Escudero, A. F. Stein, R. R. Draxler, X. Querol, A. Alastuey, S. Castillo, A. Avila

    Research output: Contribution to journalArticleResearchpeer-review

    51 Citations (Scopus)

    Abstract

    A source apportionment technique has been applied to determine the geographical distribution of emissions in Northern Africa contributing to dust outbreaks that yield high PM10 levels at Spanish regional background stations. Seven dust episodes have been analyzed in this study. Total suspended particles have been sampled and chemically analyzed for these events at La Castanya background station (Montseny, NE Spain) and differences in the composition of airborne dust have been studied. The dominant role of northern and western source areas (Tunisia, Algeria, Mauritania and the Western Sahara) contrasted with the negligible contribution of major emission source areas such as the Bodelé depression, Libya, Niger, and Sudan. During the simulated events using the dust module of the HYSPLIT model, material from the latter regions is persistently transported across the Atlantic but not towards Western Europe. As a consequence, the composition of the dust turned out to be quite homogeneous since the mixing of dust occurs from various source areas with similar chemical composition. However, differences in Ca/Al ratios have been found in a number of samples that are mainly explained by vertical transport segregation of clay minerals (relatively richer in Al) from coarser dust particles (Ca-carbonate). © 2010 Elsevier B.V.
    Original languageEnglish
    Pages (from-to)518-527
    JournalAtmospheric Research
    Volume99
    Issue number3-4
    DOIs
    Publication statusPublished - 1 Mar 2011

    Keywords

    • Air quality
    • Crustal aerosols
    • Modeling
    • Particulate matter

    Fingerprint

    Dive into the research topics of 'Source apportionment for African dust outbreaks over the Western Mediterranean using the HYSPLIT model'. Together they form a unique fingerprint.

    Cite this