Published in

American Chemical Society, Environmental Science and Technology, 23(44), p. 8814-8820, 2010

DOI: 10.1021/es1022625

Links

Tools

Export citation

Search in Google Scholar

Assessing the Performance of Methods to Detect and Quantify African Dust in Airborne Particulates

This paper is available in a repository.
This paper is available in a repository.

Full text: Download

Green circle
Preprint: archiving allowed
  • Must obtain written permission from Editor
  • Must not violate ACS ethical Guidelines
Orange circle
Postprint: archiving restricted
  • Must obtain written permission from Editor
  • Must not violate ACS ethical Guidelines
Red circle
Published version: archiving forbidden
Data provided by SHERPA/RoMEO

Abstract

African dust (AD) contributions to particulate matter (PM) levels may be reported by Member States to the European Commission during justification of exceedances of the daily limit value (DLV). However, the detection and subsequent quantification of the AD contribution to PM levels is complex, and only two measurement-based methods are available in the literature: the Spanish-Portuguese reference method (SPR), and the Tel Aviv University method (TAU). In the present study, both methods were assessed. The SPR method was more conservative in the detection of episodes (71 days identified as AD by SPR, vs 81 by TAU), as it is less affected by interferences with local dust sources. The mean annual contribution of AD was lower with the TAU method than with SPR (2.7 vs 3.5 ± 1.5 μg/m(3)). The SPR and TAU AD time series were correlated with daily aluminum levels (a known tracer of AD), as well as with an AD source identified by the Positive Matrix Factorization (PMF) receptor model. Higher r(2) values were obtained with the SPR method than with TAU in both cases (r(2) = 0.72 vs 0.56, y = 0.05x vs y = 0.06x with aluminum levels; r(2)=0.79 vs 0.43, y = 0.8x vs y = 0.4x with the PMF source). We conclude that the SPR method is more adequate from an EU policy perspective (justification of DLV exceedances) due to the fact that it is more conservative than the TAU method. Based on our results, the TAU method requires adaptation of the thresholds in the algorithm to refine detection of low-impact episodes and avoid misclassification of local events as AD.