The reconstruction of aerosol light absorption by particle measurements at remote sites: An independent analysis of data from the IMPROVE network — II

Abstract

The author's mutual validation of the IMPROVE measures of light absorption — the light absorption coefficient σ a and the TOR carbon measures — at remote sites in the western United States, has identified more light-absorbing carbon (LAC) than the current interpretation of TOR admits. Further comparison of σ a with the new determination of LAC allows us to identify fine soil as the remaining significant contributor to light absorption at these remote sites, and thus to fully reconstruct σ a there. This reconstruction also confirms the accuracy of the blank corrections to the carbon measurements. Using σ a or the new reconstruction of it given herein as the appropriate measure of light absorption allows more correct reconstructions of aerosol light extinction σ e and of organic mass; the latter provides evidence that the newly identified LAC is also essentially elemental carbon (EC). The new interpretation of the TOR carbons for the remote western sites also reveals apparently much less pyrolysis than previously though occurring during TOR analysis, for most of the aerosol samples collected at these sites. A very small minority population, comprising less than 5% of the samples and occurring mostly in the summer and autumn, is also identified, containing a larger proportion of supposed pyrolyzable organics. The differences in apparent makeup between the two populations strongly suggest that the majority population represents a widespread background of aerosol light absorption which averages 5 Mm −1 and is probably due primarily to diesel fuel emissions transported from urban areas and highways, while the minority population is probably due to wood fires. A number of possible explanations are offered for why the newly identified EC is not currently recognized in the TOR analysis. In particular, it is claimed that sample darkening during thermal analysis is not a reliable quantitative indication of pyrolyzable organics, particularly in the remote aerosols considered; and that optical monitoring of the sample during thermal analysis should be corrected for expected strong multiple scattering effects, as the IMPROVE measurement of σ a is already corrected. It is also hypothesized that internal mixing of some EC with an oxygen-containing species, most probably sulfate, during aerosol transport is the source of the oxygen that allows some of the EC to evolve at a lower temperature in TOR analysis than previously thought, and in a pure helium atmosphere. Further tests are suggested to confirm the new interpretations offered here. The primary importance of the present results, besides unifying a number of IMPROVE analyses and removing major discrepancies in the data, is to highlight what the author believes are the two most important measurement problems in aerosol research today: (1) the failure to recognize and eliminate multiple scattering effects in the optical measurement of light absorption and in optical monitoring of a sample during thermal carbon analysis; and (2) the failure to accurately distinguish between light-absorbing and non-light-absorbing carbons in thermal analysis.