10th International Aerosol Conference September 2 - September 7, 2018 America's Center Convention Complex St. Louis, Missouri, USA
Abstract View
Quantitative Comparison of Correction Algorithms Applied Filter-Based Black Carbon Measurements during the FIREX Campaign
HANYANG LI, Gavin McMeeking, Andrew May, The Ohio State University
Abstract Number: 882 Working Group: Carbonaceous Aerosol
Abstract Light absorption by black carbon (BC) particles in the atmosphere is predominantly measured using filter-based techniques. In these instruments, light is transmitted through a filter; changes in transmission are quantified as “attenuation”, which is then related to light absorption. However, measured light attenuation may be overestimated or underestimated due to filter-loading artifact during sampling and light scattering artifact by filter fibers and embedded aerosols. Consequently, several correction algorithms (derived based on ambient aerosols where the concentration was roughly constant) have been introduced to minimize such artifacts. Moreover, to convert absorption to mass concentration, a mass absorption coefficient (MAC) is required. Hence, another uncertainty associated with the quantification of filter-based BC is the selection of an appropriate MAC. While it is widely accepted that applying correction algorithms and site-specific MACs improve the agreement among different filter-based BC instruments, it is uncertain whether any correction algorithm and MAC can be regarded as ‘benchmarks’ as both factors may depend on properties of aerosols and BC concentrations. These uncertainties may be even larger for biomass burning aerosols, which are inherently different (chemically and optically) than most typical ambient aerosols.
Experiments were conducted during 2016 FIREX Campaign to investigate BC emissions from simulated wildfires in the laboratory including three filter-based absorption measurements: a 7-wavelength aethalometer (Aeth-7), a Continuous Light Absorption Photometer (CLAP), and a Tricolor Absorption Photometer (TAP). In this presentation, we will summarize MACs of filter-based absorption measurements during the Campaign and quantitatively compare four published Aeth-7 correction algorithms and two Particle/Soot Absorption Photometers (PSAP) corrections (applicable to CLAP and TAP). While evaluations of correction schemes have been conducted in the past, none have focused on biomass burning smoke (and none have systematically compared schemes applied to CLAP and TAP); thus, our dataset allows us to explore how well these correction algorithms compare to a reference in situ light absorption measurement (a Droplet Measurement Techniques (DMT) Photoacoustic Extintiometer (PAX)).
Our analysis indicates that MAC strongly depends on aerosol properties. For example, an increase in MAC is observed due to the presence of brown carbon. Furthermore, filter-loading and light scattering artifacts can be clearly identified after a filter spot change, especially during fires with low single scattering albedo (relative to the atmosphere). Finally, applying correction algorithms results in high correlations between instruments. Although we cannot be certain that any correction algorithm gives us an unequivocally correct value, we will assess the degree of agreement between different algorithms, accounting for differences in chemical and optical properties. We are continuing to explore the role of the aerosol optical properties of biomass burning on differences among correction algorithms in order to provide a robust set of recommendations that can be used to constrain the uncertainty of filter-based BC measurement techniques and improve emission inventory of BC emission factors.