Aerosol light absorption, black carbon, and elemental carbon at the Fresno Supersite, California

Abstract

Particle light absorption ( b ap), black carbon (BC), and elemental carbon (EC) measurements at the Fresno Supersite during the summer of 2005 were compared to examine the equivalency of current techniques, evaluate filter-based b ap correction methods, and determine the EC mass absorption efficiency (σ ap) and the spectral dependence of b ap. The photoacoustic analyzer (PA) was used as a benchmark for in-situ b ap. Most b ap measurement techniques were well correlated ( r ≥ 0.95). Unadjusted Aethalometer (AE) and Particle Soot Absorption Photometer (PSAP) b ap were up to seven times higher than PA b ap at similar wavelengths because of absorption enhancement by backscattering and multiple scattering. Applying published algorithms to correct for these effects reduced the differences to 24 and 17% for the AE and PSAP, respectively, at 532 nm. The Multi-Angle Absorption Photometer (MAAP), which accounts for backscattering effects, overestimated b ap relative to the PA by 51%. BC concentrations determined by the AE, MAAP, and Sunset Laboratory semi-continuous carbon analyzer were also highly correlated ( r ≥ 0.93) but differed by up to 57%. EC measured with the IMPROVE/STN thermal/optical protocols, and the French two-step thermal protocol agreed to within 29%. Absorption efficiencies determined from PA b ap and EC measured with different analytical protocols averaged 7.9 ± 1.5, 5.4 ± 1.1, and 2.8 ± 0.6 m 2/g at 532, 670, and 1047 nm, respectively. The Angström exponent (α) determined from adjusted AE and PA b ap ranged from 1.19 to 1.46. The largest values of α occurred during the afternoon hours when the organic fraction of total carbon was highest. Significant biases associated with filter-based measurements of b ap, BC, and EC are method-specific. Correcting for these biases must take into account differences in aerosol concentration, composition, and sources.