Correcting micro-aethalometer absorption measurements for brown carbon aerosol

Abstract

Portable aethalometers are commonly used for online measurements of light-absorbing carbonaceous particles (LAC). However, they require strict calibration. In this study, the performance of a micro-aethalometer (MA200 with polytetrafluoroethylene filter) in charactering brown carbon aerosol (BrC) absorption was evaluated in comparison with reference materials and techniques that included bulk solution absorbance and Mie-theory based particle extinction retrieval via broadband cavity enhanced spectrometer (BBCES). Continuous-wavelength resolved (300–650 nm) imaginary refractive index (kBrC) was derived with these methods for various BrC proxies and standard materials representing a wide range of sources and absorbing abilities, including the strongly absorbing nigrosin, pahokee peat fluvic acid (PPFA), tar aerosol from wood pyrolysis, humic-like substance (HULIS) separated from wood smoldering burning emissions, and secondary organic aerosols (SOA) from photochemical oxidation of indole and naphthalene in the presence of NOx. The BrC and nigrosin optical results by bulk solution absorption are comparable with the properties retrieved from BBCES. The MA200 raw measurements provide reliable absorption Ångström exponent (AAE) but overestimate kBrC largely. The parameterized overestimates against reference methods depend on light absorption strength, so that the MA200 overestimates more for the less absorbing BrC. The correction factor for MA200 can be expressed well as an exponential function of kBrC or particle single scattering albedo (SSA), and also as a power-law function of the MA200 raw results derived BrC mass absorption efficiency (MAE). The ensemble correction factor regressed for all these BrC and nigrosin is 2.8 based on bulk absorption and 2.7 using BBCES result as reference. Simple radiative forcing (SRF) calculations for different scenarios using the correction for MA200, show consistent SRF when using the aethalometer results after the kBrC-dependent correction.