Uncertainty in Light Scattering Measurements by TSI Nephelometer: Results from Laboratory Studies and Implications for Ambient Measurements

We performed a series of laboratory studies to assess uncertainties in aerosol optical property measurements. We used a pulsed cavity ring-down spectrometer (CRD) to measure the aerosol light extinction coefficient, a photo-acoustic absorption spectrometer (PAS) for light absorption, and a commercial integrating nephelometer (TSI, model 3563) for the scattering coefficient. The test aerosols generated for the study consisted of both non-absorbing and slightly-to-highly absorbing polydisperse submicron particles, delivered to the systems as pure compounds or mixtures. For aerosols with single scattering albedo (ω) > 0.7, we found that the ω values calculated by combining the three instruments agreed within expected uncertainty. For aerosols with ω < 0.7, the discrepancy was outside of the boundaries of existing uncertainties. Mie theory calculations of the nephelometer correction factors indicate that the current literature expressions for nephelometer truncation correction are not adequate at ω < 0.7 due to the large real part of the refractive index (n) of absorbing particles. New corrections are given for the case when n is known. Based on our instrumental capabilities we conclude that the lowest uncertainty in ω is obtained by combining CRD and PAS data. The implications for the uncertainty of aerosol optical properties when using nephelometer-based scattering coefficient in the case of, e.g., highly absorbing aerosols, coarse mode particles, and measurements at high relative humidity are discussed.