New in situ aerosol hyperspectral optical measurements over 300–700 nm – Part 2: Extinction, total absorption, water- and methanol-soluble absorption observed during the KORUS-OC cruise

Abstract

This two-part study explores hyperspectral (300–700 nm) aerosol optical measurements obtained from in situ sampling methods employed during the May–June 2016 Korea–United States Ocean Color (KORUS-OC) cruise conducted in concert with the broader air quality campaign (KORUS-AQ). Part 1 focused on the hyperspectral measurement of extinction coefficients (σext) using the recently developed in situ Spectral Aerosol Extinction (SpEx) instrument and showed that second-order polynomials provided a better fit to the measured spectra than power law fits. Two dimensional mapping of the second-order polynomial coefficients (a1, a2) was used to explore the information content of the spectra. Part 2 expands on that work by applying a similar analytical approach to filter-based measurements of aerosol hyperspectral total absorption (σabs) and soluble absorption from filters extracted with either deionized water (σDI-abs) or methanol (σMeOH-abs). As was found for σext, second-order polynomials provided a better fit to all three absorption spectra sets. Averaging the measured σext from Part 1 over the filter sampling intervals in this work, hyperspectral single-scattering albedo (ω) was calculated. Water-soluble aerosol composition from the DI extracts was used to examine relationships with the various measured optical properties. In particular, both σDI-abs(365 nm) and σMeOH-abs(365 nm) were found to be best correlated with oxalate (C2O42-), but elevated soluble absorption was found from two chemically and optically distinct populations of aerosols. The more photochemically aged aerosols of those two groups exhibited partial spectra (i.e., the longer wavelengths of the spectral range were below detection) while the less-aged aerosol of the other group exhibited complete spectra across the wavelength range. The chromophores of these groups may have derived from different sources and/or atmospheric processes, such that photochemical age may have been only one factor contributing to the differences in the observed spectra. The differences in the spectral properties of these groups was evident in (a1, a2) maps. The results of the two-dimensional mapping shown in Parts 1 and 2 suggest that this spectral characterization may offer new methods to relate in situ aerosol optical properties to their chemical and microphysical characteristics. However, a key finding of this work is that mathematical functions (whether power laws or second-order polynomials) extrapolated from a few wavelengths or a subrange of wavelengths fail to reproduce the measured spectra over the full 300–700 nm wavelength range. Further, the σabs and ω spectra exhibited distinctive spectral features across the UV and visible wavelength range that simple functions and extrapolations cannot reproduce. These results show that in situ hyperspectral measurements provide valuable new data that can be probed for additional information relating in situ aerosol optical properties to the underlying physicochemical properties of ambient aerosols. It is anticipated that future studies examining in situ aerosol hyperspectral properties will not only improve our ability to use optical data to characterize aerosol physicochemical properties, but that such in situ tools will be needed to validate hyperspectral remote sensors planned for space-based observing platforms.