Evaluation of scatter corrections for ac-9 absorption measurements in coastal waters

Röttgers, Rüdiger and McKee, David and Woźniak, Sławomir B. (2014) Evaluation of scatter corrections for ac-9 absorption measurements in coastal waters. Methods in Oceanography. ISSN 2211-1220

Full text not available in this repository.Request a copy from the Strathclyde author

Abstract

Determinations of inherent optical properties of natural waters are fundamental in marine optical research. In situ measurements of light absorption are mostly obtained with an instrument that uses a reflective tube design to reduce concomitant errors induced by light scattering (ac-9, WETLabs Inc.). The remaining, generally still substantial, error is commonly corrected using one of a number of different approaches, each of which is based on a set of assumptions. Until now, the errors in these measurements have only been theoretically examined using Monte Carlo modeling Leymarie et al. (2010). The study presented here used a lab-based point source integrating cavity absorption meter (PSICAM) which avoids scattering errors. The PSICAM data were used to evaluate the absorption determination with an ac-9 in coastal waters for each of the scattering correction approaches. The results showed that the assumption of negligible absorption at wavelengths >700 nm is not valid in coastal waters and that, as a result, ac-9 measurements strongly underestimate absorption at longer wavelengths (>600 nm). An empirical relationship between uncorrected (for scattering) ac-9 measurements and the true absorption at 715 nm was included in the correction scheme; this improved the quality of ac-9 data at longer wavelengths but showed overestimation at shorter wavelengths. However, additional inclusion of a scatter correction for the ac-9 attenuation measurement resulted in a significant improvement of the proportional scatter error correction across the spectrum. Despite these innovations, variations in scattering properties can, combined with low absorption at specific wavelengths, result in relatively large percentage errors for individual measurements.