Uncertainty budgets for liquid waveguide CDOM absorption measurements

Ina Lefering*, Rüdiger Röttgers, Christian Utschig, David McKee

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

20 Citations (Scopus)
89 Downloads (Pure)

Abstract

Long path length liquid waveguide capillary cell (LWCC) systems using simple spectrometers to determine the spectral absorption by colored dissolved organic matter (CDOM) have previously been shown to have better measurement sensitivity compared to high-end spectrophotometers using 10 cm cuvettes. Information on the magnitude of measurement uncertainties for LWCC systems, however, has remained scarce. Cross-comparison of three different LWCC systems with three different path lengths (50, 100, and 250 cm) and two different cladding materials enabled quantification of measurement precision and accuracy, revealing strong wavelength dependency in both parameters. Stable pumping of the sample through the capillary cell was found to improve measurement precision over measurements made with the sample kept stationary. Results from the 50 and 100 cm LWCC systems, with higher refractive index cladding, showed systematic artifacts including small but unphysical negative offsets and high-frequency spectral perturbations due to limited performance of the salinity correction. In comparison, the newer 250 cm LWCC with lower refractive index cladding returned small positive offsets that may be physically correct. After null correction of measurements at 700 nm, overall agreement of CDOM absorption data at 440 nm was found to be within 5% root mean square percentage error.

Original languageEnglish
Pages (from-to)6357-6366
Number of pages10
JournalApplied Optics
Volume56
Issue number22
Early online date31 Jul 2017
DOIs
Publication statusPublished - 1 Aug 2017

Keywords

  • absorption
  • oceanic optics

Fingerprint

Dive into the research topics of 'Uncertainty budgets for liquid waveguide CDOM absorption measurements'. Together they form a unique fingerprint.

Cite this