The validation and measurement uncertainty of an automated gas chromatograph for marine studies of atmospheric dimethylsulfide

Abstract

An automated gas chromatograph (GC) is described that can reliably quantify atmospheric dimethylsulfide (DMS) in near real-time at low nmol m⁻³ concentrations or pmol mol⁻¹ (ppt) mixing ratios. Features of the automated GC include: removal of atmospheric oxidants and moisture; cryogenic pre-concentration of DMS; methylethylsulfide internal standard calibration; and pulsed flame photometric sulfur specific detection. This automated instrument is suitable for field deployment; it was recently used to obtain a continuous DMS dataset at Heron Island in the southern Great Barrier Reef over 14 days during the austral summer. Detailed analysis of the measurement uncertainty of this automated GC was conducted, according to Eurachem/CITAC guidelines, by quantifying and combining the uncertainties of the components that contributed to the analytical result. This gave a relative standard uncertainty of 6.5% which was expanded, using a coverage factor of two, for an interval containing approximately 95% of the expected distribution of values. When applied to the Heron Island summer dataset, the mean atmospheric DMS concentration and expanded uncertainty was 3.9 ± 0.5 nmol m⁻³ (n = 651). This value and uncertainty interval falls within the range of mean atmospheric DMS values obtained from other studies at the Great Barrier Reef. Wider adoption of chemometrics to quantify atmospheric DMS measurement uncertainty will enable improved comparison of data and assist climate modelling.