Nonmethane Hydrocarbon Intercomparison Experiment (NOMHICE): Task 4, ambient air

The Nonmethane Hydrocarbon Intercomparison Experiment (NOMHICE) was designed to assess the accuracy and comparability of nonmethane hydrocarbon (NMHC) measurements from research groups around the globe. This is being accomplished by conducting a series of intercomparisons, called Tasks, of prepared mixtures or collected ambient air. This paper presents results for an ambient air challenge sample as part of the fourth NOMHICE installment (Task 4). Twenty‐three laboratories participated in Task 4, and 30 overall analytical results are compared. The air sample provided a wide dynamic range of mixing ratios (parts per trillion by volume (pptv) to parts per billion by volume (ppbv)) as well as large number of compounds (>100). Coelutions of NMHCs with other volatile organic compounds (VOCs) such as oxygenated VOCs are shown to make the analysis of whole air samples more challenging than for prepared standard mixtures. Individual canisters containing the air sample were prepared and analyzed by the NOMHICE group at the National Center for Atmospheric Research (NCAR‐NOMHICE), sent to participants for analysis, and reanalyzed upon return to NCAR‐NOMHICE. The mixing ratio of propylene increased by 10% (14 pptv) in the canisters with time, and some of the less volatile compounds decreased in the canisters with time; however, the majority of the compounds were stable throughout the experiment. Participants were asked to identify and quantify as many compounds as possible with their analytical techniques and to submit their results to NCAR‐NOMHICE scientists. Fifty‐four compounds were chosen for the intercomparison. Eight hundred eighty‐three measurements were compared overall; the average of the mean ratios of the participants' results to NCAR‐NOMHICE results was 1.03. The participants' results were combined and averaged for each individual NMHC measured and compared to the reference results; thirty‐three of the 54 compounds agreed to within ±20% of the reference results. Individual analyses from participant laboratories were compared and ranked with respect to agreement with the reference values; the ranking was reconciled with the analytical procedures employed for each analysis. From this, recommendations were derived for preferred analytical techniques and practice. Recommendations include but are not limited to the following: (1) National Institute of Standards and Technology (NIST) standards or NIST‐traceable standards should be used and, for mass spectrometric