NOAA/CSIRO Flask Air Intercomparison Experiment: A strategy for directly assessing consistency among atmospheric measurements made by independent laboratories

Since 1992, the National Oceanic and Atmospheric Administration (NOAA), Climate Monitoring and Diagnostics Laboratory of the United States and the Commonwealth Scientific and Industrial Research Organization (CSIRO), Division of Atmospheric Research of Australia have routinely analyzed the same air sample collected biweekly at Cape Grim, Tasmania. Comparisons of measurements of atmospheric CO2, CH4, CO, H-2, and the stable isotopes of CO2 (delta C-13, delta O-18) are used to assess the consistency of observations independently made by the two laboratories. Results demonstrate that conventional intercomparison strategies based on occasional exchange of high-pressure cylinders are not sufficient to ensure adequate comparability between flask data records. The NOAA/CSIRO flask air intercomparison experiment has higher time resolution and captures artifacts that are specific to flask measurements. This ongoing experiment provides a stringent quality control test of individual laboratories' experimental methods and internal calibration schemes, and is a valuable tool in working toward reliable integration of atmospheric data from independent laboratories. Results from the first 7 years of this intercomparison show that NOAA and CSIRO Cape Grim measurement records of CH4 are consistent to within 1 nmol mol(-1) (0.04%) and the CO2 records to similar to0.2 mu mol mol(-1) (0.06%). CH4 and CO2 observations from the two programs may be combined into more extensive data sets for specific purposes. Several causes of the observed differences in the measurements of delta C-13 and delta O-18 of CO2 have been identified and independently corroborated; combined data sets of these isotopomers appear possible. Possible explanations are provided for the observed differences in measurements of CO and H-2 from the same sample, which show significant variability with time.