NOTICE (March 2018): This website provides access to the CDIAC archive data temporarily. It will be gradually transitioned into data packages in the new ESS-DIVE archive. This site will continue to operate in parallel during and after the transition, and will be retired at a future date. If you have any questions regarding the data or the transition, please contact

image image image image

Atmospheric Carbon Monoxide Mixing Ratios NOAA Climate Monitoring and Diagnostics Laboratory Cooperative Air Sampling Network, 1988-1993



P. C. Novelli and K. A. Masarie

DOI: 10.3334/CDIAC/atg.db1011


Individual site files provide CO mixing ratios in parts per billion (ppb) (ppb = parts in 109 by mole fraction) based on measurements from the NOAA/CMDL Cooperative Air Sampling Network beginning 1988. Data are provided through June 1993 for stations at which the first sample was collected before July 1991. Below is a list of locations from which samples of air are collected and analyzed for CO. Data not included in this ftp directory may be obtained by contacting the project PI. A listing of all sites in the air sampling network used for CO measurements (as of 9/94) are given in the table below. The table includes the three letter code used to identify each site, the location of the site, latitude and longitude (in degrees and minutes), the altitude of the sampling location (in meters above mean sea level), the start and end dates for sampling CO, the country, and the cooperating agency. Shipboard sampling occurs every 5 degrees of latitude for two cruise tracks: the Wellington Star (PAW), and California Star (OPC) between North America and Australia or New Zealand, and every 3 degrees latitude between Hong Kong and Singapore for Carla A. Hills and Great Promise (SCS) (data not yet available). Sampling frequencies are approximately weekly for the fixed sites, and average one sample every 3 weeks per latitude zone for PAW and OPC, and about one sample every week per latitude for SCS. The air samples are collected by two general methods: flushing and then pressurizing glass flasks with a pump or opening a stopcock on an evacuated glass flask. (See Lang et al., 1990a,b for details of the sampling network, the equipment and procedures.) During each sampling event a pair of flasks is filled in series.

All samples were analyzed for CO at the NOAA/CMDL laboratory in Boulder by gas chromatography with mercuric oxide reduction detection, and all measurements are referenced to the CMDL CO scale (Novelli et al., 1991, Novelli et al., 1994).

Through most of the period 1988-1991, one flask of a sample pair was analyzed for CO, and when there was suitable pressure remaining in the flask, two aliquots or more were analyzed. We have used the difference in CO mixing ratio between the two aliquots as an indication of the precision of the measurement. Details of the analytical procedures through 1990 are described in Novelli et al., 1992. From the beginning of CO measurements in flask samples (mid 1987) to December 1990 we used a single point calibration sequence as the response characteristics of the CO instrument had been shown to be linear over a range of 0 to 1000 ppb CO (Novelli et al., 1991).

In January 1991, we changed the instrument which was used for the analysis of CO in flask air samples. The new instrument exhibited a non-linear response over the range of 0 to 250 ppb CO. Therefore it was necessary to change from a single-point calibration routine to a multiple standard calibration scheme ( a multi-point calibration procedure based upon that used for calibration of standards, as described in Novelli et al., 1994, was used for flask analysis).

In October 1991 we began analyzing a single aliquot from both members of a flask pair (rather than 2 or more aliquots from a single flask). The principle reasons for the change were to simplify flask handling procedures (the CMDL carbon dioxide and methane projects also measure both flasks of the sample pair) and to have flask pair agreement (the difference in mixing ratio between the two flasks collected simultaneously) as an additional diagnostic to use in evaluating the quality of the data.