Currently GLOBALVIEW-CO2C13 is based solely on measurements made by the University of Colorado (CU), Institute of Arctic and Alpine Research (INSTAAR), Stable Isotope Laboratory (SIL). This section describes the INSTAAR in-house standards and ongoing efforts to assess the comparability of calibration scales and atmospheric observations with other labs.
Extraction of carbon dioxide from samples of atmospheric air is accomplished cryogenically using methods described along with the dual inlet mass spectrometry in Trolier et al., 1996, and Vaughn et al., 2004. Standard isotopic notation is used for CO2C13 data that are reported as 'delta' values relative to a standard, in units of 'per mil' (parts per thousand, or ‰. The 'delta' notation is for 13C can be written as
δ13C = [ (13C/12C) sample / (13C/12C) reference - 1 ] x 1000
where the ratios are the ion beam currents measured on the mass spectrometer, the 'sample' is the unknown CO2 gas and the 'reference' CO2 gas is a bellows volume that is used to calculate a delta value of many unknown samples relative to the reference gas. Samples with known delta values are also run against the reference gas, thereby allowing all of the unknown samples to be reduced and expressed in parts per mil, relative to a the VPDB-CO2 (Vienna -Pee Dee Belemnite) carbonate scale (Coplen, 1995). Corrections to this value are also made to compensate for the undesired co-ionization of N2O and 17O which interfere with the CO2 signal (Assonov and Brenninkmeijer 2003; Brand et al., 2009).Standard Scale
Previously measured high-pressure cylinders of compressed whole-air standards are analyzed in the same analysis run as samples, to scale the measurements of unknown atmospheric CO2, to known values of the cylinders. Original subsets of the cylinders (currently some 20 in all) have been calibrated relative to international standards of NBS-19 and NBS-20. Subsequent cylinders are calibrated as secondary reference materials to the originals. Monitoring of the scale is accomplished with an array of compressed air cylinders with a range of isotopic values that are measured every 6 months.
The cylinders themselves are obtained from the NOAA ESRL Global Monitoring Division and are filled to a pressure of 2000 psi in aluminium tanks (LUXFER size AL150, bare aluminium with brass pack-less valves, Scott Marrin, Riverside, CA, USA). The cylinders are filled from a clean air site at Niwot Ridge, Colorado, USA and the details of the preparation of these standards can be found at http://www.esrl.noaa.gov/gmd/ccgg/refgases/airstandard.html.
The absolute value of the δ13C scale used for the cylinder values is only as accurate as the determinations of NBS-19 carbonates, to which they have been calibrated. While a few labs with meticulous attention to detail have achieved δ13C values from carbonates repeatable to ±0.02‰ historically carbonate determinations have been limited in most labs to ~ ±0.03‰ or worse (Huang, 2003). This stems from a variety of reasons including vagaries of acid preparation, homogeneity of the carbonate, and problems associated with the error in extrapolating fractionation factors necessary for the high temperature (90°C) carbonate-acid reaction typically used. The target accuracy of the δ13C value derived from compressed cylinders of whole air may be ±0.015‰, but the scale upon which they are based is known only to ±0.03‰ at best.
Given that constraint, efforts have been made to calibrate between labs, using a variety of comparison schemes including exchange of data from co-sampling at flask sites and flask intercomparisons, where labs analyze the same flask sample on a routine basis. Additionally, all compressed air cylinder standards are linked to other labs via the inter-calibration of 6 "Classic" high-pressure cylinders that rotate through 4 different isotope labs through out the world: A) Commonwealth Scientific and Industrial Research Organization - Atmospheric Research, (CSIRO) Aspendale, Victoria, Australia; B) Institute of Arctic and Alpine Research (INSTAAR) University of Colorado, Boulder, Colorado, USA and National Oceanic and Atmospheric Administration, Earth System Research Laboratory (ESRL), Global Monitoring Division (GMD), Boulder, Colorado, USA; C) Scripps Institution of Oceanography (SIO) University of California, San Diego, California, USA; and D) Center for Atmospheric and Oceanic Studies, Tohoku University (TU), Sendai, Japan. Inter-comparison of flask measurements is also done by CarboEurope. These efforts towards tighter comparisons between labs is critical for both calibration and assimilation of global data obtained from different laboratories.