Measurements Outside Scale Range

Updated February 2021

The WMO CO2 X2019 scale covers the nominal range 250 - 800 ppm. NOAA offers calibration services outside of this range for applications related to atmospheric monitoring, for example firn or ice core sampling (down to 125 ppm) and ocean pCO2 measurements (up to 3000 ppm).

Historically, NOAA kept a set of primary standards to cover CO2 values outside the core WMO scale range. These primary standards were value assigned by infrequent manometric measurements. The manometer was designed to perform best at near ambient atmospheric levels. The analysis of the primary standards well outside of the range of the scale often was less precise and potentially suffered from biases due to running the manometer under different conditions to accommodate the mole fractions these cylinders covered.

NOAA transitioned from the NDIR based calibration system (co2cal-1) to a new system based on laser spectroscopy (Tans et al., 2017) in 2016. Initially cylinders outside of the range of the scale were calibrated using the separate sets of primary standards. Cylinders calibrated against these separate standards were consistent with the core WMO range calibrations, but they were not traceable to the standards actually used to define the WMO CO2 X2019 scale.

In 2020, NOAA changed the procedure for calibrating cylinders targeted outside of the WMO scale range. We decided to extrapolate the calibration curves used to define the WMO scale for all measurements of cylinders outside the WMO scale range. There are a couple of reasons for wanting to do this. First, manometric determinations of the extended range standards involve significant effort. These primary standards have not been analyzed frequently enough on the manometer to establish tight relationships to the WMO scale. The effort to do so is not sustainable by the CCL. Second, running dedicated calibration curves significantly slows down and complicates the analysis work flow. The intended usage for standards outside of the range of the WMO scale is often for applications where the uncertainty of the calibrations is not a limiting factor and the WMO goal of transferring the scale with very low uncertainty isn't required. Third, from an analytical perspective, we believe the linearity of the CRDS analyzer used as part of the laser spectroscopic calibration system makes extrapolating the calibration curves a reasonable approach considering the intended usage.

One caveat is that the laser spectroscopic calibration system measures the three major isotopologues of CO2 (16O12C16O, 16O13C16O, and 16O12C18O) to fully account for isotopic differences among the standards and between the standards and the unknown sample tanks. We do not feel comfortable extrapolating the calibration curves of the minor isotopologues beyond the calibrated ranges. Therefore, we do not use the minor isotopologue measurements for cylinders outside of the WMO scale range to calculate total CO2 and we do not report δ13C and δ18O values for these cylinders. The total CO2 reported for these cylinders is based on the measured 16O12C16O mole fraction, corrected assuming the cylinder has near ambient isotopic composition (δ13C = -8.5‰ and δ18O = -1.5‰).

For perspective, a cylinder with a measured 16O12C16O mole fraction of 2500 ppm and δ13C = -35‰ and δ18O = -35‰ would show a bias of about 1.1 ppm if we calculated the total CO2 assuming δ13C = -8.5‰ and δ18O = -1.5‰. This is within the uncertainty of our manometric determinations of primary standards at this mole fraction and cylinders prepared by NOAA are generally not this depleted in 13C.

Table 1 lists results from a May 2020 test where we ran extended range primary standards, which have manometric assigned values and isotopic values based on IRMS measurement, on the extrapolated WMO range calibration curve and corrected assuming the near ambient isotopic content. The measured values agree well with the manometric values. We plan on continuing to measure these cylinders as target tanks to evaluate the system performance when extrapolating the calibration curves.

cylinder Manometric (ppm) Manometric std dev (ppm) δ13C (‰) δ18O (‰) Measured (ppm) Diff (ppm)
CA05380 124.950.26 -8.1 -6.9 125.10 0.15
CB11627 198.020.20 -8.5 -0.7 198.01-0.01
CA057901060.980.77-25.2-26.91061.29 0.31
CA083451291.480.22 -8.9-30.01291.22-0.26
CA038001554.490.86-31.2-26.41555.16 0.67
CA028561776.130.66-32.0-27.41775.98-0.15
CC3048592092.750.95 -8.9-35.22090.42-2.33
CA037502518.231.30-33.9-29.52516.92-1.31

We also note that the revision of the NOAA calibration results associated with the transition from the WMO CO2 X2007 to WMO CO2 X2019 scale includes a reassesment of all historical standards used for calibration services outside of the core WMO scale range. The revised values on what we now refer to as the Extended WMO CO2 X2019 scale where retroactively applied to all calibrations performed by NOAA. Changes introduced by this revision are larger above 1000 ppm and we caution that the linear X2019 to X2007 relationship does not apply outside of the core range of the WMO scale.

References
  • Tans, P. P., A. M. Crotwell, and K. W. Thoning (2017), Abundances of isotopologues and calibration of CO2 greenhouse gas measurements. Atmospheric Measurement Techniques, 10(7), 2669–2685, https://doi.org/10.5194/amt-10-2669-2017.