Revision of the international calibration scale for CO2-in-air: WMO-X2005

1 Climate Monitoring and Diagnostics Laboratory, NOAA 2 Cooperative Institute for Research in Environmental Sciences, University of Colorado at Boulder

The fifteen Primary Standards that define the current MO Mole Fraction Scale for CO2-in-Air have been in existence since early 1991. They became the basis of the WMO Scale during the WMO CO2 Experts meeting in 1995. The first calibrations of these cylinders were carried out by the Scripps Institution of Oceanography, the WMO Central Calibration Laboratory (CCL) until 1995. CMDL has until now carried out six manometric calibration episodes of the Primaries starting in 1996. During each such calibration episode the CO2 mole fraction in dry air of each Primary was determined three or more times. The absolute uncertainty of the CMDL calibrations is 0.068 (one sigma). The repeatability (standard deviation of repeated individual measurements of each cylinder) of the calibrations has improved over time from 0.12 micromol/mol to 0.04 micromol/mol. The average of all cylinders during each episode varies between episodes with a standard deviation of 0.04 micromol/mol, which suggests small systematic variations of procedure between episodes, but not a trend. The “null” hypothesis, namely that there has been no drift of the mole fraction, cannot be rejected for any of the Primaries with the statistics we have. Therefore, we assume that there has been no drift until now, and the assigned value for each primary is the average obtained for that cylinder during all CMDL calibration episodes.

The above does not imply that the defined WMO Scale has not drifted. The early assigned values of the Primaries were based on infrared calibrations by Scripps against the WMO Scale as maintained by them. From 1996 through the fall of 2001 they were determined both by Scripps and CMDL calibrations. After the latter date the assigned values were determined exclusively by CMDL manometric calibrations. The average of all assigned values to the Primaries increased by 0.16 micromol/mol from 1996 to 2005. In the ambient range of 345-415 micromol/mol, the average increased by 0.14 micromol/mol. From late 2001 until 2005 the average of all assigned values of the Primaries has decreased by 0.01 micromol/mol for both the full range and the restricted (ambient) range.

In June 2002 CMDL received revised numbers from Scripps for the four calibrations by them of our fifteen WMO Primaries that they had carried out from 1991 to 1999. Their revised scale was called X99A. The average of all Scripps calibrations of all of our Primaries on the X99A scale was 0.06 micromol/mol higher than the average of all CMDL calibrations, while their average in the ambient range (345-415) was 0.02 micromol/mol higher than ours.

In addition, in 2005 we carried out, at the request of Y. Tohjima of the National Institute of Environmental Studies (NIES), calibrations of five new gravimetric standards (approx. 350 and 390 micromol/mol) made by one-step dilution by Japan Fine Products Company. The average of the NIES values was lower than CMDL’s by 0.004 micromol/mol, and the standard deviation of the individual comparisons was 0.02 micromol/mol. Based on the assumption that the fifteen WMO Primaries have been stable, we have reassigned a constant value (the average of the measured values during the six CMDL calibration episodes) to each of them going back to 1994. We call this revised scale WMO-X2005. The assignments of the CMDL secondary standards, that derive their values directly from the Primaries by repeated infrared comparative measurements, have also been recalculated going back to 1994. All other calibrations of CO2 reference gas cylinders were then recalculated based on the revised values of the secondaries. With few exceptions the other calibrations have always been done relative to the secondary standards. All laboratories and individuals for whom we have calibrated CO2 standards in the past will be able to read the previous and the newly revised values of their reference gases from the CMDL website upon entering the cylinder number of the reference gases, or they may contact Mr. Duane Kitzis at CMDL.

Guess they didn't want to do a press release about this one.

The calibrations for the CO2 measuring equipment were wrong.

The numbers we've been fed about the rise in CO2 for the last 50 years are questionable because of faulty calibration of equipment.

"The “null” hypothesis, namely that there has been no drift of the mole fraction, cannot be rejected for any of the Primaries with the statistics we have. Therefore, we assume that there has been no drift until now, and the assigned value for each primary is the average obtained for that cylinder during all CMDL calibration episodes."

"The above does not imply that the defined WMO Scale has not drifted."

"The average of all Scripps calibrations of all of our Primaries on the X99A scale was 0.06 micromol/mol higher than the average of all CMDL calibrations, while their average in the ambient range (345-415) was 0.02 micromol/mol higher than ours."

"Based on the assumption that the fifteen WMO Primaries have been stable, we have reassigned a constant value (the average of the measured values during the six CMDL calibration episodes) to each of them going back to 1994. We call this revised scale WMO-X2005."

"All laboratories and individuals for whom we have calibrated CO2 standards in the past will be able to read the previous and the newly revised values of their reference gases from the CMDL website..."

Keep in mind that the increases measured and reported so far at Mauna Loa were only 0.00013% average annual of the total air sample, or 1.3 parts per million.

More on calibration. 1996

Department of Commerce, Mountain Administrative Support Center, Procurement Division, MC3, 325 Broadway, Boulder, CO 80303-3328

"The calibration shall consist of a determination by the Scripps manometer of the mole fraction of CO2 in each NOAA reference gas cylinder..."

"...it is the intent of the Government to award a sole source contract under the authority 41 U.S.C. 253 (c)(1) to the University of California at San Diego. Scripps Institute of Oceanography, 2314 Ritter Hall, LaJolla, CA 92093, as this source is considered to be the only responsible source and no other type of supplies or services will satisfy agency requirements."

Not only were Scripps AKA Mauna Lua the only collector of data, they were the only calibrator of the equipment.

If Scripps was wrong, then there was no independent check of the calibration, because they did it themselves, and all the other later CO2 measurers presumably had their equipment calibrated from Scripps.

The average increase Mauna Lua measured was 1.3 parts per million by volume. That is 0.00013%.

Anyone confident that their measurements and calibration errors weren't responsible for that?

http://www.fbodaily.com/cbd/archive/1996/10(October)/31-Oct-1996/Bsol001.htm

I'm just trying to put the magnitude of the calibration errors in prespective with observed changes in CO2 levels. You cite Mauna Lua as the only collector of data. So what you are basically saying is that there is only 1 source for benchmark atmospheric CO2 levels in the world? And that, based on the figures your posted article cites, there is an apparent downward drift in the calibration figures over time and that this downward drift is of similar size to the observed increase in atmospheric CO2 levels?

Are the Scripps calibration standards supposed to be world wide gold calibration standards used by everybody doing climate change work? Or are they for U.S. researchers? In other words does Europe have a gold standard calibration set or are we relying only on 1 measurement center on which we base all our measurements? If that is the case, then all the previous CO2 measurements need to be redone. How many papers using the faulty calibrations will be revised and republished? How many key fiundings, long accepted by the climate change establishment will be significantly questioned. On that, I won't hold my breath what knowning the politics of climate research today.

The magnitude of the recalibration is relatively small 0.16 ppmvAs.

It is the fact that

"....there has been no drift of the mole fraction, cannot be rejected for any of the Primaries with the statistics we have.Therefore, we assume that,there has been no drift until now...... The above does not imply that the defined WMOScale has not drifted".

They are talking about an unknown amount of "drift" possibly (probably?) over and above the amount they have recalibrated for.

Their own page (extracts shown as follows) on preparation and stability is littered with examples of drift potential, and those are only the ones they know about now. What others remain as yet undetected?

The manufacturing target of the CO2 test canisters is "within 5 umol/mol for near ambient concentrations." But this appears to be greater than the annual measured change of 1.3 ppmv.

"The zero air from both trapping agents was tested and found to be less than 5 umol/mol for CO2."

"Aquasorb is phosphorus pentoxide (P2O5). However, CO2 was found to increase by 2-5 umol/mol with this new drying material. This problem and the safety issue of cycling phosphoric acid in a high pressure system prompted the decision to discontinue use."

"The steel cylinders used in the past exhibited too much drift in CO2 mixing ratios, with half of these drifts on the order of 0.05 umol/mol CO2 per year [Komhyr, 1985] could be at different rates or of a different sign in each cylinder."

"There are some cases (8%) with evidence of drift greater than 0.045 umol/mol per year. For this reason, it is recommend that standards be recalibrated during long use and at the end of their useful volume. The drifts, as shown in the histogram, cannot be predicted. Drift in most of our standards is often undetectable for histories of less than 2.5 years. In some documented cases the drift was attributed to H2O mixing ratios greater than 5 umoles mole-1, stainless steel valves, or valves with packing materials, which preferentially absorb trace gases. There may be an increasing trend in the CO2 with decreasing cylinder pressure below 20 atmospheres. This is different in each cylinder or may not be present at all."

"There have been some cases of major drift (0.05 per mil 13C per week) of the CO2 stable isotopes, attributed to the Teflon paste used as the cylinder valve sealant and also to unknown regulator contamination. There may be some evidence for long-term drift of CO in aluminum cylinders as exhibited in smaller high-pressure cylinders [Paul Novelli, CMDL, personal communication, 1999]. This drift, however, is on the order of the measurement precision and has not yet been well quantified."

"Drifts have persisted for months due too evacuating the cylinder down to vacuum before filling. Drifts due to spiking, discussed below, have been eliminated by filling the cylinder immediately after spiking and filling the cylinder laid down. There is no evidence to show drifting due to the filling process described below. As a result of these experiences and to remain conservative about long-term unknown effects, the following steps have been adopted by CCGG..."

"The zero air is created with specific chemical traps for each trace gas. Carbon dioxide can be removed with sodium hydroxide, Ascarite (Thomas). Carbon monoxide is removed with Schutze reagent. The zero air from both trapping agents was tested and found to be less than 5 umol/mol for CO2.."

"Occasionally the targeted concentration is missed by an amount that requires adjustment. If the concentration of the standard is closer to ambient than the target concentration, it is easiest to blow this off and start over. Experimenting with introducing small volumes of high pressure (high or low concentrations) into the air stream have proven very time consuming due to the time necessary for the cylinder to become evenly mixed before obtaining a usable measurement from the standard."

"There is evidence that as the volume of pressurized air is used up, the trace gas can come off the walls in disproportionate amounts. This has sometimes been seen in cylinders with CO2 when a standard is used at pressures below 20 atm. Thus CO2 mole fraction may increase with pressure loss. This effect is variable and may not be measurable in all tanks."

NOAA Technical Memorandum ERL-14
CMDL/CARBON CYCLE GREENHOUSE GASES GROUP STANDARDS
PREPARATION AND STABILITY
Duane Kitzis
Conglong Zhao
http://www.cmdl.noaa.gov/ccgg/refgases/stdgases.html

Remember they are claiming to show accurately changes of 1.5ppmv over a year.

If they have made a mistake, or there is drift in the primary calibration, then all the figures produce by this method may be wrong.

Science should be about independent testing and replication, but if all the testers are using the same equipment, using the same calibration of the primary, then there is in fact no independent verification of the method or the calibration.

See also http://www.cmdl.noaa.gov/ccgg/refgases/reg.guide.html Guidelines For Standard Gas Cylinder and Pressure Regulator Use, for how complicated all this is to work in practice.

As far as I know in the early years SCRIPPS and now NOAA/ESRL's Global Monitoring Division (formerly CMDL) of the National Oceanic and Atmospheric Administration are the only source for the testing calibration gases and the equipment and methodology. I don't know whether the Europeans or others have their own methods, but I've never seen any data from them.

Their page showing possible additonal source of error is:

http://www.cmdl.noaa.gov/ccgg/refgases/airstandard.html

And their guidelines for use are here:

http://www.cmdl.noaa.gov/ccgg/refgases/reg.guide.html

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.