The Jena Reference Air Standard – JRAS, produced at the Max Planck Institute for Biogeochemistry in Jena, Germany,
is a stable isotope standard consisting of CO₂ generated from a calcite and mixed into CO₂-free air. It is closely
linked to the VPDB-scale and is well suited to serve as a primary scale anchor for CO₂–in-air measurements. The
World Meteorological Organization – Global Atmosphere Watch (WMO-GAW) has asked MPI-BGC to act as a Central
Calibration Laboratory (CCL) in an effort to unify the stable isotope scale anchors across the worldwide community of
laboratories measuring stable isotopes of CO₂ in air. The effort is supported and funded by the IMECC project
(Infrastructure for Measurements of the European Carbon Cycle). IMECC is an Integrated Infrastructure Initiative
(I³) under the Sixth Framework Program of the European Commission.

JRAS is currently distributed to 11 laboratories all over the world, representing not only different countries and
regions but also different analytical strategies. For example; the laboratory at the Max Planck Institute for
Biogeochemistry in Jena, measures CO₂ in atmospheric samples by extracting 600 ml of air, while the laboratory in
the Department of Climate and Environmental Physics at the University of Bern measures air trapped in ice by
collecting as little as 1 ml of air per sample. These differences in analytical procedures are reflected in the
precision results obtained by the different laboratories, as shown by the data presented below.

The JRAS project does not aim to compare the capabilities of the laboratories; But rather to provide a rigorous
test of the suitability of JRAS as the unifying scale anchor. This is done by identifying the offsets between
local isotope scales and JRAS.

The offsets are defined as the difference between the values obtained for the JRAS gases on the local scales and
their assigned values. These offsets will be monitored in time to produce a measure of scale consistencies. For
this purpose the standard deviation of the long term average of the offset will be used. The scale consistency will
also serve as a measure of the reliability of JRAS as the scale anchor. If JRAS is found to be a reliable scale
anchor, the scale differences between laboratories can then be eliminated by adopting JRAS as a global scale anchor
for CO₂ in air measurements.

A JRAS set contains two 5-L glass flasks each equipped with a single PCTFE valve. Two different calcites are used
to produce the CO₂; MAR-J1 (MARble - Jena #1, purchased from a local vendor in Jena) with a δ¹³C_VPDB-air of 1.96 ‰
and a δ¹⁸O_VPDB-air of -2.58 ‰, and OMC-J1 (Otavi Meieberg Calcite-Jena #1), a calcite slab from the Meieberg
section of the Otavi platform in northern Namibia. The OMC-J1 has a δ¹³C_VPDB-air of -4.37 ‰ and a δ¹⁸O_VPDB-air
of -8.93 ‰. These calcites have been chosen based on their isotope composition. MAR-J1 is very close to the δ¹³C
value of NBS19, the scale anchor for the VPDB scale, and is therefore providing a close connection between JRAS and
the VPDB-scale. The second calcite, OMC-J1, is simulating atmospheric CO₂. It is the calcite of suitable homogeneity
and abundance with a δ¹³C value closest to that in atmospheric air that we have been able to find. With a δ¹³C
of -4 ‰ it is a compromise in this regard. However, this shortcoming will be explored during the project by the
additional introduction of a pure southern-hemisphere air sample from Cape Schanck Lighthouse, Australia.

The CO₂ from the calcites is produced by dissolution in concentrated phosphoric acid and subsequent mixing with
CO₂ free air using an automated preparation system (ARAMIS).
One preparation results in a batch of three 5-L flasks of CO₂ mixed into synthetic air. Each flask is first analyzed
for CO₂ and N₂O concentrations and finally for the CO₂-stable isotope composition. The isotopic analysis is repeated
3 times per flask resulting in 9 measurements per batch. The average of these measurements is used to assign the
isotope values of the batch.

PLEASE NOTE:
The offsets in the figures are calculated as the difference between the measured result on the local scale and the JRAS
value assigned to the batch from which the flask originated, ie. (local - JRAS).
For the long term averages the data points are measurements that are separated in time by five days or more.
Measurements made within a shorter time frame has been averaged to yield a single data point. The long term
consitency is one standard diviation of the data points used for the long term average.