(for a time at least) by terrestrial plants, particularly trees. The air–ocean–land
partition of burned carbon has been measured in several ways. Likely the most
accurate are estimates from measures of the amounts of oxygen removed from the
atmosphere by the burning carbon (C + O 2 → CO 2 ), with some additional reduction
from oxidizing the hydrogen (producing water) in petroleum and natural gas. It is also
necessary to estimate the release of both O 2 and N 2 from the ocean caused by
warming on decadal scales. The net change in atmospheric oxygen partial pressure is
only a small decrement to the 20.9% in the atmosphere, too small to measure directly,
particularly given the variation in actual [mol O 2 /volume] driven by varying
atmospheric pressure. However, the ratio O 2 : N 2 can be very precisely measured (i)
by interferometry from the shift in air refractive index as the ratio changes (Keeling et
al. 1998a), or (ii) by mass spectrometry (Bender et al. 1996). The final measurements
are very precise, but both techniques involve refined gas-handling procedures because
oxygen adsorbs to collector surfaces and valve gaskets, reacts significantly with some
metal ducting (e.g. stainless steel) and passes through valves and orifices at different
rates than nitrogen. Moreover, standards and reference air must be produced with
precise proportions of O 2 , N 2 , Ar, and CO 2 then protected from interaction with their
containers. By extended trials, those problems have been solved sufficiently (e.g.
Keeling et al. 1998a) to provide useful data.
(^) The use of ratios measured relative to arbitrary standards produces somewhat
abstruse arithmetic for evaluating temporal changes in the atmosphere and relating
them to the disposition of fossil-fuel carbon dioxide to different global reservoirs –
arithmetic well presented by Manning and Keeling (2006). The first-order variable
used is:
(^) in units termed “meg”, parts per million. The global reduction in the ratio from 1998
to 2002 was ∼21 meg/year (slope of the fitted curve in Fig. 16.7), after removal of
smooth and opposite seasonal variations in the northern and southern hemispheres
(Fig. 16.8). A budgetary argument for changes in atmospheric CO 2 and O 2 allows
estimation of the ocean and land sinks for CO 2 :
(^)
(^) with F = moles fossil C burned, Oc = ocean CO
2 uptake, L = land CO 2 uptake, αF =
O 2 : CF ratio for fossil-fuel burning (∼1.39), αB = O 2 produced by an average mole of
terrestrial organic matter produced by photosynthesis (∼1.1) and Z = the air : sea
exchange of O 2 (small but toward the air due to upper ocean warming). Quantities on