Atmospheric O/N Ratio

The atmospheric reservoirs of O and CO are linked by processes that involve the formation and destruction of organic matter such as photosynthesis, respiration, and combustion. On times scales shorter than many thousands of years, these organic oxidation-reduction reactions are the main source of variability in atmospheric O abundance. These reactions also produce and destroy CO, but the chemistry of atmospheric CO is further complicated by reactions with seawater. In seawater, CO dissolves to form carbonic acid which can react to form basic compounds like carbonate and bicarbonate ions. These acid-base reactions have no effect on oxygen abundance so that atmospheric oxygen variations essentially reveal how atmospheric carbon dioxide would behave if the acid-base reactions did not occur.

The difference between the geochemistry of atmospheric O and CO can be quantified in terms of the relative fluxes of O and CO expected from certain types of processes, as summarized in Table 2. One important difference between CO and O is that the uptake of fossil-fuel CO by the ocean essentially proceeds through reaction of dissolved CO with carbonate ions and therefore involves no O. Another difference is that marine photosynthesis and respiration can produce much larger changes in atmospheric O than CO, especially on short time scales. Here the difference depends mainly on the fact that CO exchange between the atmosphere and oceans proceeds much more slowly than O exchange. CO is exchanged slowly because most of carbon in the oceans is in the form of carbonate and bicarbonate ions which are not exchanged across the air--sea interface.

Two techniques are now available for measuring
[4]changes in atmospheric oxygen, one involving interferometry [ Keeling, 1988; Keeling and Shertz, 1992], the other mass spectrometry [ Bender et al., 1993]. Both methods determine changes in atmospheric oxygen
[4]through changes in the O/N ratio of air. Changes in the O/N ratio are mainly caused by changes in O because N is constant to a very high level. Like isotopic ratios, the O/N ratio is expressed as deviations from a reference

The resulting deviations are multiplied by and the result is expressed in a new unit called a ``per meg.'' In these units 1/0.2095 = 4.8 per meg is equivalent to 1 part-per-million by volume (ppmV) because O comprises 20.95% of air by volume [ Machta and Hughes, 1970].

Measurements on air samples collected at three sea-level sites using the interferometric technique were reported by Keeling and Shertz [1992], and are shown here in Figure 4. Significant seasonal variations in (O/N) are evident at all three sites. An interannual decrease in O/N is clearly evident in the La Jolla data. Concurrent CO data are also shown.

One process leading to seasonal variations in O/N is the seasonal uptake and release of O due to photosynthesis and respiration of terrestrial ecosystems. These exchanges of O are closely tied to exchanges in CO with an exchange ratio of approximately -1.05:1 (O:CO). The seasonal variations in CO in the northern hemisphere are almost entirely caused by these terrestrial exchanges, and they can be used to correct for the effects of terrestrial exchange on the O/N variations [ Keeling and Shertz, 1992]. The residual variations in O/N must be oceanic in origin. The oceanic component is especially pronounced in the southern hemisphere where the seasonal O/N variations are accompanied by only very small variations in CO.

Oxygen is released to the atmosphere by the oceans at middle and high latitudes in the spring and summer when the net rate of photosynthesis in surface waters exceeds the rate of respiration. Oxygen is removed from the atmosphere in the fall and winter when marine photosynthesis rates are lower and when deeper water, undersaturated in oxygen, mixes upwards to the surface. These seasonal air--sea O fluxes are linked to the rate at which organic material is produced and exported from the euphotic zone [ Jenkins and Goldman, 1985; Keeling et al., 1993] and they are linked to changes in dissolved inorganic carbon (DIC) in the water. Seasonal heating and cooling of the upper ocean also contributes to seasonal variations in atmospheric O/N because of the solubility temperature dependence of O and N [ Keeling and Shertz, 1992].

Measurements of seasonal variations in O/N will be useful constraining estimates of the annual net photosynthetic production of organic carbon in the euphotic zone. To succeed this application also requires taking account of transport within the atmosphere and transport of O between the euphotic zone and deeper waters. Atmospheric oxygen data may be especially helpful in determining productivities over large regions because the air mixes so rapidly.

Measurements of O/N ratios will also be useful for determining the mechanisms by which excess carbon dioxide produced from fossil-fuel burning is being removed from the atmosphere. Over the long-term, we can represent the global budget for atmospheric CO according to

where CO is the annual averaged change in atmospheric CO, F is the source of CO from burning fossil fuels, C (virtually negligible) is the CO source from cement manufacturing, O is the oceanic CO sink, and B is the net source of CO from terrestrial ecosystems (B can be positive or negative), all in units of moles yr. Likewise, we can represent the budget for atmospheric oxygen according to

where O is the change in atmospheric oxygen, H is the O sink owing to the oxidation of elements other than carbon (predominately hydrogen) in fossil fuels, and represents the O:C exchange ratio for terrestrial biomatter.

Adding Eqs. (3) and (4), and solving for O yields

The last term on the right-hand side of Eq. (5) can be evaluated by solving Eq. (4) for B, although this term is virtually negligible since . Solving Eq. (4) for B yields

These equations show how observations of the change in atmospheric oxygen combined with estimates of fossil-fuel CO production and O consumption can be used to directly calculate the net exchange of CO with the oceans and with the land biosphere.

Using preliminary estimates of the O trend based on the data shown in Figure 4, Keeling and Shertz [1992] derive an oceanic uptake of gT C/yr (1 gT = 10g) and a net terrestrial carbon sink of gT C/yr for the 1989--1991 period. This estimate is clearly preliminary, and the uncertainties are too large to make these results very useful in constraining CO sinks. The primary source of uncertainty comes from uncertainty in the O trend, and this uncertainty should decrease as longer records are obtained.

Bender et al. [1994b] have extended our knowledge of variations in atmospheric O/N ratio back over the past decade by measurements on air samples extracted from glacial firn at Vostok Station, Antarctica. The detected O/N variations imply that the terrestrial biosphere was neither a large source nor sink of CO over this longer period, agreeing with Keeling and Shertz [1992], although the uncertainties in this preliminary work are again quite large. Attempts to extend the records even further into the past from air extracted from bubbles in the glacial ice have so far been frustrated by processes which fractionate O relative to N in the ice bubbles or during the extraction process [ Craig et al., 1988; Sowers et al., 1989; Bender et al., 1995].

Although uptake of anthropogenic CO by the oceans has no effect on atmospheric O, it is possible that natural variability in the oceans could lead to net O exchange with the oceans on interannual time scales. This possibility, which was neglected in Eq. (4), would complicate the use of O/N data for discriminating between terrestrial and oceanic sinks for CO. Such air--sea exchanges are especially likely on the 3 to 6 year time scale of the El Nino phenomenon [ Keeling and Severinghaus, 1994] which means that the O/N records will probably need to span several El Nino events before the data can be used to place firm constraints on the sources and sinks of anthropogenic CO.