E&ES 359 Climate Change J. C. Varekamp

In the 1980-1990 period we released on the order of 5.4 GtC from fossil fuel burning and 1.0 GtC from deforestation, for a total of 6.4 GtC/yr. From the observed data of CO₂ in the atmosphere, we can deduce that about 3.3 GtC stayed in the atmosphere (about 52 %), and models predict that about 2.1 GtC disappeared into the oceans (about 33 %). That leaves 1.0 GtC unaccounted for, and the only place where such a massive amount of carbon could have gone is in the biosphere (greening and browning of the earth, about 15 %).

It is hard to document the quantitative increase in terrestrial biosphere mass and the ocean uptake term is model dependent as well. How do we figure this all out?

1. The Keeling (p) curve from 1957 till now shows the steady increase in atmospheric CO₂ and the seasonal oscillation caused by the waxing and waning of the northern hemispheric terrestrial biomass (remember our discussion here about distribution of land and biosphere). There are slight indications that the seasonal amplitude has increased over time, suggesting an ‘earth greening’ process.
2. In 1988 Keeling (f) discovered a method to measure small changes in atmospheric O₂ and his atmospheric O₂ curve is out of phase with the Keeling (p) curve, which makes sense. We can predict the decrease in atmospheric O₂ from the overall anthropogenic C flux, and the observed decrease is smaller than the predicted one: evidence for greening!! The observational record is too short to make better predictions for a partition of C sinks between ocean and the terrestrial biosphere. Some papers have appeared arguing that most of the CO₂ went into the biosphere and not into the ocean, but these are highly debated.

The ocean uptake term can be approached from several angles. From thermodynamics we can calculate the C storage capacity of ocean water for the various carbonate compounds. This is a theoretical upper limit and does not tell us anything about the rate of uptake. The carbonate reactions are

CO₂ + H₂O è H₂CO_3èH⁺ + HCO₃^- è H⁺ + CO₃⁼

These are known as the K_CO2, K₁ and K₂ reactions, and the reaction constants are known for a given temperature. Total dissolved carbonate, SCO_2, is H₂CO₃ + HCO₃^- + CO₃⁼, and usually HCO₃^- is the largest of these three. We can calculate for a given pCO₂ all the concentrations using the reaction constants and their equations, and the charge balance equation. The pH is also determined by the pCO₂: when SCO₂ increases, the pH goes down. The oceanic CO₂ uptake capacity is usually expressed with the Revelle buffer factor, which is defined as the relative increase in SCO₂ compared to the relative increase in atmospheric pCO₂ or (DpCO₂/pCO₂) / (DSCO₂ / SCO₂ ) = ~11.

The changes in oceanic SCO₂ are difficult to determine from carbonate measurements in sea water because the increases are small. Current SCO₂ is about 140 ppm HCO₃^- and varies by a few ppm from spot to spot dependent on biological activity. The parameter of interest is the CO₂ invasion rate, I, or how fast does CO₂ dissolve in ocean water (reaction KCO₂, K₁ and K₂ above) if there is a certain degree of undersaturation in the water relative to the pCO₂ in the atmosphere. The uptake of CO₂ can then be modeled as a function of I for some thickness of a mixed surface layer, and then the diffusion (D) of that CO₂ through water to deeper layers. This would be a 1-dimensional, static ocean with no currents. The greatest resistance (slowest process) is the diffusion-into-deep-water term in these equations. Mixing in the surface waters is fast relative to the dissolution speed of CO₂ so the top layer remains homogeneous. The factor I is obtained from the step wise modeling of absorption of CFC’s (a very different gas than CO₂), Rn (a naturally occurring radioactive gas, different from CO₂), Tritium (from nuclear experiments, but also a very different gas from CO₂), from ¹⁴C generated during the bomb tests (1951-1964), and from natural ¹⁴C and its dilution by fossil fuel carbon in the atmosphere. The figure handouts show some of the details of these approaches: the bomb ¹⁴C pulse is the simplest: we can measure how far the bomb ¹⁴C has penetrated into the ocean (about 380 m), which gives the diffusion coefficient D, and we can measure the mean amount of ¹⁴C in the mixed layer, knowing for every year the ¹⁴C/C ratio in the atmosphere, which gives I (about 0.064 mole m^-2 yr^-1 Dmatm-pCO₂^-1). The Dmatm-pCO₂ is the difference between atmospheric pCO₂ and the virtual pCO₂ in ocean water ([H₂CO₃]/KCO₂). We can also model the abundances of natural ¹⁴C in the ocean, which becomes diluted with fossil fuel "dead" carbon. All of these approaches give similar I and D values. In our class experiments we determine directly the I value for a given temperature for different Dmatm-pCO₂ values.

In addition to the static ocean model, we can consider the thermohaline convection component which drags surface waters to the ocean bottom. The surface waters may have absorbed some excess CO₂ before they are dragged down, which is another sink for excess CO_{2 atm}. The "conveyor belt of water" takes out 20 Sverdrups, which is 20 10¹⁶ m³/sec and we can calculate that as a potential annual CO₂ sink of about 0.5 GtC (depends on the degree of equilibration with the atmosphere on the surface). Using all these numbers and models leads to the annual values of about 2.0 GtC into the ocean.

In addition to the CO₂ dissolution fluxes, there is the biological carbon pump, which transports CO₂ down into the ocean after photosynthesis. This process is strongly dependent on the availability of the nutrients NO₃^- and P, and is not directly influenced by the pCO₂ in the atmosphere. The combustion of fossil fuels also delivers NO_x to the atmosphere which then forms NO₃^-, which may lead to additional productivity and an enhanced biological carbon pump in the oceans. The strength of that mechanism to remove excess anthropogenic carbon is a matter of debate.