In which the EPA explains to Judith Curry why Tom Segalstad is wrong, wrong, wrong.
Several commenters state that CO2 has a short lifetime in the atmosphere (0711.1, 0714.1): for example, a commenter (1616) claims that the lifetime of CO2 can be at most 20 years based on the 12% annual exchange of CO2 with the surface ocean and 10% exchange between the surface and deep ocean as shown in the National Aeronautics and Space Administration (NASA) carbon cycle diagram, and two commenters (3440.1, 3722) state that the overwhelming majority of scientific papers support a residence time of seven years in contrast to the TSD and IPCC. Several commenters (e.g. 3722) cite Professor Segalstad who has stated, based on his work on CO2 residence times (Segalstad 1997), that the assumption of a 50- to 200-year lifetime by IPCC results in a “missing sink” of 3 Gt of carbon a year, which is evidence that IPCC is mistaken.
Another commenter submitted Essenhigh (2009), which developed a box model and also found that the lifetime of CO2 was on the order of a few years.
EPA reviewed the information presented, as well as the work by Segalstad, and finds that it does not address the lifetime of a change in atmospheric concentration of CO2, but rather the lifetime in the atmosphere of an individual molecule of CO2. These are two different concepts. As stated in the First IPCC Scientific Assessment, “The turnover time of CO2 in the atmosphere, measured as the ratio of the content to the fluxes through it, is about 4 years. This means that on average it takes only a few years before a CO2 molecule in the atmosphere is taken up by plants or dissolved in the ocean. This short time scale must not be confused with the time it takes for the atmospheric CO2 level to adjust to a new equilibrium if sources or sinks change.
This adjustment time ... is of the order of 50–200 years, determined mainly by the slow exchange of carbon between surface waters and the deep ocean” (Watson et al., 1990). The magnitudes of these large balanced sources and sinks are addressed in response 2-2, and are similar to those represented in the NASA carbon cycle diagram. Newer research has only extended and confirmed this statement from the first IPCC assessment report (Denman et al., 2007). A recent approximation for this perturbation lifetime is sometimes represented as the sum of decay functions with timescales of 1.9 years for a quarter of the CO2 emissions, 18.5 years for a third of the CO2, 173 years for a fifth of the CO2, and a constant term representing a nearly permanent increase for the remaining fifth (Forster et al., 2007).
The “missing sink” that was referred to by a commenter is also addressed in response 2-2, and is now called the “residual land sink.” The magnitude of this sink is about 2.6 Gt of carbon per year, with significant uncertainty. Denman et al. (2007) included a hypothesis that a portion of this sink is due to the increased growth of undisturbed tropical forest due to CO2 fertilization, but the carbon accumulation of natural systems is hard to quantify directly. The uncertainty in determining the size and nature of this residual sink does not contradict the assessment literature conclusions about the perturbation lifetime of CO2 concentration changes in the atmosphere, but is reflected in the carbon cycle uncertainty for future projections of CO2 (see responses regarding carbon cycle uncertainty in Volume 4 on future projections).
The box model in Essenhigh (2009) is clearly flawed: the results from this model as reported in the paper include a lifetime for CO2 containing the 14C isotope that is a factor of 3 different from the lifetime of CO2 containing the 12C isotope. This difference in lifetimes is not scientifically compatible with the immense difficulty involved in isotope separation. The model assumes that each “control volume” (each volume represents either the ecosystem, the surface ocean, or the deep ocean) is perfectly mixed, which is contrary to the observations of oceanic CO2 which show that storage of carbon in the ocean is only at 15% of the equilibrium value, and that the mixing time between the surface ocean and intermediate and deep oceans is on the order of years to centuries (Field and Raupach, 2004). Additionally, the paper uses only historical fossil fuel emissions of CO2, without including land use change CO2, and contains the same confusion about “residence lifetime” and “adjustment lifetime” that has been addressed above.
A common analogy used for CO2 concentrations is water in a bathtub. If the drain and the spigot are both large and perfectly balanced, then the time than any individual water molecule spends in the bathtub is short. But if a cup of water is added to the bathtub, the change in volume in the bathtub will persist even when all the water molecules originally from that cup have flowed out the drain. This is not a perfect analogy: in the case of CO2, there are several linked bathtubs, and the increased pressure of water in one bathtub from an extra cup will actually lead to a small increase in flow through the drain, so eventually the cup of water will be spread throughout the bathtubs leading to a small increase in each, but the point remains that the "residence time" of a molecule of water will be very different from the "adjustment time" of the bathtub as a whole.
This analogy does not hold for other GHGs: methane, HFCs, and N2O are actually destroyed chemically in the atmosphere, unlike CO2 where the carbon is not destroyed but merely shifted from one reservoir to another, and therefore the residence lifetime of these gases is fairly close to the adjustment lifetime of their concentrations in the atmosphere.
Similarly, any given molecule of CO2 is only expected to stay in the atmosphere for a few years before it moves into the oceans or ecosystem, but the change in atmospheric concentration due to combustion of fossil fuels can persist for much longer. Indeed, because the oceans and ecosystems are finite, some small fraction of CO2 emissions will have a perturbation lifetime in the atmosphere of thousands of years (Karl et al., 2009).