This (rather dodgy) graph shows a number of estimates of the residence time of atmospheric CO2:

enter image description here

My question concerns why there is so much variation (Machta, 1974 about 3 years, Seuss and Druffe, 1983 about 12 years) in the estimates based on the decay of bomb 14C, given that it seems reasonably easy to estimate from the decay after the nuclear test ban treaty in 1963:

enter image description here

Is there some complicating factor?

I've been doing some thinking and research about this and some issues that might be relevant are:

  • small 14C generated from nuclear power reactors ,
  • the carbon cycle is not at equilibrium, but a naive analysis might
    assume that.
  • 1
    $\begingroup$ The different locations for the measurements might not be comparable. Sources of large mounts of CO2 established during the time series were recorded might alter the delta-14C values close by. The noise in too short time series might also lead to different results for the residence time. Maybe "residence time" and "half life" were mixed together in the figure. Also estimating "constants" within a factor of four might not be a so bad result for a complex system like the earth ;-) . $\endgroup$ Oct 19, 2016 at 14:57
  • $\begingroup$ Cheers daniel.neumann, in the first figure they are confusing residence time (the average time a molecule of CO2 remains in the atmosphere) and the adjustment time (the timescale on which atmospheric CO2 responds to changes in the sources and sinks, e.g. fossil fuel emissions). Sadly this is a very common confusion on climate blogs! My question was only about the green bars relating to bomb 14C, rather than the others, but this was a nice indication of the spread if you ignore the red bar! $\endgroup$ Oct 19, 2016 at 15:10

1 Answer 1


Judging by the general quality of the first graph, I don't think that it is the most reliable source. However as I don't feel like reading trough all papers, I will accept the numbers as they are stated. However once you include uncertainties of the observed residence time, I would assume most of the studies are not significantly different from each other.

The variability in the calculated residence time can be due various reasons; the measurement location, the number of observations used and the length of the observational record.

The measurement location is probably the most important source of variation as results may vary a lot depending on where you measure. Multiple observation locations measuring what is called background air masses would be needed to get a good indication of the global average residence time of $CO_2$ in the atmosphere.

I would say you are spot on with the complicating factors. Local influences are important as $CO_2$ emissions have different radiocarbon signatures depending on their origin. $CO_2$ from fossil fuel burning contains no radiocarbon while the nuclear industry emits low amounts of pure $^{14}C$ (not necessarily in the form of $^{14}CO_2$ though). The natural sources and sinks of $CO_2$ (biosphere and ocean) are also not in equilibrium with the atmosphere since the atomic bomb tests in 1950s and 1960s. Immediately after the bomb tests, the large amount of $^{14}C$ produced in the atmosphere meant that it was enhanced in $^{14}CO_2$ compared to the biosphere and the ocean. Since then, $^{14}CO_2$ has decreased in the atmosphere, as it was taken up by the biosphere ant the oceans. This lead to an such a large increase in $^{14}C$ in the biosphere and the surface ocean, that they are now enhanced compared to the atmosphere.

This graph including the oceans should explain what I mean.

Radiocarbon signature in ocean and atmosphere.

  • $\begingroup$ I checked a couple of the green bars on the first figure and they seemed O.K., I suspect it is only the red bar that is wrong, my coauthor of the linked article made a corrected version static.skepticalscience.com/pics/… . I'm hoping to find the time to reproduce some of the calculations and need to find out more of what I don't know I don't know ;o) $\endgroup$ Oct 21, 2016 at 16:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.