At some point the fact that climate models are reliably unreliable should make discussions of exactly what’s wrong with them moot. As moot, one might say as inflation bursts loose again, as debating the finer defects in Keynesian economics and the computer models it spawned. But it’s still worth pointing out, given the extraordinary faith many people have in them, that they rely on unsound connections between inaccurate facts to reach inaccurate conclusions. Hence Charles Rotter points on Watts Up With That to a new paper that starts, in the very first sentence of the abstract, “The terrestrial carbon cycle is a major source of uncertainty in climate projections.” Which is already heretical in admitting there is uncertainty in climate projections. But wait until you get to the stuff about how what we think we do know about it isn’t just wrong, it’s contradictory and nobody even noticed. Some settled science you’ve got there.
It has been pointed out before that there’s something fishy about the whole carbon cycle as customarily presented with great certainty. Including by us at CDN, who have complained that you can find all sorts of graphics online about how much carbon is released and sometimes absorbed by the oceans, plants, the soil and also naughty human fossil fuels, cement (a surprisingly large source) and changing land use patterns, and they tend to agree very precisely about how much humans are adding. But in the background they often disagree massively about what nature is up to which should make you suspicious about any of their numbers. You should be even more suspicious that there is this widespread “consensus” that half of human CO2 is absorbed and half is not. It’s a suspiciously round number, for one thing, and just weird for another.
How would nature know to absorb all the natural CO2 and only half of ours, or any other proportion, when they are indistinguishable to plants? It’s a sort of natural Calvinism, a religious dogma that Gaia is harmonious and we are utter rubbish. It seems far more likely that if atmospheric CO2 is rising, nature is absorbing the same proportion of both natural and man-made CO2, say 95%, and leaving 5% of each hanging around ruining everything. Except then of course you’d have natural fluctuations and bad natural CO2 which would never do because it is generally conceded that nature cranks out about 19 times as much CO2 as we do every year. But in a good way.
Hang on, say these researchers. “We combine carbon-cycle flux estimates and partitioning coefficients to show that historical estimates of global GPP and RS are irreconcilable.” Do you though? GPP is “gross primary productivity” and RS is “soil respiration” (with the “S” subscripted if typography permits). And the point here is that the traditional method of estimating the former from the latter gives figures much higher than people typically use for GSS.
We will digress briefly to note that the paper’s “Introduction” begins “The terrestrial carbon sink removes about a quarter of anthropogenic CO2 emissions but is highly variable in time and space depending on climate.” A quarter? Wasn’t it meant to be half? And variable? Wasn’t it meant to be steady? But we do digress, because the substance of this finding is elsewhere.
It’s that basically the conventional numbers for the carbon cycle are simply wrong. And nobody checked. “To date, however, no attempt has been made to quantify how consistent these independent GPP and RS estimates are at the global scale.” What a thing to overlook in what is now more than two decades of models we are told are indeed ready for prime time. And bear in mind that since human CO2 is only a small share of the annual natural output and absorption, relatively small errors in estimating the natural side leave you with no real idea of our relative contribution. Which is kind of important if you’re trying to tell us to a tenth of a degree what temperature changes various levels of human GHG output will cause in eight decades or so, which alarmists are.
Reading the fine print in the new study, you discover that both RS and GPP are very loose estimates indeed, based in the former case on “upscaled in situ measurements” which probably means you actually measure it somewhere and then assume that changes in soil composition, altitude, temperature and so forth can be guessed away, and in the latter on “satellite-driven models” which probably means they aren’t worth the screens they are displayed on. But it should at least have been possible to cram them into approximate juxtaposition by kluging various assumptions.
Incredibly, nobody bothered. “These independent lines of the analysis demonstrate that GPPlit and Rslit, the historical global flux estimates reported in the published scientific literature, are almost certainly inconsistent with each other.” Oh darn. Everything the models say is wrong, including what proportion of our nasty CO2 nature spits back each year. And so, the authors say, more research is needed. Of the sort they do.
To which we add that more humility is also needed, of the sort alarmist politicians, activists and journalists don’t do.