We really may have to start that feature given that this thing that won’t die just won’t stop dying. Thus on Oct. 14 Reuters tells us “An Atlantic Ocean current system weakened by global warming could recover over time if the climate cools, a study by the Danish Meteorological Institute showed on Wednesday.” Apparently there is no “irreversible tipping point for AMOC, as suggested by some previous studies”, since all the short-term changes are eventually reversible. But we die nonetheless because the study also said “that the Greenland ice sheet that contributes to weakening the AMOC as it pours ⁠cold freshwater into the ocean could not be restored once it had melted – even if the warming trend was reversed.” And the whole thing could “destabilise global rainfall patterns” and speed up sea-level rise.” Could. Could. Could. Unless it doesn’t. But in any case it’s all computer simulations of the AMOC that don’t actually simulate it since we don’t understand how ocean currents work or when or why they do what they do. Change a parameter and get a different result. But none of it’s remotely real-world.

We’re not making it up: the “Abstract” of the actual study, in Science Advances, says in its 2nd sentence that:

“Here, we use a carbon dioxide emission–driven Earth system model with an interactive GrIS to assess AMOC reversibility under a range of idealized carbon dioxide emission pathways that produce near-linear global warming up to 10 kelvins (K), stabilization across 1.5 to 9 K, and subsequent cooling.”

It’s the AMOC of the mind, not the one out there at sea. Moreover, the first sentence of the “Abstract” says:

“The Atlantic Meridional Overturning Circulation (AMOC) is considered vulnerable to irreversible weakening or collapse under global warming and related Greenland ice sheet (GrIS) melt, yet its resilience to either forcing remains poorly constrained.”

It doesn’t say why it’s “considered vulnerable” if it’s “poorly constrained”, which is how scientists say our model doesn’t tell us anything meaningful about what might happen unless it doesn’t. It just tells us how the model behaves while pretending to model the real world:

“We find that although the AMOC reaches threshold-defined collapsed states, these do not represent dynamical tipping in this model”.

Which to paraphrase Nero Wolfe is merely a comment on the model, not on the AMOC. The model of course reflects the assumptions of the programmers about the AMOC because they told it to. Not all models of all things have that problem: some of them actually incorporate things that have been measured empirically in the real world and then turned into equations that actually represent reality in such a wise that you can feed in past conditions and get present ones, or feed in present ones and get short-term predictions you can test that turn out well. But climate models surprisingly often do not do so because they do not attempt to do so, on everything from the behaviour of clouds to these ones on ocean currents. As the “Introduction” that follows the Abstract comments:

“AMOC observations from the RAPID array at 26.5°N since the early 2000s reveal large interannual variability with only a marginally significant negative trend, and the observational record remains too short to robustly separate any forced trend from natural variabilities.”

So the empirical observations are just a bunch of statistical noise with no signal. But then models of what warming will do fed into models of what Greenland ice will do can be fed into models of what the AMOC will do and tell you whatever you want to hear. And it’s all fun and games until someone loses a theory which they never do because none of it is empirical. As the Introduction also says:

“Statistical analyses of proxy records of AMOC strength and model-derived physics-based early-warning indicators have raised the possibility of approaching a tipping point, although these signals remain debated and the proxies contested.”

Which if they taught English in schools would read that our guesses about the past and the physics say it could do something drastic unless it doesn’t, but there’s massive disagreement about those guesses because they’re so insubstantial. The paper goes on and on about what they don’t know but in such impenetrable prose that if the journalists over at Reuters even tried to read it they clearly didn’t understand it.

Despite his general respect for tradition Cato the Elder, according to Cicero, once commented that he did not understand how two soothsayers could look one another in the face without laughing. According to us it’s a mystery when it comes to climate scientists as well when they take all this uncertainty, dress it up in incomprehensibility, and get exactly the conclusion they were looking for when they started out. Thus here the Abstract ends with a lovely non sequitur:

“Our findings underscore the urgency of rapid emission cuts to limit climate overshoot, AMOC weakening, and irreversible ice sheet loss.”

As we told them to.

As for the journalists, the Reuters story does include various passages such as “The 7 C rise used in the modelling used for scientific interest is far beyond the current projections for the warming trend” which might have you tossing the thing out back with the various shoulder-bones and livers used in Rome to predict what might happen unless it didn’t. And also:

“Willem Huiskamp at the Potsdam Institute for Climate ⁠Impact Research, who was not involved in the study, said a more stable AMOC would be good news, but said he doubted the real-world current was as stable as some models suggested.”

That it would be wise to look at the real-world current instead of the guts of a computer model was not however mentioned.