A look back at an article about the imminent “tipping point” of the Atlantic overturning circulation:
At the end of July 2023 we had reported on an essay that thought to have identified the “tipping point” of the Atlantic overturning circulation (AMOC): Supposedly (most likely) it will be already in 2057. The author stirred up a lot of dust, we described it. Our summary at the time:
DD23 is a statistical exercise, very remote from any evidence of physical significance to AMOC itself. The conclusions therein are not supported by the content of the paper.”
In the meantime (mid-September 2023), another unreviewed (preprint) paper appeared and finds very similar things. It uses several sea surface temperature (SST) data series and another AMOC “fingerprint,” a dipole that avoids including warming itself in the result of the AMOC decrease it is looking for.
And lo and behold, it comes to quite different conclusions. Depending on which series one uses – especially in the early years up to about 1950, they are very patchy in direct observation and what is missing is filled in by various “infill methods” which have a strong impact on the statistics as they underlie the paper in question – one comes to completely divergent results. For example, the AMOC “collapse” is delayed by nearly a hundred years if one uses the same “fingerprint”, only a different SST series (Instead of HadISST1 ERSSTv5).
The other “fingerprint”, also based on SST, pushes the “collapse” determined by the methods in the paper to between 2100 to the year 3300 and beyond! At the end, the authors, who interestingly include Niklas Boers of the Potsdam Institute (PIK), who is probably concerned about the scientific reputation of the institution, write:
We emphasize that these uncertainties, originating from underlying modelling or mechanistic assumptions as well as from the employed empirical data, need to be taken into account and propagated thoroughly before attempting to estimate a future tipping time of any potential Earth system tipping element.”
ANY estimate of the timing of a “tipping point” must take the uncertainties into account, which are so substantial that they are all unsound when made.
Bet you won’t read that in the media? They had their hands full, with the active assistance of Prof. Rahmstorf, also from the PIK, to propagate the near end of the AMOC.
For this, attentive readers has this blog.