pith. machine review for the scientific record. sign in

arxiv: 2604.20341 · v1 · submitted 2026-04-22 · ⚛️ physics.geo-ph · stat.AP

Recognition: unknown

Extrapolation from historical data cannot reliably predict the time of a potential AMOC collapse

Alessandro Cotronei, Andreas Morr, Brian Groenke, Christof Sch\"otz, Eirik Myrvoll-Nilsen, Martin Rypdal, Maya Ben-Yami, Niklas Boers, Sebastian Bathiany

Pith reviewed 2026-05-09 23:05 UTC · model grok-4.3

classification ⚛️ physics.geo-ph stat.AP
keywords AMOCcollapsetipping pointuncertaintyextrapolationsea surface temperaturestochastic modelbifurcation
0
0 comments X

The pith

Accounting for uncertainties makes AMOC collapse predictions unreliable and far in the future.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper challenges a recent estimate that the Atlantic Meridional Overturning Circulation might collapse in the mid-21st century by showing that the statistical extrapolation used is highly sensitive to several types of uncertainty. These include the choice of a simple one-dimensional model, statistical uncertainties in fitting it to data, the suitability of sea surface temperature as a proxy for the full circulation, and variations in how the observational data are processed. Synthetic experiments and checks with alternative data sets demonstrate that the collapse time can shift by thousands of years when these factors are considered.

Core claim

Using synthetic experiments and alternative fingerprints, the tipping times from the DD23 extrapolation are shown to be highly sensitive to structural, statistical, proxy, and data uncertainties, extending several millennia into the future when these are propagated.

What carries the argument

Propagation of four uncertainty categories—structural model form, statistical fit, proxy representativeness, and data preprocessing—through the maximum likelihood fit of a stochastic fold-bifurcation model to SST fingerprints.

If this is right

  • The reported confidence interval from 2037 to 2109 underestimates the true uncertainty range.
  • AMOC tipping time cannot be reliably estimated from current historical SST data alone.
  • Structural model assumptions strongly affect the inferred bifurcation timing.
  • Data preprocessing choices and proxy selection can alter collapse estimates by millennia.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Early warning methods based on low-dimensional fits to limited observations may systematically underestimate uncertainty in complex systems.
  • Similar sensitivity analyses should be applied to other proposed climate tipping point predictions.
  • Longer or more direct observations of ocean circulation would be required to tighten the bounds on possible collapse times.

Load-bearing premise

The four categories of uncertainty examined are sufficient to capture the dominant sources of error, and synthetic experiments adequately mimic real AMOC behavior.

What would settle it

Long-term direct observations of AMOC volume transport that show a collapse occurring within or outside the expanded uncertainty range from this analysis.

read the original abstract

Ditlevsen and Ditlevsen [Nature Communications, 2023] (DD23 hereafter) propose a statistical framework to estimate the timing of a potential collapse of the Atlantic Meridional Overturning Circulation (AMOC) based on extrapolating information from observed sea-surface temperature (SST) variability. By fitting a stochastic one-dimensional fold-bifurcation model to an SST-based fingerprint of the AMOC using Maximum Likelihood Estimation (MLE), they conclude that a collapse is most likely to occur in the middle of the 21st century, with a reported 95% confidence interval covering the time span from 2037 to 2109. Given the profound implications of such a claim for both climate and society, it is essential to thoroughly test the robustness of this result, to critically assess the underlying assumptions and uncertainties, and to estimate the extent to which the reported confidence interval reflects the true limits of current knowledge. Here we examine the sensitivity of DD23's results and argue that four types of uncertainty are insufficiently explored in their analysis: (i) structural uncertainty associated with the assumed low-order bifurcation model, (ii) statistical uncertainty in their model fit, (iii) uncertainty in the representativeness of SST-based fingerprints as proxies for the high-dimensional AMOC dynamics, and (iv) uncertainty in the underlying data, arising from non-stationary observational coverage and dataset preprocessing. Using synthetic experiments and a systematic analysis of alternative fingerprints and observational products, we show that the tipping times estimated by DD23 are highly sensitive to the uncertainties listed above, and extend several millennia into the future when these uncertainties are thoroughly propagated.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The manuscript critiques the Ditlevsen and Ditlevsen (2023) extrapolation of AMOC collapse timing from SST fingerprints using a stochastic fold-bifurcation model fitted by MLE. It identifies four under-explored uncertainty sources—structural model form, statistical fit, proxy representativeness of SST fingerprints, and data preprocessing/non-stationarity—and uses synthetic experiments plus alternative fingerprints/observational products to show that propagated uncertainties shift the 95% confidence interval for collapse from 2037–2109 to several millennia into the future.

Significance. If the synthetic experiments and alternative-proxy tests are representative, the result demonstrates that low-order statistical extrapolations for AMOC tipping are too sensitive to modeling choices to support precise near-term predictions. This has direct implications for how such statistical forecasts are communicated in climate science and policy contexts. The systematic multi-product analysis is a strength when the central claim holds.

major comments (3)
  1. [Synthetic experiments section] Synthetic experiments section: the generation of synthetic time series from low-order stochastic fold models is not validated against CMIP6 AMOC indices or paleoclimate reconstructions for metrics such as lag-1 autocorrelation decay rates or decadal spectral power. Without this calibration, it is unclear whether the reported multi-millennial extension of confidence intervals correctly bounds epistemic uncertainty or is an artifact of the simplified dynamics.
  2. [Alternative fingerprints analysis] Alternative fingerprints and observational products analysis: the quantitative shifts in MLE-derived tipping times when switching fingerprints or datasets are presented without accompanying uncertainty estimates on the alternative fits or direct statistical comparison to the original DD23 MLE parameters. This makes it difficult to determine whether the sensitivity is robust or driven by specific preprocessing choices.
  3. [Discussion] Discussion of uncertainty categories: the claim that the four examined categories suffice to capture dominant errors rests on the assumption that interactions between structural and statistical uncertainties are negligible; no sensitivity test explores joint variation of model form and data preprocessing, which could alter the propagated intervals.
minor comments (2)
  1. [Figures] Figure captions for the synthetic experiment results should explicitly state the number of realizations and the exact parameter ranges used for the stochastic fold model to allow reproducibility.
  2. [Results] The manuscript cites DD23 but does not include a side-by-side table comparing the original MLE parameters, likelihood values, and tipping-time estimates to the re-fitted values under alternative fingerprints.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their thoughtful and constructive review of our manuscript. We address each of the major comments point by point below, providing clarifications on the scope of our analysis and indicating revisions made to strengthen the presentation of results and limitations.

read point-by-point responses

  1. Referee: [Synthetic experiments section] Synthetic experiments section: the generation of synthetic time series from low-order stochastic fold models is not validated against CMIP6 AMOC indices or paleoclimate reconstructions for metrics such as lag-1 autocorrelation decay rates or decadal spectral power. Without this calibration, it is unclear whether the reported multi-millennial extension of confidence intervals correctly bounds epistemic uncertainty or is an artifact of the simplified dynamics.

    Authors: We appreciate this point. The synthetic experiments were intentionally constructed from the same low-order stochastic fold-bifurcation model used in DD23 to isolate the effects of statistical and structural uncertainties within that modeling framework. Their purpose is to demonstrate that, even when the data-generating process is exactly the assumed model, modest changes in data length, noise realization, or fitting procedure produce tipping-time distributions that extend over millennia. We did not attempt to calibrate the synthetic series to CMIP6 or paleoclimate metrics because doing so would require treating the low-order model as an adequate representation of real AMOC dynamics—an assumption our paper questions. We have revised the manuscript to (i) state this scope limitation explicitly in the synthetic-experiments section and (ii) add a short discussion of how the reported intervals should be interpreted as conditional on the model class rather than as unconditional epistemic bounds. A full cross-validation against CMIP6 indices lies outside the present scope but would be a valuable follow-up study. revision: partial

  2. Referee: [Alternative fingerprints analysis] Alternative fingerprints and observational products analysis: the quantitative shifts in MLE-derived tipping times when switching fingerprints or datasets are presented without accompanying uncertainty estimates on the alternative fits or direct statistical comparison to the original DD23 MLE parameters. This makes it difficult to determine whether the sensitivity is robust or driven by specific preprocessing choices.

    Authors: We agree that uncertainty estimates and parameter-level comparisons would improve interpretability. In the revised manuscript we now report bootstrap-derived 95 % confidence intervals for the MLE tipping times obtained from each alternative fingerprint and observational product. We have also added a table comparing the fitted bifurcation parameter, noise intensity, and estimated critical threshold between the original DD23 fit and the alternatives, together with two-sample Kolmogorov–Smirnov tests on the underlying residual distributions. These additions allow readers to judge whether the observed shifts in tipping time are statistically distinguishable from sampling variability under the same model class. revision: yes

  3. Referee: [Discussion] Discussion of uncertainty categories: the claim that the four examined categories suffice to capture dominant errors rests on the assumption that interactions between structural and statistical uncertainties are negligible; no sensitivity test explores joint variation of model form and data preprocessing, which could alter the propagated intervals.

    Authors: The referee is correct that interactions between uncertainty sources could in principle produce still wider intervals. While our main analysis treats the four categories separately to isolate their individual contributions, we have added a supplementary sensitivity experiment in which we simultaneously (a) augment the model with a linear trend term (structural change) and (b) apply alternative preprocessing choices (e.g., different detrending windows and observational products). The resulting joint confidence intervals remain multi-millennial, consistent with the separate analyses. We have updated the discussion to note that neglecting interactions is likely to underestimate total uncertainty and therefore reinforces, rather than weakens, the central claim that precise near-term collapse dates cannot be reliably inferred from the present statistical approach. revision: yes

Circularity Check

0 steps flagged

No significant circularity in the derivation chain

full rationale

The paper critiques DD23 by propagating uncertainties through independent synthetic data generation from low-order models, systematic variation of fingerprints and preprocessing, and alternative observational products. No step reduces a claimed result to its own fitted inputs or self-citations by construction; the sensitivity analysis introduces external variation rather than re-deriving DD23 outputs tautologically. The central claim remains independent of the target extrapolation.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claim rests on standard domain assumptions about the adequacy of low-order bifurcation models and SST proxies for AMOC dynamics; no new free parameters or invented entities are introduced.

axioms (2)
  • domain assumption A one-dimensional stochastic fold-bifurcation model is a sufficient representation of AMOC tipping dynamics for the purpose of timing estimation.
    This is the structural assumption critiqued in section (i) of the abstract and tested via synthetic experiments.
  • domain assumption SST-based fingerprints are representative proxies for the high-dimensional AMOC state.
    Invoked when the paper examines uncertainty type (iii) and alternative fingerprints.

pith-pipeline@v0.9.0 · 5636 in / 1371 out tokens · 54903 ms · 2026-05-09T23:05:51.109811+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

49 extracted references · 36 canonical work pages

  1. [1]

    Nature Communications14(1), 4254 (2023) https://doi.org/10.1038/s41467-023-39810-w

    Ditlevsen, P.D., Ditlevsen, S.: Warning of a forthcoming collapse of the Atlantic meridional overturning circulation. Nature Communications14(1), 4254 (2023) https://doi.org/10.1038/s41467-023-39810-w

    work page doi:10.1038/s41467-023-39810-w 2023

  2. [2]

    Applied Mathematical Sciences, vol

    Kuznetsov, Y.A.: Elements of Applied Bifurcation Theory, 4th edn. Applied Mathematical Sciences, vol. 112. Springer, New York (2004). https://doi.org/10. 1007/978-1-4757-3978-7

    2004

  3. [3]

    Frontiers in Climate8, 1761461 (2026) https: //doi.org/10.3389/fclim.2026.1761461

    Cotronei, A., Myrvoll-Nilsen, E., Rypdal, M.: Evaluating model uncertainty in critical threshold estimations from time series data: Application to the Atlantic meridional Overturning Circulation. Frontiers in Climate8, 1761461 (2026) https: //doi.org/10.3389/fclim.2026.1761461

    work page doi:10.3389/fclim.2026.1761461 2026

  4. [4]

    Jour- nal of Physical Oceanography24(9), 1911–1920 (1994) https://doi.org/10.1175/ 1520-0485(1994)024⟨1911:ASBMOS⟩2.0.CO;2

    Cessi, P.: A simple box model of stochastically forced thermohaline flow. Jour- nal of Physical Oceanography24(9), 1911–1920 (1994) https://doi.org/10.1175/ 1520-0485(1994)024⟨1911:ASBMOS⟩2.0.CO;2

    1911

  5. [5]

    Physical Review X14(2) (2024) https://doi.org/ 10.1103/PhysRevX.14.021037

    Morr, A., Boers, N.: Detection of Approaching Critical Transitions in Natural Systems Driven by Red Noise. Physical Review X14(2) (2024) https://doi.org/ 10.1103/PhysRevX.14.021037

    work page doi:10.1103/physrevx.14.021037 2024

  6. [6]

    Stochastic properties of the frequency dynamics in real and synthetic power grids,

    Morr, A., Riechers, K., Gorj˜ ao, L.R., Boers, N.: Anticipating critical transitions 12 in multidimensional systems driven by time- and state-dependent noise. Physical Review Research6(3), 033251 (2024) https://doi.org/10.1103/PhysRevResearch. 6.033251

    work page doi:10.1103/physrevresearch 2024

  7. [7]

    Science Advances10(31), 4841 (2024) https://doi.org/10.1126/sciadv.adl4841

    Ben-Yami, M., Morr, A., Bathiany, S., Boers, N.: Uncertainties too large to predict tipping times of major Earth system components from historical data. Science Advances10(31), 4841 (2024) https://doi.org/10.1126/sciadv.adl4841

    work page doi:10.1126/sciadv.adl4841 2024

  8. [8]

    Philosophical Transactions of the Royal Society of London, Series A: Mathematical and Physical Sciences236(767), 333–380 (1937) https: //doi.org/10.1098/rsta.1937.0005

    Neyman, J.: Outline of a Theory of Statistical Estimation Based on the Classical Theory of Probability. Philosophical Transactions of the Royal Society of London, Series A: Mathematical and Physical Sciences236(767), 333–380 (1937) https: //doi.org/10.1098/rsta.1937.0005

    work page doi:10.1098/rsta.1937.0005 1937

  9. [9]

    Psychonomic Bulletin & Review 23(1), 103–123 (2015) https://doi.org/10.3758/s13423-015-0947-8

    Morey, R.D., Hoekstra, R., Rouder, J.N., Lee, M.D., Wagenmakers, E.-J.: The fal- lacy of placing confidence in confidence intervals. Psychonomic Bulletin & Review 23(1), 103–123 (2015) https://doi.org/10.3758/s13423-015-0947-8

    work page doi:10.3758/s13423-015-0947-8 2015

  10. [10]

    The Annals of Applied Statistics6(4) (2012) https://doi.org/10.1214/12-AOAS571

    Efron, B.: Bayesian inference and the parametric bootstrap. The Annals of Applied Statistics6(4) (2012) https://doi.org/10.1214/12-AOAS571

    work page doi:10.1214/12-aoas571 2012

  11. [11]

    Test3(1), 5–124 (1994) https://doi.org/10.1007/BF02562676

    Berger, J.O., Moreno, E., Pericchi, L.R., Bayarri, M.J., Bernardo, J.M., Cano, J.A., De la Horra, J., Mart´ ın, J., R´ ıos-Ins´ ua, D., Betr` o, B., Dasgupta, A., Gustafson, P., Wasserman, L., Kadane, J.B., Srinivasan, C., Lavine, M., O’Hagan, A., Polasek, W., Robert, C.P., Goutis, C., Ruggeri, F., Salinetti, G., Siva- ganesan, S.: An overview of robust B...

    work page doi:10.1007/bf02562676 1994

  12. [13]

    Journal of Climate23(21), 5678–5698 (2010) https: //doi.org/10.1175/2010JCLI3389.1

    Kanzow, T., Cunningham, S.A., Johns, W.E., Hirschi, J.J.-M., Marotzke, J., Baringer, M.O., Meinen, C.S., Chidichimo, M.P., Atkinson, C., Beal, L.M., Bryden, H.L., Collins, J.: Seasonal variability of the atlantic meridional over- turning circulation at 26.5°n. Journal of Climate23(21), 5678–5698 (2010) https: //doi.org/10.1175/2010JCLI3389.1

    work page doi:10.1175/2010jcli3389.1 2010

  13. [14]

    Frontiers in Marine Science6(JUN), 1–18 (2019) https: //doi.org/10.3389/fmars.2019.00260 13

    Frajka-Williams, E., Ansorge, I.J., Baehr, J., Bryden, H.L., Chidichimo, M.P., Cunningham, S.A., Danabasoglu, G., Dong, S., Donohue, K.A., Elipot, S., Heim- bach, P., Holliday, N.P., Hummels, R., Jackson, L.C., Karstensen, J., Lankhorst, M., Le Bras, I.A., Susan Lozier, M., McDonagh, E.L., Meinen, C.S., Mercier, H., Moat, B.I., Perez, R.C., Piecuch, C.G.,...

    work page doi:10.3389/fmars.2019.00260 2019

  14. [15]

    https://doi.org/10.5285/ cc1e34b3-3385-662b-e053-6c86abc03444

    Frajka-Williams, E., Moat, B.I., Smeed, D.A., Rayner, D., Johns, W.E., Baringer, M.O., Volkov, D., Collins, J.: Atlantic meridional overturning cir- culation observed by the RAPID-MOCHA-WBTS (RAPID-Meridional Over- turning Circulation and Heatflux Array-Western Boundary Time Series) array at 26N from 2004 to 2020 (v2020.1) (2021). https://doi.org/10.5285/...

    2004

  15. [16]

    Nature438, 655–657 (2005) https: //doi.org/10.1038/nature04385

    Bryden, H.L., Longworth, H.R., Cunningham, S.A.: Slowing of the atlantic meridional overturning circulation at 25°n. Nature438, 655–657 (2005) https: //doi.org/10.1038/nature04385

    work page doi:10.1038/nature04385 2005

  16. [17]

    Rahmstorf, J

    Rahmstorf, S., Box, J.E., Feulner, G., Mann, M.E., Robinson, A., Rutherford, S., Schaffernicht, E.J.: Exceptional twentieth-century slowdown in Atlantic Ocean overturning circulation. Nature Climate Change5(5), 475–480 (2015) https:// doi.org/10.1038/nclimate2554

    work page doi:10.1038/nclimate2554 2015

  17. [18]

    Nature556(7700), 191–196 (2018) https://doi.org/10.1038/s41586-018-0006-5

    Caesar, L., Rahmstorf, S., Robinson, A., Feulner, G., Saba, V.: Observed finger- print of a weakening Atlantic Ocean overturning circulation. Nature556(7700), 191–196 (2018) https://doi.org/10.1038/s41586-018-0006-5

    work page doi:10.1038/s41586-018-0006-5 2018

  18. [19]

    Drijfhout, S., Oldenborgh, G.J., Cimatoribus, A.: Is a decline of AMOC causing the warming hole above the North Atlantic in observed and modeled warming patterns? Journal of Climate25(24), 8373–8379 (2012) https://doi.org/10.1175/ JCLI-D-12-00490.1

    2012

  19. [20]

    Climate Dynamics50(7-8), 3063–3080 (2018) https:// doi.org/10.1007/s00382-017-3793-8

    Menary, M.B., Wood, R.A.: An anatomy of the projected North Atlantic warming hole in CMIP5 models. Climate Dynamics50(7-8), 3063–3080 (2018) https:// doi.org/10.1007/s00382-017-3793-8

    work page doi:10.1007/s00382-017-3793-8 2018

  20. [21]

    Science Advances6(26), 2–10 (2020) https://doi.org/10.1126/sciadv.aaz4876

    Liu, W., Fedorov, A.V., Xie, S.P., Hu, S.: Climate impacts of a weakened Atlantic meridional overturning circulation in a warming climate. Science Advances6(26), 2–10 (2020) https://doi.org/10.1126/sciadv.aaz4876

    work page doi:10.1126/sciadv.aaz4876 2020

  21. [22]

    1029/2020GL090888

    Little, C.M., Zhao, M., Buckley, M.W.: Do Surface Temperature Indices Reflect Centennial-Timescale Trends in Atlantic Meridional Overturning Circulation Strength? Geophysical Research Letters47(22), 1–10 (2020) https://doi.org/10. 1029/2020GL090888

    2020

  22. [23]

    Geophysical Research Letters45(16), 8547–8556 (2018) https: //doi.org/10.1029/2018GL078104

    Jackson, L.C., Wood, R.A.: Hysteresis and Resilience of the AMOC in an Eddy- Permitting GCM. Geophysical Research Letters45(16), 8547–8556 (2018) https: //doi.org/10.1029/2018GL078104

    work page doi:10.1029/2018gl078104 2018

  23. [24]

    Climate Dynamics58(9-10), 2249–2267 (2022) https://doi.org/10.1007/s00382-021-06003-4

    Li, L., Lozier, M.S., Li, F.: Century-long cooling trend in subpolar North Atlantic forced by atmosphere: an alternative explanation. Climate Dynamics58(9-10), 2249–2267 (2022) https://doi.org/10.1007/s00382-021-06003-4

    work page doi:10.1007/s00382-021-06003-4 2022

  24. [25]

    Geophysical Research Letters49(19) (2022) https://doi.org/10.1029/2022GL100420

    He, C., Clement, A.C., Cane, M.A., Murphy, L.N., Klavans, J.M., Fenske, T.M.: A 14 North Atlantic Warming Hole Without Ocean Circulation. Geophysical Research Letters49(19) (2022) https://doi.org/10.1029/2022GL100420

    work page doi:10.1029/2022gl100420 2022

  25. [26]

    Geophysical Research Letters49(16) (2022) https://doi.org/10.1029/2022GL097967

    Ferster, B.S., Simon, A., Fedorov, A., Mignot, J., Guilyardi, E.: Slowdown and Recovery of the Atlantic Meridional Overturning Circulation and a Persistent North Atlantic Warming Hole Induced by Arctic Sea Ice Decline. Geophysical Research Letters49(16) (2022) https://doi.org/10.1029/2022GL097967

    work page doi:10.1029/2022gl097967 2022

  26. [27]

    Journal of Climate36(6), 1–31 (2022) https://doi.org/10.1175/JCLI-D-22-0222.1

    Ghosh, R., Putrasahan, D., Manzini, E., Lohmann, K., Keil, P., Hand, R., Bader, J., Matei, D., Jungclaus, J.H.: Two distinct phases of North Atlantic Eastern Subpolar Gyre and Warming Hole evolution under Global Warming. Journal of Climate36(6), 1–31 (2022) https://doi.org/10.1175/JCLI-D-22-0222.1

    work page doi:10.1175/jcli-d-22-0222.1 2022

  27. [28]

    Nature Climate Change 10(7), 667–671 (2020) https://doi.org/10.1038/s41558-020-0819-8

    Keil, P., Mauritsen, T., Jungclaus, J., Hedemann, C., Olonscheck, D., Ghosh, R.: Multiple drivers of the North Atlantic warming hole. Nature Climate Change 10(7), 667–671 (2020) https://doi.org/10.1038/s41558-020-0819-8

    work page doi:10.1038/s41558-020-0819-8 2020

  28. [29]

    Science363(6426), 516–521 (2019) https://doi.org/10.1126/science.aau6592

    Lozier, M.S., Li, F., Bacon, S., Bahr, F., Bower, A.S., Cunningham, S.A., De Jong, M.F., De Steur, L., DeYoung, B., Fischer, J., Gary, S.F., Greenan, B.J.W., Holliday, N.P., Houk, A., Houpert, L., Inall, M.E., Johns, W.E., John- son, H.L., Johnson, C., Karstensen, J., Koman, G., Le Bras, I.A., Lin, X., Mackay, N., Marshall, D.P., Mercier, H., Oltmanns, M....

    work page doi:10.1126/science.aau6592 2019

  29. [30]

    Geophysical Research Letters49(17), 1–11 (2022) https://doi.org/10.1029/2022GL099133

    Chafik, L., Holliday, N.P., Bacon, S., Rossby, T.: Irminger Sea Is the Center of Action for Subpolar AMOC Variability. Geophysical Research Letters49(17), 1–11 (2022) https://doi.org/10.1029/2022GL099133

    work page doi:10.1029/2022gl099133 2022

  30. [31]

    Nature Communications8 (2017) https://doi.org/10.1038/ncomms14375

    Sgubin, G., Swingedouw, D., Drijfhout, S., Mary, Y., Bennabi, A.: Abrupt cooling over the North Atlantic in modern climate models. Nature Communications8 (2017) https://doi.org/10.1038/ncomms14375

    work page doi:10.1038/ncomms14375 2017

  31. [32]

    Annals of the New York Academy of Sciences1504(1), 187–201 (2021) https://doi.org/10.1111/nyas.14659

    Swingedouw, D., Bily, A., Esquerdo, C., Borchert, L.F., Sgubin, G., Mignot, J., Menary, M.: On the risk of abrupt changes in the North Atlantic subpolar gyre in CMIP6 models. Annals of the New York Academy of Sciences1504(1), 187–201 (2021) https://doi.org/10.1111/nyas.14659

    work page doi:10.1111/nyas.14659 2021

  32. [33]

    Science377(6611) (2022) https://doi.org/10.1126/science.abn7950

    McKay, D.I.A., Staal, A., Abrams, J.F., Winkelmann, R., Sakschewski, B., Lori- ani, S., Fetzer, I., Cornell, S.E., Rockstr¨ om, J., Lenton, T.M.: Exceeding 1.5 ◦C global warming could trigger multiple climate tipping points. Science377(6611) (2022) https://doi.org/10.1126/science.abn7950

    work page doi:10.1126/science.abn7950 2022

  33. [34]

    Nature Communications14(1), 8344 (2023) https://doi.org/10.1038/ s41467-023-44046-9

    Ben-Yami, M., Skiba, V., Bathiany, S., Boers, N.: Uncertainties in critical slowing 15 down indicators of observation-based fingerprints of the Atlantic Overturning Cir- culation. Nature Communications14(1), 8344 (2023) https://doi.org/10.1038/ s41467-023-44046-9

    2023

  34. [35]

    Journal of Climate26(22), 9155–9174 (2013) https://doi.org/ 10.1175/JCLI-D-12-00762.1

    Roberts, C.D., Garry, F.K., Jackson, L.C.: A multimodel study of sea surface temperature and subsurface density fingerprints of the Atlantic meridional over- turning circulation. Journal of Climate26(22), 9155–9174 (2013) https://doi.org/ 10.1175/JCLI-D-12-00762.1

    work page doi:10.1175/jcli-d-12-00762.1 2013

  35. [36]

    Journal of Climate33(16), 7027–7044 (2020) https://doi.org/10.1175/ JCLI-D-20-0034.1

    Jackson, L.C., Wood, R.A.: Fingerprints for Early Detection of Changes in the AMOC. Journal of Climate33(16), 7027–7044 (2020) https://doi.org/10.1175/ JCLI-D-20-0034.1

    2020

  36. [38]

    Nature Climate Change14(January), 43–48 (2023) https://doi.org/10.1038/s41558-023-01878-z

    Boers, N.: Reply to: Evidence lacking for a pending collapse of the Atlantic Meridional Overturning Circulation. Nature Climate Change14(January), 43–48 (2023) https://doi.org/10.1038/s41558-023-01878-z

    work page doi:10.1038/s41558-023-01878-z 2023

  37. [39]

    Scientific Data10(1), 1–16 (2023) https://doi.org/10

    Lundstad, E., Brugnara, Y., Pappert, D., Kopp, J., Samakinwa, E., H¨ urzeler, A., Andersson, A., Chimani, B., Cornes, R., Demar´ ee, G., Filipiak, J., Gates, L., Ives, G.L., Jones, J.M., Jourdain, S., Kiss, A., Nicholson, S.E., Przybylak, R., Jones, P., Rousseau, D., Tinz, B., Rodrigo, F.S., Grab, S., Dom´ ınguez-Castro, F., Slonosky, V., Cooper, J., Brun...

    2023

  38. [40]

    Journal of Geophysical Research: Atmospheres 124(14), 7719–7763 (2019) https://doi.org/10.1029/2018JD029867

    Kennedy, J.J., Rayner, N.A., Atkinson, C.P., Killick, R.E.: An Ensemble Data Set of Sea Surface Temperature Change From 1850: The Met Office Hadley Centre HadSST.4.0.0.0 Data Set. Journal of Geophysical Research: Atmospheres 124(14), 7719–7763 (2019) https://doi.org/10.1029/2018JD029867

    work page doi:10.1029/2018jd029867 2019

  39. [41]

    Journal of Geophysical Research: Atmospheres108(14) (2003) https://doi.org/10.1029/ 2002jd002670

    Rayner, N.A., Parker, D.E., Horton, E.B., Folland, C.K., Alexander, L.V., Rowell, D.P., Kent, E.C., Kaplan, A.: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. Journal of Geophysical Research: Atmospheres108(14) (2003) https://doi.org/10.1029/ 2002jd002670

    2003

  40. [42]

    Earth Sys- tem Dynamics14(1), 173–183 (2023) https://doi.org/10.5194/esd-14-173-2023

    Smith, T., Zotta, R.-M., Boulton, C.A., Lenton, T.M., Dorigo, W., Boers, N.: Reli- ability of resilience estimation based on multi-instrument time series. Earth Sys- tem Dynamics14(1), 173–183 (2023) https://doi.org/10.5194/esd-14-173-2023

    work page doi:10.5194/esd-14-173-2023 2023

  41. [43]

    Journal of Climate30(20), 8179–8205 (2017) https://doi.org/10

    Huang, B., Thorne, P.W., Banzon, V.F., Boyer, T., Chepurin, G., Lawrimore, J.H., Menne, M.J., Smith, T.M., Vose, R.S., Zhang, H.M.: Extended reconstructed 16 Sea surface temperature, Version 5 (ERSSTv5): Upgrades, validations, and inter- comparisons. Journal of Climate30(20), 8179–8205 (2017) https://doi.org/10. 1175/JCLI-D-16-0836.1

    2017

  42. [44]

    Journal of Geophysical Research: Atmospheres126(3), 1–28 (2021) https: //doi.org/10.1029/2019JD032361

    Morice, C.P., Kennedy, J.J., Rayner, N.A., Winn, J.P., Hogan, E., Killick, R.E., Dunn, R.J.H., Osborn, T.J., Jones, P.D., Simpson, I.R.: An Updated Assess- ment of Near-Surface Temperature Change From 1850: The HadCRUT5 Data Set. Journal of Geophysical Research: Atmospheres126(3), 1–28 (2021) https: //doi.org/10.1029/2019JD032361

    work page doi:10.1029/2019jd032361 2021

  43. [45]

    The Annals of Statistics52(2) (2024) https://doi.org/10.1214/24-AOS2371

    Pilipovic, P., Samson, A., Ditlevsen, S.: Parameter estimation in nonlinear mul- tivariate stochastic differential equations based on splitting schemes. The Annals of Statistics52(2) (2024) https://doi.org/10.1214/24-AOS2371

    work page doi:10.1214/24-aos2371 2024

  44. [46]

    Statistics and Computing18(4), 435–446 (2008) https: //doi.org/10.1007/s11222-008-9104-9

    ter Braak, C.J.F., Vrugt, J.A.: Differential Evolution Markov Chain with snooker updater and fewer chains. Statistics and Computing18(4), 435–446 (2008) https: //doi.org/10.1007/s11222-008-9104-9

    work page doi:10.1007/s11222-008-9104-9 2008

  45. [47]

    Comprehensive R Archive Network (2017)

    Hartig, F., Minunno, F., Paul, S.: BayesianTools: General-Purpose MCMC and SMC Samplers and Tools for Bayesian Statistics. Comprehensive R Archive Network (2017). https://doi.org/10.32614/CRAN.package.BayesianTools

    work page doi:10.32614/cran.package.bayesiantools 2017

  46. [48]

    Journal of Geophysical Research: Oceans128(4), 2022–019487 (2023) https://doi.org/10.1029/2022JC019487 https://agupubs.onlinelibrary.wiley.com/doi/pdf/10.1029/2022JC019487

    P´ erez-Hern´ andez, M.D., Hern´ andez-Guerra, A., Cana-Cascallar, L., Arum´ ı- Planas, C., Ca´ ınzos, V., Gonz´ alez-Santana, A.J., Guti´ errez-Guerra, M.A., Mart´ ınez-Marrero, A., Mosquera Gim´ enez, A., Presas Navarro, C., Santana- Toscano, D., V´ elez-Belch´ ı, P.: The seasonal cycle of the eastern boundary currents of the north atlantic subtropical ...

    work page doi:10.1029/2022jc019487 2022

  47. [49]

    The index defined in DD23 (SPG SSTs minus twice the global mean SST)

  48. [50]

    The classical SPG-based AMOC index (SPG SSTs minus the global mean SST) [18]

  49. [51]

    Additionally, we investigated the propagation of observational uncertainty by con- sidering a 200-member uncertainty ensemble of HadCRUT5 [44]

    The dipole fingerprint, calculated as the difference between averaged SSTs in the North Atlantic (45 ◦–80◦N, 70 ◦W–30◦E) and the South Atlantic (0 ◦–45◦S, 70 ◦W– 30◦E) [35, 36]. Additionally, we investigated the propagation of observational uncertainty by con- sidering a 200-member uncertainty ensemble of HadCRUT5 [44]. The tipping time tc was estimated b...

    1924