ClearFactAll topics

Do critical climate tipping points render global warming largely irreversible within 100 years even if emissions stop today?

5 peer-reviewed studies for, 6 against

Key takeaways

Stronger evidence: For

Verdict

The proposition asks whether warming and its major consequences would be largely irreversible over the next century even if emissions stopped now. On that framing, the evidence for is somewhat stronger, although the case is not one-sided.

The strongest support comes from two independent lines. Solomon and colleagues show that CO2-driven warming stays close to its peak for about 1,000 years, because ocean heat uptake offsets the slow decline in atmospheric CO2. Garbe and colleagues find strong hysteresis in the Antarctic Ice Sheet: ice lost past certain thresholds does not regrow even if temperatures return to pre-industrial levels, and West Antarctica is committed to long-term collapse at about 2°C. Armstrong McKay and colleagues add that several tipping elements are already within their uncertainty range at about 1.1°C of warming.

The evidence against is credible but mostly answers a different question. MacDougall's multi-model ZECMIP result (about -0.07°C after 50 years) and Matthews and Caldeira show that warming stops soon after emissions stop, so there is no hidden runaway warming. But a plateau is not a reversal. Stable temperatures at the peak level are consistent with Solomon's finding that warming is not undone. Tietsche and colleagues show Arctic summer sea ice recovers quickly, but that is one fast-responding element with a clear stabilizing feedback, and it is not the ice sheets or circulation systems that matter most for long-term commitments.

Ritchie and colleagues make a more important point: a brief overshoot of a threshold need not cause tipping. That limits how automatic tipping is, but it depends on temperatures being brought back down in time, which is a different scenario from emissions merely stopping at a warm peak. Weijer's review shows models do not project an AMOC collapse this century, though it concedes models may be biased toward stability, and Boers finds early-warning signals. The AMOC question is therefore unresolved rather than refuted.

Overall, the for side shows that warming and ice-sheet commitments are effectively irreversible on human timescales. The against side mainly shows that tipping is not inevitable or instantaneous and that warming does not keep climbing, so the for side's case is stronger.

Strongest study on each side

Ranked by citation count.

Strongest study for

Tipping elements in the Earth's climate system (Proceedings of the National Academy of Sciences, 2008)

3,982 citations

Lenton and colleagues introduced the formal concept of "tipping elements" — subsystems such as Arctic sea ice, the Greenland ice sheet, the West Antarctic ice sheet, the Atlantic overturning circulation and the Amazon rainforest — that can be switched into a qualitatively different state by small perturbations. Drawing on expert elicitation, they concluded several could be tipped within this century under plausible warming, with consequences that would persist far beyond the forcing that triggered them.

Strongest study against

Stabilizing climate requires near-zero emissions (Geophysical Research Letters, 2008)

537 citations

Matthews and Caldeira use an Earth system model to show that when CO2 emissions are eliminated, global temperature stabilizes almost immediately rather than continuing to climb. The long-assumed "warming in the pipeline" from past emissions is largely cancelled by declining atmospheric CO2, meaning the future trajectory of warming is controlled by future emissions rather than locked in by past ones.

Open the interactive page →

The evidence

For (5)

Tipping elements in the Earth's climate system (Proceedings of the National Academy of Sciences, 2008)

3,982 citations

Lenton and colleagues introduced the formal concept of "tipping elements" — subsystems such as Arctic sea ice, the Greenland ice sheet, the West Antarctic ice sheet, the Atlantic overturning circulation and the Amazon rainforest — that can be switched into a qualitatively different state by small perturbations. Drawing on expert elicitation, they concluded several could be tipped within this century under plausible warming, with consequences that would persist far beyond the forcing that triggered them.

Irreversible climate change due to carbon dioxide emissions (Proceedings of the National Academy of Sciences, 2009)

3,217 citations

Solomon, Plattner, Knutti and Friedlingstein show that the climate change caused by CO2 is largely irreversible for 1,000 years after emissions stop. As ocean heat uptake slows, it offsets the slow decline of atmospheric CO2, so surface temperatures stay close to their peak rather than falling. Dry-season rainfall losses and thermal sea-level rise are similarly locked in, meaning stopping emissions halts further warming but does not undo it.

Exceeding 1.5°C global warming could trigger multiple climate tipping points (Science, 2022)

1,636 citations

Armstrong McKay and colleagues reassessed over 200 studies and found that five tipping elements — including the Greenland and West Antarctic ice sheets, tropical coral reefs, and abrupt boreal permafrost thaw — are already within their uncertainty range at today's ~1.1°C of warming, and become "likely" past 1.5°C. Several of these, once triggered, commit the planet to changes that unfold over centuries to millennia regardless of later emission cuts.

Observation-based early-warning signals for a collapse of the Atlantic Meridional Overturning Circulation (Nature Climate Change, 2021)

438 citations

Boers analyzes eight independent sea-surface temperature and salinity indices and finds statistically significant early-warning signals — rising variance and autocorrelation — consistent with the AMOC losing stability over the past century. This suggests the circulation may be closer to a critical transition than climate models indicate, and a collapse would be effectively irreversible on centennial timescales.

The hysteresis of the Antarctic Ice Sheet (Nature, 2020)

315 citations

Garbe, Albrecht, Levermann, Donges and Winkelmann simulate the Antarctic Ice Sheet across a wide range of warming levels and find strong hysteresis: once ice is lost past certain thresholds, it does not regrow even if temperatures return to pre-industrial levels. At 2°C of sustained warming, West Antarctica is committed to long-term collapse, contributing meters of sea-level rise that cannot be reversed on human timescales.

Against (6)

Stabilizing climate requires near-zero emissions (Geophysical Research Letters, 2008)

537 citations

Matthews and Caldeira use an Earth system model to show that when CO2 emissions are eliminated, global temperature stabilizes almost immediately rather than continuing to climb. The long-assumed "warming in the pipeline" from past emissions is largely cancelled by declining atmospheric CO2, meaning the future trajectory of warming is controlled by future emissions rather than locked in by past ones.

Stability of the Atlantic Meridional Overturning Circulation: A Review and Synthesis (Journal of Geophysical Research: Oceans, 2019)

314 citations

Weijer and colleagues review theory, observations and models of AMOC stability. State-of-the-art climate models project a weakening of the AMOC this century but not a collapse, and no comprehensive model shows an abrupt shutdown under 21st-century forcing. While models may be biased toward stability, the current evidence does not establish that an AMOC tipping point will be crossed within 100 years.

Is there warming in the pipeline? A multi-model analysis of the Zero Emissions Commitment from CO2 (Biogeosciences, 2020)

264 citations

MacDougall and colleagues compare 18 Earth system models in the ZECMIP experiment, where CO2 emissions are abruptly stopped. The multi-model mean Zero Emissions Commitment 50 years after emissions cease is close to zero (about -0.07°C), with a range of -0.36 to +0.29°C. Once emissions stop, warming essentially stops too — there is no evidence of large committed warming or runaway feedbacks driving temperatures upward.

Overshooting tipping point thresholds in a changing climate (Nature, 2021)

223 citations

Ritchie, Clarke, Cox and Huntingford show that crossing a tipping threshold does not automatically trigger a transition. Because many tipping elements respond slowly, a temporary overshoot can be survived without tipping if the threshold is exceeded only briefly and temperatures are brought back down in time. This undercuts the idea that passing a threshold instantly locks in irreversible change.

Recovery mechanisms of Arctic summer sea ice (Geophysical Research Letters, 2011)

216 citations

Tietsche, Notz, Jungclaus and Marotzke artificially remove all Arctic summer sea ice in a climate model and find it recovers within about two years to the state set by the prevailing climate. Strong stabilizing feedbacks — chiefly increased winter heat loss from open water — mean summer sea ice has no tipping point, contradicting the popular claim that its loss is irreversible.

The debate

Lisa: Here's the uncomfortable fact. Solomon and colleagues showed in PNAS that the warming we've caused from CO2 is irreversible for about a thousand years. Stop all emissions today and temperatures don't come back down, because the ocean stops absorbing heat as fast as CO2 declines. That alone makes the warming we already have essentially permanent on any human timescale.

James: That's a statement about warming not reversing, not about tipping points making it worse. And the newer work is reassuring on the key question. MacDougall's ZECMIP study compared 18 Earth system models, cut emissions to zero, and found the warming commitment is close to zero. When emissions stop, warming stops. Matthews and Caldeira found the same in 2008. There's no big hidden pipeline of warming.

Lisa: But the question is about tipping points, and those are exactly what models struggle with. Armstrong McKay's 2022 Science review found that five tipping elements, including the Greenland and West Antarctic ice sheets, are already within their uncertainty range at today's warming. Garbe showed in Nature that the Antarctic ice sheet has strong hysteresis. Once it goes, it doesn't come back even if you cool the planet.

James: The ice sheets are the strongest part of your case, but look at the timescale. Ice sheet collapse plays out over centuries to millennia. It commits us to sea-level rise, but it doesn't make global temperatures run away within 100 years. And Ritchie and colleagues showed in Nature that crossing a threshold doesn't mean you've tipped. Slow systems can survive a brief overshoot if you bring temperatures back down in time.

Lisa: What about the AMOC? Boers found statistically significant early-warning signals in eight independent observational indices. The circulation appears to be losing stability. Models may simply be too stable.

James: That's a real concern, but Weijer's review in JGR Oceans found that state-of-the-art models project weakening, not collapse, this century. And people often point to Arctic sea ice as the classic irreversible change, yet Tietsche showed that if you remove all summer sea ice in a model, it comes back within about two years. Not every scary threshold is a true tipping point.

Lisa: So we agree the warming we've caused won't reverse on its own, and that the ice sheets may already be committed to long-term loss. Shouldn't that count as largely irreversible?

James: It counts as "doesn't reverse without carbon removal," which is serious. But the claim was that tipping points render warming largely irreversible within 100 years. The evidence says the opposite: stop emissions and warming stops, and most tipping elements stay untripped at today's levels. The danger is in the warming we haven't done yet.

Lisa: Which is exactly why every tenth of a degree matters. I'll take that as common ground.

Frequently asked questions

Will global temperatures keep rising if we stop emitting CO2?

Multi-model research finds warming essentially stops once CO2 emissions stop, with an average change of about -0.07°C after 50 years. However, other modeling shows temperatures stay near their peak rather than falling back.

Can melted Antarctic ice grow back if the climate cools?

Simulations show strong hysteresis in the Antarctic Ice Sheet, meaning ice lost past certain thresholds does not regrow even if temperatures return to pre-industrial levels. At 2°C of sustained warming, West Antarctica is committed to long-term collapse.

Is the loss of Arctic summer sea ice permanent?

One modeling study removed all Arctic summer sea ice and found it recovered within about two years to the state set by the prevailing climate. Stabilizing feedbacks suggest summer sea ice has no tipping point.

Will the Atlantic Meridional Overturning Circulation (AMOC) collapse this century?

Climate models project weakening but not collapse this century, though a review notes models may be biased toward stability. Observation-based early-warning signals suggest the circulation may be losing stability, so the question remains unresolved.

Related questions