‘Notoriously difficult to investigate and even more difficult to predict’: Thomas Stocker on tipping points

The following conversation with Thomas Stocker, a professor of Climate and Environmental Physics at the University of Bern and lead author of a 2024 article that argued for the necessity of a robust assessment of tipping points by the Intergovernmental Panel on Climate Change to establish much needed consensus on the topic, is one of several interviews conducted for the March issue of the Bulletin of the Atomic Scientists, which is all about tipping points. You can find the other conversations here, and the rest of the magazine issue here. This interview has been lightly edited and condensed.

Jessica McKenzie: Could we start by defining tipping points?

Thomas Stocker: There is, of course, the mathematical definition of a tipping point, that’s quite clear. It’s dynamical systems that have more than one stable equilibrium, and where these transitions occur from one equilibrium to the next—when you change an external parameter or a forcing or whatever—that point is called a tipping point.

Usually it’s associated with hysteresis—I don’t know whether you’re familiar with that notion. It is the physical basis of magnetic information storage: Magnetic moments (the strength and orientation of a magnetic source) orient along an external magnetic field, and if that external field is switched off, the moments remain oriented due to their magnetic interaction. It requires an external magnetic field in the opposite direction to switch orientation. It turns out that hysteresis, very similar to the physical phenomenon, occurs in many non-linear dynamical systems. Hysteretic behavior has also been shown in earth system models across the entire hierarchy of model complexity, from very simple models to the comprehensive coupled general circulation models.

In the public discussion, ‘tipping point’ has assumed a much broader meaning. In earlier IPCC [Intergovernmental Panel on Climate Change] assessments, we called them “surprises in the climate system,” where, due to slow changes in certain parameters or in certain forcing, like the CO2 or greenhouse gas concentrations in the atmosphere, the system would react by crossing such tipping points. The classical one is the Atlantic Meridional Overturning Circulation, which was really coming into the focus in the mid-1980s, when ice cores from Greenland showed a very dynamical behavior of the climate: A sequence of abrupt warmings and coolings in the temperature indicators measured on the Greenland ice cores. This suggested that the ocean circulation, bringing heat northward in the Atlantic, may be responsible for that. A reduction of this transport would result in a cooling, while a rapid resumption of the circulation could generate an abrupt warming.

This is still a very actively researched topic, now in the context of crossing possible tipping points due to anthropogenic heating. By now we have dedicated observational networks in the Atlantic Ocean, to monitor the behavior of this meridional overturning circulation and eventually provide information as to whether that system does show an approach towards a tipping point.

I think tipping points are equally relevant at a much smaller scale than the global-scale tipping points associated with AMOC, the West Antarctic ice sheet, or others. Consider precipitation characteristics (amount, timing, extent), or water resources in general. If a community dependent on the monsoon system, say in India, is experiencing a fundamental step change in the availability of water supplies, they would certainly interpret such a change as the crossing of a tipping point. In the future, regional tipping points and their impacts will require much more attention.

McKenzie: One thing that I wished I had gotten into this issue was something on biodiversity.

Stocker: Absolutely, the tipping point concept is not reserved to the physical sciences. In fact, equally important are all complex systems that support habitability on this planet. The well-functioning of ecosystems is absolutely central to our survival. Consider an ecosystem with multiple participants that all have different sensitivities to local climatic conditions. Responding to the warming, the more sensitive species will migrate to colder locations (if they can), the less sensitive stay where they are. This way you could lose key elements of an ecosystem, just as if you tore apart the ecosystem. This could ultimately lead to the gradual dysfunction of an ecosystem, culminating in its eventual loss.

McKenzie: The history of tipping points can be hard to tease out. I found it difficult to determine whether climate scientists were using the phrase before Malcolm Gladwell popularized the term.

Stocker: It came later.

When I was a postdoc in the late 1980s, early 1990s, I was researching the North Atlantic circulation, and I was very interested in that system and its vulnerability to perturbations. One finding with a climate model of reduced complexity was that this circulation exhibited two stable states: a strong circulation and a weak or collapsed one. This is the essence of hysteresis. The driving parameter was the Atlantic freshwater balance, precipitation minus evaporation, that induced the crossing of a tipping point. Our results provided a link between very simple box models, presented in the early 1960s and the simulations with a state-of-the-art climate model by [Syukuro] Suki Manabe, the Nobel Laureate of the Physics Nobel Prize in 2021, with whom we were in very lively and engaged contact.

But we never mentioned tipping points then. We said, “might the climate system have multiple equilibria, and might that actually lead to surprises in the climate system?” Already in 2001, in the third IPCC assessment, we talked about surprises in the climate system. The tipping point vocabulary came later, around 2008, with the article by Tim Lenton. That notion quickly gained traction and resonated within the community. It was a concept that, it seems, proved easier to communicate to a wider audience than complex ideas like multiple equilibria or abrupt climate change. Everybody had some kind of imagination of tipping points.

McKenzie: It’s a very evocative phrase. How has the understanding of tipping points changed since 2008, when it was introduced? Has it evolved significantly?

Stocker: Absolutely. We’ve seen quite an activity of research worldwide in this topic, with the newer generation of the coupled climate models that offered better resolution and more realism. There have been high times and low times of tipping point research, with the community vacillating between, “yes, it’s important, yes, it’s prevalent,” or “maybe not—it’s a complicated system, and therefore tipping is perhaps not so prominent.” And so, the scientific community has not yet fully come to terms with it.

What I perceive now is that there is a gap opening or widening between those colleagues—respected colleagues—who say, well, “tipping is probably not the primary issue in the climate system,” and others who find all kinds of indicators that things may tip. You can see this in what the high-profile journals publish at a regular pace: One paper will argue things are on the way to tipping, and then subsequent papers say, “Ah, no, but looking at that indicator, we don’t see that the system is already tipping.”

It’s a very lively scientific debate, and it is a debate that can become confusing for the public. Therefore, a very careful assessment is required. That’s why I proposed to IPCC to address this issue head on in the seventh cycle that started last year. Switzerland promoted the proposal of a special report that would allow the scientific community to congregate under that formal umbrella with the task of finding consensus on these different possible tipping elements: the observational evidence for it, the paleoclimatic evidence for it, the evidence in model projections, and then carry out a detailed scientific assessment in the very formalized framework of a special report. Now unfortunately, last year, IPCC decided not to commission any new special reports for the seventh cycle.

McKenzie: Do they usually do one?

Stocker: The sixth cycle did three different special reports. It’s always, of course, a significant burden to the scientific community, but it’s also a unique opportunity to shed light on complex issues that the policy makers are interested in. This time around, they said, “Well, we already have one special report in the pipeline on climate change and cities, and so we have no capacity for any other special reports,” and so that was brushed off the table.

McKenzie: How long would it be before you could revisit?

Stocker: This cycle will last for another five years or so.

McKenzie: So, IPCC won’t do anything on tipping points for at least five years…

Stocker: Well, in December, they had the so-called scoping meeting, which is where policy makers and experts get together and discuss what chapters will be defined in the three respective reports, on the physical science, the impacts, and mitigation. They proposed a chapter on tipping points and irreversibility. And this has now been confirmed by the 62nd Plenary of the IPCC. This is sort of a second-best option, in which you have an entire chapter dedicated to that issue, which allows you to invite a good number of leading experts. Their assessment work will surely generate a kind of a consensus around these important questions.

It may also be that in some of the aspects of that question, there will not be a consensus, and they will have to report, “we are split on this, and we cannot really say for these and these reasons, and here is the research strategy that would be needed to address that more comprehensively and more conclusively.” Although there was some pushback on the term “tipping point,” it has been retained in a somewhat weakened form in the following chapter title “Abrupt changes, low-likelihood high impact events and critical thresholds, including tipping points, in the Earth system.”

McKenzie: What are the key questions that you would hope the IPCC report would address?

Stocker: For that chapter, I would hope that they would, in a systematic way, go through the literature, and address all the proposed tipping elements to date. So: AMOC, West Antarctic ice sheets, Amazonian rainforest, boreal permafrost, monsoon systems, etc. Then, home in on what that means, if such a point were crossed, what impact this would have on the regional climate. That’s ultimately what we need to inform the public and the stakeholders about, regional climate, but all encompassing, not just the sort of the mean climate, but the weather aspects of the climate. In other words, the statistics of extreme events associated with the climate state or a regional climate state, and how that changes if a tipping point is crossed.

At the more fundamental level, I would hope that this assessment would reach a consensus on the likelihood of tipping. You know, it could be that they say, so far, we don’t really have evidence in the 21st century for AMOC to tip, but there is stronger evidence even in previous climates and under certain future scenarios, models indicate we have had, or could have, tipping. So really, a very careful and profound assessment on likelihoods and on uncertainties associated with tipping is required.

I am aware that it’s a difficult topic, and I suspect that this was part of the reason why, in the end, [the IPCC] decided not to have a special report, apart from the principal decision not to want special reports anymore, because it’s too much of a burden on the scientists. Tipping points is indeed a difficult topic, but on the other hand, you could say assessing climate sensitivities is difficult, and IPCC has addressed that in a very persistent and consistent way in all six previous assessments. This process allowed for the evolution of the consensus. Every assessment is a snapshot of the science. Successive assessments therefore map the evolution of climate science, and with it strengthen consensus on many issues.

McKenzie: If this is the only thing someone’s going to read about tipping points, what’s the main thing that they should take away?

Stocker: Going back to the very title of your publication, the Bulletin of the Atomic Scientists, I think it’s very important to say that what we are doing here is physics. It’s physics of the atmosphere, physics of the ocean, physics of the climate system. That is also conveyed by the title of the IPCC assessments of the first working group, which says we are assessing the physical science basis of climate change. When I give lectures on climate change, people are really surprised that climate modeling has received a Nobel Prize—not in statistics, that doesn’t exist—but in physics. Many people think these climate models are just statistical machines that make some extrapolation, and there’s really large uncertainties. We hear such statements constantly, fed and fueled by climate skeptics or climate deniers.

But look at weather forecasts. Don’t you consult the weather forecast every day? Aren’t these quite reliable? Aren’t we predicting the pathways of hurricanes and warning thousands of people in good time? That’s all physics. And of course, it’s uncertain, but it’s information that is crucial. The same holds true for future climate change, the same holds true for the tipping points. Ultimately, it is physics, and as it is with physics, instabilities are notoriously difficult to investigate and even more difficult to predict.