Predicting Tipping Points

(Subscribe to World Policy Journal here)

From the Spring 2015 issue "The Unknown"

By Richard Blaustein

At the time he attended the 2004 European Open Science Forum in Stockholm, German physicist Hans Joachim Schellnhuber had been speaking out publicly since the late 1990s about climate change impacting what he termed “the Achilles heels of the Earth system.”  Using science-speak, he talked of  “switch-and-choke points” and “large scale discontinuities” to describe the possibility that key natural systems whose stability we take for granted—like the Atlantic Ocean’s jet stream and a mechanism called the thermohaline circulation, which circulates warm water globally—could abruptly change under climate change stress. 

 Not surprisingly, Schellnhuber was gaining adherents among some scientists, including within the Intergovernmental Panel on Climate Change (IPCC) of which he was a part, but hardly with the broader public that is so essential if action is to be taken on this issue. Indeed, it was not until the 2004 meeting that he significantly broadened this kind of reach with one phrase he tried out on BBC reporter, Alex Kirby, who was at the Stockholm conference and interviewed Schellnhuber. “I tried to explain to him the switch-and-choke-points of the Earth system—the ice changes, the thermohaline system, and he didn’t quite get it,” recalls Schellnhuber. But “when I said we found practical ‘tipping points’ in the Earth system, then [with his article] this climate change ‘tipping points’ notion was born.” 

A decade later, climate change tipping points are now perceived as a serious concern. Heat waves and drought throughout the world indicate that climate change is having an impact that will likely worsen. The global warming effects can be bizarre: some scientists attribute the recent winter cold snaps to global warming as the Arctic system is thrown out of balance, with winter Arctic weather spilling over into the lower latitudeshence the growing frequency and intensity of the Polar Vortex. More ominous, however, are changes in the permafrost, which is already showing signs of melting in places like Alaska and Russia, where incidents of thaw have upended property and public roads. It is a bad sign. Northern thawing would release massive amounts of methane, compounding the greenhouse gas problem. The changes are no less worrisome in the planet’s southern extremes. In 2002, in Antarctica, the Larsen B Shelf, a 1,250 square mile area of ice larger than Rhode Island, collapsed and disintegrated into the ocean, validating glacioligist John Mercer’s 1978 prediction that global warming would first impact Antarctica’s outreaching ice shelves. With an eerie sense of déjà vu in May 2014, two widely-reported scientific papers predicted the ongoing loss and eventual collapse of glaciers in West Antarctica’s Amundsen Sea, causing the sea level to rise three to 12 feet. 

So far, climate change diplomacy has been ineffectual at addressing the global emissions problem, with only economic downturn having a large-scale mitigating effect—hardly the best policy choice.  However, an increasing number of scientists are now in preparation mode, hoping governments and international organizations will at least develop the means to anticipate tipping points—when natural systems flip to harsher, altered states. In fact, in this decade, a robust university-based scientific discourse on predicting tipping points is debuting. Some are urging that these investigations be broadened and that government networks prepare for a changing, less-familiar natural world.      

Predictive Science

After the 2004 BBC interview, Schellnhuber received a little more governmental support for his efforts to build collaborations between British and German climate scientists. Some 36 scientists gathered the next year at the British Embassy in Berlin to discuss the science of climate change tipping points. British climate scientist Timothy Lenton, a friend and collaborator of Schellnhuber, played a prominent role in the Berlin meeting that now appears to be a watershed event. New collaborations, publications, and wider interest grew from the Berlin Embassy gathering.

Lenton, Schellnhuber, and five other authors began working on a paper that laid out the basic science of climate change tipping points,  highlighting nine areas of concern, such as the altering of the Indian summer monsoon season, the Amazon rainforest switching over to savanna, and melting ice sheets in Greenland and Antarctica. “Tipping elements in the Earth’s climate system” even touched on early warning signals for these abrupt changes.  

Lenton, now at the University of Exeter, is proud of that meeting and his ensuing research, but recalls that early time as a kind of a wilderness spell. Not only was there little public interest in climate change tipping points, the larger climate change scientific community was resistant. “It would be fair to say there was a lot of skepticism from the conservative ranks of climate science to the whole concept of tipping points,” Lenton recalls. “We were sometimes kind of accused of alarmism to be even drawing attention to what were seen as high impact-low probability events.” 

Recently, however, much has changed. Last spring the IPCC released a report where for the first time it spoke explicitly of tipping points: “With increasing warming, some physical systems or ecosystems may be at risk of abrupt and irreversible change.” The report points out that coral reef and Arctic ecosystems are already undergoing irreversible changes. And it warns that tipping point risks “increase disproportionately” as temperature rises above 5.4 degrees Fahrenheit, “due to the potential for a large and irreversible sea-level rise from ice sheet loss.” 

Just a few months earlier, the U.S. National Research Council, too, had made headlines by releasing a report on tipping points. University of Colorado climate scientist Jim White led the committee of scientists that put together “Abrupt Impacts of Climate Change: Anticipating Surprises,” and says he briefed agencies and foundations on the findings through much of 2014. The report gave an up-to-date assessment of tipping points, with mass species extinctions and critical water shortages among the points we will witness in the next few decades. Some fearful impacts not expected this century but highly probable after the 21st century include a critical decline of oceans’ oxygen levels, which would have serious toxic effects.  

White’s group urged more work on tipping point early warning signals and proposed the creation of an Abrupt Change Early Warning System within the U.S. government, built upon existing government capacity for data collection and natural disaster prediction. The early warning system idea includes new modeling and other means for foretelling climate change tipping points. 

Lately, in fact, there has been some considerable  university-based exploration of early warning signals of abrupt change, focusing on a phenomenon called “critical slowing down,” which occurs when a system takes longer and longer to recover from a disturbance. 

Beginning this decade, a few researchers have shown that critical slowing down precedes a tipping point, such as when a microorganism population rapidly heads to extinction. The investigations raise the possibility that under climate change stress, some natural systems heading toward abrupt change would send off early warning signals. 

In the not-too-distant future, critical slowing down might very well join tipping points and become part of the climate change vernacular.

Debate Over Tipping Points

While these investigations have advanced, some scientists have criticized the way the term “tipping point” has been used, saying it is applied in a vague or overreaching way. Indeed, tipping points, sometimes depicted as a global dominos crescendo, are not so closely linked, and their time frame varies, with some way off on the horizon. Lenton says he and his Berlin colleagues clarified climate change tipping points as self-accelerating transformations set off when natural thresholds are crossed. He stresses that tipping points often infer geological timescales longer than human history and are not necessarily irreversible. Lenton admits there is “quite a bit of slightly sloppy use of the term in the scientific discourse, as well as in the wider discourse that will make some of the more precisely minded scientists uncomfortable.”

University of Maryland-Baltimore landscape ecologist Erle C. Ellis has criticized the whole Earth global tipping point scenario, pointing out that the Earth’s biosphere lacks the requisite interconnectivity for a tipping point response of global magnitude. Ellis is anything but a climate change skeptic, and he says that human-caused climate change will have serious ecological impacts. “There’s no denying the concept of tipping points, and there are tipping points in ecosystems that are well-known in nature,” says Ellis. But he points out that there are many types of responses to stress “for systems as complex as ecosystems and social-ecological systems,” and that the ecological flips that do occur will probably be varied. Because of this discontinuity, Ellis emphasizes that scientists and policy makers should concentrate on local and regional impacts. 

Similarly, glaciologist Mark Serreze, director of the National Snow and Ice Data Center, finds the term tipping point too loosely used, especially about Arctic summer ice. “That’s the issue, how you frame it,” Serreze says. “The term tipping point has not been well defined.” In the Arctic, says Serreze, “what we’ve seen is that you can have periods of rapid ice loss, but it is not a threshold so to speak—it is not really a tipping point.” Specifically, Serreze points to Arctic summer ice loss in 2007 as huge. But 2008, Serreze recalls, was a year of significant ice recovery, and 2009 was also a summer with higher sea ice. “Then we start going down again,” Serreze says.  “2012 was a record low, but then 2013 quite a recovery. Now all the way from record low to sixth lowest on the record, so it is going from really bad to a little better. It shows you have variability in the system.”

Arctic sea ice could snap back more resiliently in the summer if governments reduce greenhouse gas emissions and stabilize atmospheric levels. But the Arctic is getting warmer with massive ice loss when measured across recent decades. And governments have not shown any ability to stabilize—let alone reduce—greenhouse gas levels. The effects could be serious. Arctic scientist Julienne Stroeve points out that with reduced Arctic sea ice, a “huge amplified warming” occurs in the Arctic. “We know, in theory, that with this change in temperature gradient between the pole and the equator, [come] changes in large scale atmospheric circulation patterns,” Stroeve says. More extreme weather for the Western Hemisphere is one very real possibility.

“That is not to say that there are not systems where there may be some kind of critical threshold, such as in permafrost,” Serreze clarifies.

Once permafrost warms up enough to form an unfrozen layer, a dynamic sets in causing the permafrost to dwindle rapidly. Kevin Schaefer, a permafrost expert, adds that permafrost thawing will dramatically compound the greenhouse gas problem, as permafrost  contains twice as much carbon as currently in the atmosphere, with great potentiality for release. He points out that experts predict permafrost greenhouse gas emissions to begin mid-century. “They won’t be huge but they will continue for several centuries,” says Schaefer. “And once you turn them on, you can’t turn them off. So it is a classic tipping point.” 

While there are some points of disagreement, the science of tipping points is progressing with new knowledge gained.  White, for example, says that just in the past 10 years, the understanding of glacier dynamics has expanded dramatically. Schellnhuber is heartened and calls tipping points the “next frontier of climate change research.” 

This tipping point frontline becomes better grounded when it sticks to a system-by-system approach. In fact, with this tight focus, the science of early warning signals is advancing.  

A Background in Ecology

In 1978, ecologist C.S. Holling and colleagues changed the way ecologists view nature when they looked at Douglas fir forests prone to spruce budworm infestations. The forests had two states. In one, there is enough foliage to attract mainly birds who keep spruce budworms in check; and in the other, pests defoliate the forest and thrive until they just about eat themselves out of house-and-home, giving space for natural predators to rebound. The spruce budworm study also went into much detail explaining the mathematics of the cyclical pest outbreaks. 

The Douglar fir study is fundamental to understanding tipping points. It showed  that stresses can abruptly tip a stable ecosystem over to its “alternative state.” A pristine lake can unexpectedly flip to become a turbid and lifeless body of water; the flowing and dynamic thermohaline circulation can become stagnant, a phantom of itself, which in geological time it has done. 

Scientists, such as lake ecologists Marten Scheffer and Stephen Carpenter, have carried on from these early studies, examining the abrupt transitions found in nature. After his university training, Scheffer worked on turbid lakes in the late 1980s at the Netherlands Institute for Inland Water Management and Waste Water Treatment. Scheffer says he learned in those early days that a few European biologists concluded that lakes were “trapped in a turbid state by some feedback mechanisms” and demonstrated “you could tip those lakes back” with a sort of “shock therapy” of removing certain abundant fish. Scheffer started to experiment on nearby turbid lakes, where flourishing carp and breen species fed on food in sediments, stirring up more turbidity while foraging. By drastically removing the carp and breen species, Scheffer and colleagues were able to flip the lakes back to their clear state with a new assemblage of fish. Vegetation took hold on the lake bottom, keeping the sediment in place. “In the end, it turned out we could convincingly show that in a complex system like a lake, you could have those tipping points,” says Scheffer. He also began to work through the theory and mathematics of tipping points, which he notes were not too sophisticated. “We were not thinking of early warning signals in those days,” Scheffer explains.    

Around the same time, Stephen Carpenter was working on lake die-off  at the University of Wisconsin, trying to frame water pollution problems in a way that would help policy makers. “As a natural scientist, I knew that there were lots of phenomena in living systems that have big jumps,” says Carpenter.

Together, Carpenter, Scheffer, and a few colleagues became well known for their work on nature’s abrupt changes.  In 2003, they started thinking about how to predict these tipping points, and were drawn to the phenomenon known in physics and mathematics as “critical slowing down.”

Critical Slowing Down

Since then, a small circle of researchers have generated a vigorous discourse on the signals of an approaching tipping point and focused on critical slowing down. In a natural system, such as an ecosystem or local precipitation region, the occurrence of critical slowing down is a signal of imminent tipping point change. Critical slowing down can be thought of as a loss of resiliency with major implications. The slowing down refers to a system taking more time to stabilize itself from a formerly minor perturbation. Even if changes in recovery appear minor to the human eye, there are statistical and mathematical ways to establish that a system is in a critical slowing down mode. For example, a deprived organism’s weight gain may look fine, but subtleties in rate of return indicate otherwise. There are statistical standards that guide the researchers who look at critical slowing down.

The investigations are varied and far-reaching. For example, Lenton found evidence of critical slowing down in some points in the Earth’s tumultuous geological history. For example, he has described critical slowing as occurring in the Younger Dryas cooling period—a 1,300-year period of cold climatic conditions between 10,800 and 9,500 B.C.E. that took place within a larger warmer period. In fact, the Younger Dryas snap might have been linked to a temporary collapse of the large-scale ocean circulation. Similarly, Scheffer has begun a Netherlands Earth System Science Center project studying past and future climate change tipping points that includes a focus on critical slowing down.

The most concrete part of the critical slowing down research is the work done by researchers who have set up biology experiments to test for critical slowing down. There have been fewer than a dozen of these investigations, most involving microorganisms in a lab setting. However, their results are published in the top scientific journals with findings that shed light on how the natural world responds to subtle and profound stresses. 

Over a period of one month, Annelies Veraart, a microecologist now with the Netherlands Institute of Ecology, and colleagues subjected photosynthesizing bacteria to increasing light. At a point a light threshold was reached and the bacteria, who communally shade one another during tolerable stresses, were damaged, their self-shading ceased, and they began dying rapidly. The group propelled ultimately into complete death, indeed extinction. Veraart and colleagues were able to collect data of critical slowing down by correlating the light stresses with declines in biomass indicating weakening recovery, which could in turn be directly linked with critical slowing down, which her team measured on six occasions until the bacteria went extinct. 

“This decrease in recovery rate really started early, far from when the population collapsed,” Veraart says. “This critical slowing down starts early. You can’t really tell when you look at the population, but if you look at data of recovery rates you would see the resiliency decreasing."

Scheffer says it’s well-established that chemical systems send off critical slowing down signals and believes that this experiment led by Veraart and the other microorganism critical slowing down studies are proof-of-concept for critical slowing down in living systems. “Altogether we showed that, yes, in living systems you can indeed see these theoretically predicted phenomenon,” says Scheffer. “It is somehow surprising how universal it can be.”

Scheffer says the potential for this and other early warning methods is great. “It helps with the tool kit we have to understand which coral reefs, which lakes, tropical rain forests, which grass ecosystem or boreal forests are close to a tipping point and which are further away…In many cases you can protect the system…Like with a lake we can control the nutrients and make the ecosystem more resilient against the expected effects of climate change.”   

“I think our experiment showed you really have to monitor an ecosystem’s health.” Veraart says. “These early warning signals might show which ecosystems are most fragile and perhaps allow you to do something” to protect them.   

At Peter and Paul Lakes at the University of Notre Dame Environmental Research Center, which extends over parts of northern Minnesota and Wisconsin, Carpenter and his team explored for three years whether abrupt changes in a lake’s food chain would send out early warning signals. With Paul Lake set as the control, at the experiment’s start, Peter Lake had 39 largemouth bass. In mid-2008, Carpenter and team added 12 largemouth bass. They then added 15 more on both the 169th and 203rd days of 2009. Carpenter explains they had to add the bass slowly to get good statistical results. “We were going to have to be careful not to shoot past the threshold too fast.”

As the experiment got underway, Carpenter says that the small prey fishes—pumpkinseeds, golden shiners, fathead minnows, dace, brook stickleback, and central mudminnow—kept to the lake’s periphery to avoid being eaten. But their behavior changed a lot, and “their numbers began to dwindle, too, as the predators began to chip away on them.” Soon, zooplankton grew in size as they were not being eaten by the prey fish. As the zooplankton grew, the chlorophyll, which were the focus for critical slowing down, changed just as Carpenter’s models predicted. Over two and a half years, the frightened prey fish in the lake’s periphery embodied one ecological state about to give way to another. When the tipping point occurred in 2010, the food chain dramatically shifted from one being dominated by plankton-eating fish to one ruled by fish-eating fish.

Carpenter says his Peter Lake experiment “suggests we might have some methods to evaluate the resilience of ecosystems in a wide range of settings and to compare the resilience of different ecosystems.” However, he cautions that much more research is needed. “So far I think our experiment is the only large scale field experiment to demonstrate this phenomenon.”

“It is obvious critical slowing down should be one of the tests of whether you are approaching a tipping point,” says Schellnhuber who nevertheless believes it is conceivable that a small, apparently-stable ecosystem could be upended by a larger, encompassing landscape undergoing change. So when the larger system tips over, it carries with it the smaller bit of nature, with no advance warning. “In the end, you have to understand the system and calculate the safe space,” Schellnhuber says. 

Surprising Hope

Know your natural systems. That is the first axiom for tipping point early warning signals. And with this orientation the science of tipping points has gotten off the ground since Schellnhuber’s first 2004 BBC interview. Moreover, important scientific centers, such as the U.S. National Research Council and the IPCC have placed tipping points on their top list of concerns. Scientists are beginning to look at ecosystems’ vulnerabilities and the early warning signals they send off, which could provide the space for some local or regional action to ease problems before they become broader. 

And, notwithstanding caveats, critical slowing down is still a frontrunner as the one generic test for detecting when natural systems approach their thresholds. Both Scheffer and Lenton, along with their colleagues, say that critical slowing down does not fit every situation, but it might very well be that first across-the-board-measurement for predicting tipping points for our planet.

Lenton, for example, is especially concerned about predicting changes in the monsoon season and believes that the fast monsoon dynamic might very well send out critical slowing signals. In fact, Lenton and colleagues, in 2014, conducted an analysis of monsoons in a past ice age and also in computer modeling and found signs of critical slowing down. Progress is urgent, Lenton says, because an abrupt change for the monsoons “is the most worrisome tipping point potentially in terms of human impact.” 

White also believes that in a political culture where climate change is a partisan issue, the proposal for an Abrupt Change Early Warning System, which he calls the ACEWS,  could be a unifier. White says the ACEWS “is about resilience, which people can get behind” and that “it stands a decent chance of cutting across ideological lines, political lines, and motivating, if not change, at least then the knowledge that’s needed to effect change.” Lenton agrees and is especially happy about the early warning system idea. 

Moreover, for Lenton and for White, the time is not too late, and they feel there is still opportunity for mitigation—actually lowering greenhouse gas emissions and then bringing down atmospheric levels. “I hope we do get together and do mitigate,” says Lenton. “Then we are essentially casting off a whole range of future temperature rises in which there lurk probably several tipping points. So we might not eliminate all possibility of tipping points, but at least we cut out a chunk of them.” 

Predicting tipping points is what Lenton has focused on most intensively recently, and he feels that this new frontier is not only compelling but offers great promise. It just needs some broader support. “I honestly think we can do something useful here to help societies on this planet and help reduce the risk that these events pose,” Lenton says. “For that to happen, it needs to get some social-political momentum behind it, for people to believe it is possible and to want to see us try.” 



Richard Blaustein is a science and environmental journalist based in Washington, D.C. His Twitter is @richblaustein.

[Photo courtesy of Ismael Alonzo]

Related posts