16,000 years ago, Iceland was in the grip of ice. The sea was 100 meters lower than it is now, and the land extended far beyond the current coast. But too little benefit. The island was covered in an icecap which extended well beyond the current coast. At its peak (21.5 thousands years ago) the ice may have been 1500 meters thick (some say 2000 meters). In the west, the peak extent was as far as 100 km from the current coast. The edge was around the location where the shelf around Iceland reaches 200 meter depth. The outer parts of the ice were floating on the freezing-cold Atlantic ocean.
2500 years later, Iceland was unrecognizable. The ice had withdrawn from the coast and covered only the central regions. Sea levels had risen but were still below modern levels. Volcanoes erupted on the newly exposed coastal regions, perhaps surprised by the rapidity of the change. This was the time when the big glaciers over Canada and Europe were melting, and the sea had rising by some 50 meters over 1500 years. The sea also had warmed. The floating ice shelf very quickly disintegrated and the ice had withdrawn to the central regions of Iceland. Tundra vegetation was becoming established. It was a new land.
The warmer period did not last. It was followed by the cold Younger Dryas, which suddenly plunged Iceland back into the ice age. Between 13,000 and 12,000 years ago, the ice extended back to the coast, although it remained well within the earlier limits. After 12,000 BP, the glaciers again retreated, but only by about 20% over the next 800 years. After that, the retreat speeded up and over some 2500 years the glaciers disintegrated.
By now the glaciers were limited to the higher mountains (invariably volcanic) only. The ice did not disappear entirely, as indicated by the fact that Katla’s eruptions during this time produced subglacial tephra. But again the good times came to an end, and by 5000 BP the glaciers began to grow again. There was a growth spurt between 2500 and 2000 years ago, and then more extensive growth during the little ice age, from 1300 to 1800 AD. But that too did not last. And now, in our age of modernity we are again in a phase of rapid melting; the predictions are that Iceland’s glaciers will almost entirely disappear in the next 200 years, something that didn’t happen even during the glacial minimum 5000 years ago. As a sign of the times, the Okjökull glacier died in 2014. Perhaps the country will soon need to find a new name.
The history of Iceland shows that major changes can happen fast. The first major retreat of the ice was around 21 thousand years ago. At this tine, over a period of only 750 years, Iceland’s ice age glaciers contributed 50 cm to sea level rise. 15 thousand years ago there was a second phase where the glaciers halved within a 1000 years. And a thousand years (one millennium) is not long: it is less than the period during which Iceland has been inhabited. There have been changes within the past millennium too. The Vikings came when the ice was still subdued: some areas they settled later became uninhabitable as the ice advanced. Some farms were overrun. But these changes were not as rapid as has happened in the past. This is the cycle of the ice: it grows slow but melts fast. Advancing glaciers give you time. Retreating glaciers change the world almost instantly.
And the world today is not like yesterday. The climate is changing faster than it has for many millennia. What happened in Iceland 15,000 years ago, is happening in Greenland today. If Greenland would do an Iceland, and see a complete and rapid collapse of its glacier, sea level would rise by 7 meters. Only 5% of the world’s population lives within 5 meters of sea level, but this still means that half a billion people would be submerged. All harbours would be lost, obviously. The USA would lose 3% of its farmland; Vietnam would lose 23%. It is not only the wealthy ocean-view villas that are at risk from Greenland.
Global warming and sea level rise are happening fast. The world of 2020 is already notably different from that of 2010; even the change is changing. But is it a crisis? Or is it relevant for a distant future but are the reports of our impending demise overblown? What does science say about the future?
Recently, people have pointed out potential climate tipping points: points of no return beyond which changes become rapid and irreversible – at least on our time scales. As the U.N. Secretary-General António Guterres said last week, “the point of no return is no longer over the horizon. It is in sight and hurtling toward us.” Are those tipping points real? Was the sudden collapse of Iceland’s glaciers such a tipping point, induced by the warming seas? And have we already tipped the balance? This post will discuss some of those proposed tipping points. Don’t expect answers: this is about the questions.
Seven climate tipping points
Ice sheets are inherently unstable. The top tends to be too cold to melt, sticking up as high as the proverbial mountains. But ice can flow, and just like a mountain made out of syrup, glaciers spread out over time and become lower. New snow is needed every year to keep the ice cap as high as it was. If the ice cap begins to lower, either from melt or from lack of snow, the temperature at the top begins to increase. For every 100 meter lowering, temperatures go up by 1 C. Once the top of the ice cap begins to melt, the process quickly accelerates, and becomes unstoppable.
Ice sheets have two more vulnerabilities. The bottom of an ice cap is at zero degrees C, and water can collect there. This lubricates the flow down the hill. If more melt water reaches down there, it can speed up the flow down-hill. If the outer edge forms a floating ice shelf, it is liable to melting by the sea water underneath. A warming sea attacks the shelf from below. An ice shelf can disintegrate within just a few years. And because that removes a lot of weight, it allows the glacier behind, which formed the shelf, to speed up.
So glaciers can disappear fast. In contrast, forming a glacier takes a long time, because it cannot grow faster than the amount of snow that falls per year. (A point in case is modern Iceland: the glaciers are mainly along the south coast even though this is the warmer region: this is where the precipitation is highest. The colder northern part is much drier.)
So there is a strong case that there is a tipping point for glaciers. The fact that Iceland’s glacier collapsed so quickly at the end of the ice age, and did not fully reform during the Younger Dryas, stems from that. But what is the tipping point, and what would happen when it is reached?
For Greenland’s ice, the tipping point is around 1.5 C of warming. That much warming is already unavoidable, and there is a strong case that Greenland’s glacier is doomed. But this doom takes time. At 1.5 C, it would still take 10,000 years for the process to unfolds. At 2C, it might take 1000 years. Greenland’s ice will add 7 meters to our sea level, but it will not happen instantly.
There are on-going discussions about the fate of Antarctica’s ice. The West Antarctic ice sheet is largely grounded below sea level, because its weight has depressed the land by that much. That makes it vulnerable. Some models indicate that it has already passed its tipping point, caused by the rising and warming seas. That will add another 3 meters to sea level, but again not instantly.
There are probably three separate tipping points, one for Greenland, once for West Antarctica and one for the much more robust ice sheet of East Antarctica. The latter seems safe for now. Greenland is doomed, but looking at the past suggests that Greenland will need time to melt. The main ice sheet is vulnerable only from above. West Antarctica is a different matter. Long ago, Iceland’s ice collapsed in part due to sea level rise. Between 14.7 and 14.3 BP, a 400-year period, global sea level rose by 40 mm per year, or 4 meters per century. This lifted the near-shore ice and disintegrated it, within only a few hundred years. Current sea level rise is around 5 mm per year, having doubled since 1990. This is still much less than what attacked Iceland – but West Antarctica is both larger and more vulnerable.
So over a millennium, the coast is toast. But if we do push West Antarctica over the edge, a 3-meter flood could come faster than expected, although it is still unlikely to happen this century. This tipping point is a danger – the question is when, not if.
During the 17th century, sea ice was a common occurrence around Iceland. The northern coast could sometimes be ice-bound until June, and at times sea ice reached the southern coast. Nowadays the sea ice normally stays well away to the north. This is due to rapid warming of the sea in the region. The warm Atlantic current has pushed the cold waters further north, and the sea is now several degrees warmer than it used to be.
This local effect can’t be extrapolated to the rest of the Arctic. But sea ice is fragile. It survives the summer because it is so thick. Once it melts from below, the thickness is no longer sufficient and it can very suddenly and unexpectedly break up. Models indicate that the shift to an ice-free Arctic summer happens around 2 C of warming. Once this happens, the water warms further because sunlight gets absorbed in the water rather than reflected by the ice. This tipping point is probably unavoidable already. A commercial pilot a few years ago showed me pictures of the North Pole, taken on his New York – Hong Kong route. He was shocked by how broken the ice was. This was in May, when just a few years earlier this would have been solidly frozen. Now there is open water at the pole.
Does it matter? There will be ecological changes where some animals will suffer and some will do well. The main effect may be local warming, with much warmer summers around the Arctic. That sounds like a good thing. But it will affect the surrounding land, and will for instance accelerate ice melt in Greenland. And there is the disappearance of the permafrost in Alaska, Canada and Siberia.
This brings us to the next tipping point. When will the permafrost go? Across the north, permafrost is retreating, down into the earth and north towards the pole. The melt leaves a damaged landscape, with features similar to sink holes. This is a fragile land at the best of times; removing its foundations while soaking the upper layers is a recipe for subsidence. (One group has warned of the risk of defrosted ancient bacteria causing epidemics. That is silly. It is unlikely to the extreme and there are far better sources for new diseases available.)
But would the permafrost count as a tipping point? It is not irreversible – lower the temperature and the land will refreeze. Also, once melted the land will become vegetated, and over time the damaged landscape will fix itself. However, the risk is in what lays beneath. The frozen tundra has locked in CO2 and methane, and the melt releases these relics. CO2 emissions from the permafrost are probably dwarfed by what we put out ourselves. But methane is a different matter. It is a powerful, albeit short-lived greenhouse gas. Worldwide emissions are rising and we are not fully sure why. Melting permafrost may be a contributor. The methane concentration has more than doubled since pre-industrial days and is expected to double again by 2100. Still, methane accounts for ‘only’ 20% of current warming. Methane lasts for about a decade in the atmosphere. Over that period, it enhances all other effects of global warming – including the melting of the permafrost. So there is some ‘tipping point’ where the process become self-sustaining. However, there isn’t a clear sudden onset, the changes are mainly along the outer areas of the permafrost regions, and CO2 contributes much more to the warming. The effect is real and significant, and needs to be included in climate models, but as a ‘tipping point’ it seems overstated.
Three further aspects related to the living world have been mentioned as tipping points. The first is the boreal forest, the band of trees circling the world which is better known as the taiga. It covers 10% of the world surface and contains half of all remaining wilderness area. It is at risk from two factors, insect infestations killing the trees and fires killing the forests, making use of the newly deaded wood. Recent summers saw tremendous fires in the Siberian taiga (often caused by people). The fires put CO2 into the atmosphere. To put that in context: the taiga contains twice as much carbon as the world’s tropical forests, much of it in the soil underneath its feet. (This is in contrast to the tropical forests where the soil is carbon poor and almost all the carbon is in the trees.) It has been called the carbon that the world forgot. The permafrost and the taiga are closely linked and overlap, but the taiga covers a much larger area than the tundra permafrost.
In the past, the boreal forest acted as a carbon sink: it took up more carbon than it put into the atmosphere. But as temperatures are rising, the boreal forests are changing and are becoming a net carbon source. The ‘tipping point’ is where it becomes a net source of carbon, and it may happen as early as the 2020’s. But it is a gradual change, and the precise time at which this point is reached is less important. The importance lies in the size of the carbon reservoir that it acts on.
The second one is the Amazon rain forest. This forest is often considered doomed. The forest is dependent on the seasonal rains, and these are being replaced by droughts. It is plausible that much of the Amazon will cease to be a rain forest. From the point of the biosphere this is a crucial change, as so many of the world’s life forms are here and much will not survive. But the models do not predict that the entire rain forest would be lost. The core region with its two rainy seasons per year seems safe. But especially the southern areas, with one rainy season, are at risk. There is a back story here. The risks to these areas were already obtained from IPCC models 12 years ago. At the time it was ridiculed by the skeptics as misleading – the Sunday Times at the front of the queue. However, these were indeed the regions affected by major fires this year, caused by drought – just as had been predicted. I must have missed the apology from the Sunday Times.
The forest is to some degree self-sustaining: the frequent rain showers are fed by water evaporated by the trees themselves. Remove the trees, and the rain will only fall once and rivers will quickly carry the water off to the Atlantic ocean. There is indeed a tipping point here: remove enough trees, and the lessening of the rains will kill off the rest. The tipping point appears to fall around 30-40% of deforestation; we are half-way there. It will impact the world’s of life’s diversity. Beetle-dom will never be the same. The link to global climate is clear for the southern regions. For the core of the Amazon basin it is not clear-cut: the tipping point appears to be mostly about the deforestation rather than temperatures. This is not unlimited: too much heat will kill off any forest. But less extreme warming may have more complex effects.
It is worth pointing out that the tropical rain forests are comparatively recent. They formed after the ice age. Their fragility may therefore not be such a surprise.
The third case is that of the coral reefs, the rain forests of the oceans. Here the effect is crystal clear: coral bleaching is already killing much of the coral. Ocean heat waves kill the symbiotic algae on which the corals depend. It leaves the coral ghostly white, and starving as the link to 80% of its food supply is lost. 29% of the Great Barrier Reef has been lost to bleaching. (Sometimes it seems that the Australian government, not quite the most green in the world, would prefer the whole thing to be gone). If the bleachings occur more frequently than new coral can grow, the reef does not recover. Before 2010, the Great Barrier Reef was in balance, with regrowth keeping up with the occasional bleaching events. But this is the case no longer, and we can say that for Australia, this tipping point was reached in 2010.
This tipping point may not be a permanent one. There are coral reefs in hot waters in the Persian Gulf, and so reefs in principle could re-establish themselves from these species. This may require our helping hand. But ocean heat waves are not the only risk factor to the coral reefs. Ocean acidification makes reefs chemically unstable, and their calcium skeletons will dissolve into the acidic water. This tipping point may be reached in 50 years. Acidification may be the most serious threat and it seems hard to avoid. The bottom line is that if you still need to see a coral reef (it really is a wonderful experience, if you get to see an undamaged one!), you’d better hurry.
We will give the IPCC the last word, in their 2019 report. They have high confidence that ‘critical thresholds for some ecosystems (e.g. kelp forests, coral reefs) will be reached at relatively low levels of future global warming‘.
This saviour of Iceland is also the dark horse of our warming world. In the Atlantic ocean, warm, salty waters from the tropics are carried north by currents such as the Gulf stream; in the north Atlantic these waters cool and sink, and flow along the ocean floor, back to the tropics. In the process it creates the biggest waterfall in the world in the Denmark Strait. The Denmark Strait cataract is 150 kilometers wide, plunges 3.5 kilometers down, and carries 5 million cubic meters of water per second. And no one will ever see it in action. The warmth released by the sinking waters is the sole reason why Iceland isn’t still under a kilometer of ice. The strength of this so-called Atlantic meridional overturning circulation (AMOC) has varied over time. It has slowed down by about 15% since the 1950’s. The cause of this is not known but a possibility is the increase of fresh water entering the northern oceans, coming from glacial melt. The fresh water forms a low-salt layer which floats on top, and slows the cooling and sinking of the warm currents which now flow underneath. At the same time this would also warm the Arctic oceans, and contribute to the collapse of the sea ice.
The possibility has been raised that the AMOC may collapse entirely under global warming , which would be a major escalation of climate change. The physics of the collapse is plausible: the mechanisms work, and the question is only whether they are strong enough to induce collapse. If it does happen, in theory, Iceland could plunge back into a Younger Dryas – in fact the dominant model for the cause of the Younger Dryas is exactly that. This is a different tipping point, as it would tip part of the world back into a much colder past. This experiment has never been tried under current CO2 levels. As a scientist I am highly curious as to which of the two would win, the oceanic cooling or the CO2 warming. (My gut feeling is that CO2 would largely compensate for the oceanic cooling, at least for Europe, but winters would be freezing and stormy) (which they aren’t now). It would safeguard Greenland, and it would slow sea level rise. But this comes at the cost of losing Iceland – and Western Antarctica would remain at high risk.
Although cessation of the Gulf stream is often described as the biggest risk of global warming, it is a slow process and we are likely to loose our coast line well before AMOC can help us. But by the 24th century it becomes more likely. The IPCC (2019) writes: ‘By 2300, an AMOC collapse is as likely as not for high emission pathway and very unlikely for lower ones […] Nevertheless, the human impact of these physical changes have not been sufficiently quantified and there are considerable knowledge gaps in adaptation responses to a substantial AMOC weakening’. For the 21st century, they consider a sudden collapse as very unlikely. This appears to be one tipping point that can be avoided, and if we let it happen we only have ourselves to blame.
Lessons from Iceland
Iceland has always been at the forefront of our battle with climate. It lies at the border of the polar and Atlantic climates, a border which has been pushed far north by the Atlantic ocean currents. Any small change either in those currents or in the world’s climate, puts Iceland solidly on one side or the other. During the ice age, it single-handedly maintained one of the world’s major glaciers, with the size of France. It lost much of this within only a thousand years – twice. From Iceland we have learned how quickly a glacier can disintegrate. The Vikings found a green and fertile land, but within a few centuries they had used up the thin post-ice-age soils, and also saw the ice and jokulhaups encroach. They saw sea ice arrive from the north and they lost the sea route to the Greenland colony to the ice. But in recent years the warmth has returned, the sea has warmed by several degrees and the newly expanded ice caps are now on the brink of collapse. Sadly, the soil remains fragile.
In Iceland we can see what could happen in our rapidly warming world. Some of the tipping points are real, and some are overstated although still representing a real change to our world. Further sea level rise seems inevitable and anyone living within 1 meter of sea level may want to be cautious about a long-term mortgage. Which path we follow is still up to us. At the moment we are on a track that will give in excess of 5 C warming by the end of the century, but that can still be limited. But not by in-action.
And as always, volcanoes have the final word. We have not had a significant eruption since before 1900. That cannot last. One day the skies will darken again and a mountain will be lost. Just wait. And waiting is what Iceland does best. After both glacial collapses, 22,000 and 15,000 years ago, a phase with large eruptions followed. How much of this was pent-up pressure released by the melting ice, and how much was decompression melt and came later, is up for discussion. But now the glaciers are melting and the experiment is repeated – again. Some time, Antarctica will re-erupt too. But first, Iceland will have its turn. Katla is ready – and waiting.
Albert, December 2019