The dangers of a collapse of the main Atlantic Ocean circulation, known as Amoc, have been “greatly underestimated” and would have devastating and irreversible impacts, according to an open letter released at the weekend by 44 experts from 15 countries. One of the signatories, Stefan Rahmstorf, an oceanographer and climatologist who heads the Earth system analysis department at the Potsdam Institute for Climate Impact Research in Germany, explains here why he has recently upgraded his risk assessment of an Amoc breakdown as a result of global heating – and what that means for Britain, Europe and the wider world.
What is Amoc?
Amoc, or the Atlantic meridional overturning circulation, is a system of ocean currents that brings heat into the northern Atlantic. Warm surface water from the tropics flows north and releases its heat in the subpolar Atlantic, south of Greenland and west of Britain and Ireland. Then it cools and sinks to a depth of between 2,000m to 3,000 metres before returning south as a cold current. Amoc is one of our planet’s largest heat transport systems, moving the equivalent of 50 times the human energy use, and it has a particularly strong impact on the climate in Europe, affects the ocean’s CO2 uptake and oxygen supply, as well as rainfall patterns in the tropics.
How is Amoc different to the Gulf Stream?
They are connected because the northwards flow of Amoc goes via the Gulf Stream, which is a warm and swift Atlantic Ocean current that originates in the Gulf of Mexico, then flows through the Florida straits, up the coast of the US and then across towards Europe. Amoc contributes just 20% to the Gulf Stream water flow but most of the heat transport, since Amoc’s deep return flow is very cold. It works like a central heating system.
What is happening to Amoc?
There are indications that Amoc has been slowing down for the last 60 or 70 years due to global heating. The most ominous sign is the cold blob over the northern Atlantic. The region is the only place in the world that has cooled in the past 20 years or so, while everywhere else on the planet has warmed – a sign of reduced heat transport into that region, exactly what climate computer models have predicted in response to Amoc slowing as a result of greenhouse gas emissions.
Are there other indications that Amoc is weakening?
Yes. There is a region of excessive heating along the east coast of North America, which is predicted by climate models and oceanographic theory as a result of a slowing Amoc, which pushes the Gulf Stream closer to the shore.
Another indicator is a reduction in the salt content of seawater. In the cold blob region, salinity is at its lowest level since measurements began 120 years ago. This is probably linked to Amoc slowing down and bringing less salty water and heat from the subtropics.
Why is the salt content significant?
When the water is less salty, it is less dense, which makes it harder to sink down. That is important because the sinking process is what drives Amoc. The fresher the water, the slower it gets.
What is driving the change in salinity?
Firstly, salinity is directly affected by global heating, which enhances the water cycle so there is more evaporation in the subtropics and more precipitation in the subpolar oceans. This leads to a freshening of the subpolar ocean. Then there are additional contributions from the melting of sea ice and the loss of continental ice from the Greenland ice sheet, which is freshwater that flows into the ocean.
It is an amplifying feedback: as Amoc gets weaker, the subpolar oceans gets less salty, and as the oceans gets less salty then Amoc gets weaker. At a certain point this becomes a vicious circle which continues by itself until Amoc has died, even if we stop pushing the system with further emissions.
When might Amoc weakening reach a point of no return?
The big unknown here – the billion-dollar question – is how far away this tipping point is. It is very difficult to answer because the process is non-linear and would be triggered by subtle differences in salinity, which in turn depend on amounts of rainfall and cloud cover over the ocean as well as Greenland melting rates. These are hard to model accurately in computers so there is a big uncertainty relating to when the tipping point will be reached.
What is the range of forecasts?
Until a few years ago, the general thinking in the Intergovernmental Panel on Climate Change (IPCC) was that the probability of crossing the tipping point this century was less than 10%. Since then, there have been a number of studies suggesting a collapse would probably be triggered this century, possibly in the next few decades. So my risk assessment has really changed. I am now very concerned that we may push Amoc over this tipping point in the next decades. If you ask me my gut feeling, I would say the risk that we cross the tipping point this century is about 50/50.
Is there any possibility it has already happened?
I wouldn’t rule it out completely, because it would be very hard to tell from observations. Nothing dramatic happens at the tipping point. That just means Amoc is then doomed and it will slowly die, but that process could take 50 to 100 years. Because the Amoc is already weakening we can’t be entirely sure whether we already passed a tipping point, but I would say this is most likely not the case, so it is not too late to prevent this.
What would be the warning signs of Amoc collapse?
We need to keep monitoring the flow of water in the Atlantic, which is being done with the Rapid project. We should also monitor deep winter mixing in the northern Atlantic and Nordic seas. If the deep mixing starts to decline a lot, that could be an early indicator that we are approaching a tipping point. There are some signs of this, but we don’t have enough data yet to be sure.
What would be the consequences of Amoc breakdown?
This has happened repeatedly in Earth’s history, most recently during the last ice age, when big ice masses slid into the ocean – so-called Heinrich events – adding meltwater that diluted the salinity of the north Atlantic. These are among the most massive upheavals of climate conditions in Earth’s history.
The effects include a cooling of the northern hemisphere, particularly northwestern Europe. There would also be a shift of the tropical rainfall belt to the south, which is bad because the rains will move away from the rainforests to regions that are not used to so much rainfall. So this will mean droughts in some regions and floods in others.
Amoc collapse would also have a major impact on the northern Atlantic sea level, which would rise by half a metre or so, in addition to the rise caused by global heating. It would also reduce the CO2 uptake of the ocean because Amoc sinking in the northern Atlantic takes a lot of CO2 down into the deep oceans where it is safely locked away from the atmosphere.
Amoc collapse would also change the nutrient supply and reduce the oxygen of the deep oceans. This would have a massive effect on marine biology and the entire ecosystem of the northern Atlantic.
Many of these things are happening already, aren’t they?
Yes, to some extent. This is partly because Amoc is weakening and so is its counterpart in the southern hemisphere, the Antarctic bottom water formation, according to research by Australian colleagues.
Could the cooling effect of Amoc collapse offset the heating caused by human emissions?
I can’t think of anywhere that will be better off. If it were just a case of averages, then somewhere like Germany might see a balance. But weather is not a climate average; it is seasonal and highly variable. Within the average you can get warm air from the south or cold polar air outbreaks from the north. These contrasts will be more pronounced if Scandinavia and Britain cool while Spain and Italy warm. This will drive much greater variability in the weather, which is bad for agriculture, and it will cause more storms. I would expect major extreme weather events that we have not seen in the past.
The key thing about climate change is that both the ecosystem of the Earth as well as human settlements and infrastructures are highly adapted to what the climate was like in previous centuries. So any change – whether global heating or global cooling – will always be bad because it will lead to maladaptation. Think of the tremendous flooding we have seen somewhere in the world almost every week in the last months. If it had been like that for centuries, then river and sewer systems would be adapted to take up that water. But because we are not used to that, there are disasters. That is the problem of climate change.
How certain is the science about Amoc collapse?
It is well established that Amoc is weakening and that a tipping point exists. The uncertainty is about when we will cross that threshold. We also have very few studies about what the combined effect of Amoc collapse and global heating would exactly look like.
It is a question of risk assessment. I compare it to being told that there is a 10% chance of an airplane crashing. Would you get on that plane? I wouldn’t. The disastrous consequences are unacceptable.
Why haven’t the IPCC made more of Amoc risks?
They have not done enough risk assessment because they tend to focus on the most probable scenarios for future climate change. Some colleagues say we shouldn’t talk about extreme possibilities like an Amoc collapse because it sounds alarmist and might distract people from more certain impacts of global heating, which are bad enough. But I think those extreme risks are part of the whole picture that we need to consider, to make responsible and rational decisions.
How long would an Amoc collapse last and how survivable would it be?
The last time, it took about 1,000 years to recover, though the past is not a direct analogue because there is also massive CO2 forcing this time – CO2 is already higher than any time in 15m years. There are physical reasons why some form of deep overturning circulation will eventually come back.
One thing is for sure: humanity will not die out. But for some countries that will be in the midst of this, like Norway, and Scotland, the risks will be existential and raise the question whether people can continue to live there or whether most of them would rather move.
How does the Amoc threat compare to other climate tipping points?
That is hard to tell. It is a trade-off between more distant futures and things that are already happening.
We have already crossed the tipping point of many coral reefs, which are now in middle of global die-off. This is very depressing because it is already too late to do anything about it, though marine biologists have warned about the risks for a long time. The Amazon rainforest is also dangerously close to a tipping point. As we speak, it is going through the worst drought on record.
Then in the very long run, we have the ice sheet tipping points in Greenland and west Antarctica. From Greenland alone, this will lead to a seven-metre global sea level rise that will wipe all major coastal cities off the map. But that will occur over many centuries because ice sheet melt is a slow process.
Amoc is on an intermediate timescale because it unfolds over decades to 100 years.
I am worried about all of these things to be honest. And the conclusion for all of them is the same: this is all driven mainly by fossil fuel emissions and also deforestation, so both must be stopped. We must stick to the Paris agreement and limit global heating as close to 1.5C as possible. I don’t think it is my job to talk about my feelings, but I do have two children and I am very worried about what future they will live in. I sometimes joke that physicists don’t have feelings. But even physicists care about their kids.
