We know that our planet has experienced warmer periods in the past, during the Pliocene geological epoch around three million years ago.
Our research, published today, shows that up to one third of Antarctica’s ice sheet melted during this period, causing sea levels to rise by as much as 20 metres above present levels in coming centuries.
We were able to measure past changes in sea level by drilling cores at a site in New Zealand, known as the Whanganui Basin, which contains shallow marine sediments of arguably the highest resolution in the world.
Using a new method we developed to predict the water level from the size of sand particle moved by waves, we constructed a record of global sea-level change with significantly more precision than previously possible.
The Pliocene was the last time atmospheric carbon dioxide concentrations were above 400 parts per million and Earth’s temperature was 2°C warmer than pre-industrial times. We show that warming of more than 2°C could set off widespread melting in Antarctica once again and our planet could be hurtling back to the future, towards a climate that existed three million years ago.
Last week we saw unprecedented global protests under the banner of Greta Thunberg’s #FridaysForFuture climate strikes, as the urgency of keeping global warming below the Paris Agreement target of 2°C hit home. Thunberg captured collective frustration when she chastised the United Nations for not acting earlier on the scientific evidence. Her plea resonated as she reminded us that:
With today’s emissions levels, that remaining CO₂ budget [1.5°C] will be entirely gone in less than eight and a half years.
At the current rate of global emissions we may be back in the Pliocene by 2030 and we will have exceeded the 2°C Paris target. One of the most critical questions facing humanity is how much and how fast global sea levels will rise.
According to the recent special report on the world’s oceans and cryosphere by the Intergovernmental Panel on Climate Change (IPCC), glaciers and polar ice sheets continue to lose mass at an accelerating rate, but the contribution of polar ice sheets, in particular the Antarctic ice sheet, to future sea level rise remains difficult to constrain.
If we continue to follow our current emissions trajectory, the median (66% probability) global sea level reached by the end of the century will be 1.2 metres higher than now, with two metres a plausible upper limit (5% probability). But of course climate change doesn’t magically stop after the year 2100.
To better predict what we are committing the world’s future coastlines to we need to understand polar ice sheet sensitivity. If we want to know how much the oceans will rise at 400ppm CO₂, the Pliocene epoch is a good comparison.
Back in 2015, we drilled cores of sediment deposited during the Pliocene, preserved beneath the rugged hill country at the Whanganui Basin. One of us (Timothy Naish) has worked in this area for almost 30 years and identified more than 50 fluctuations in global sea level during the last 3.5 million years of Earth’s history. Global sea levels had gone up and down in response to natural climate cycles, known as Milankovitch cycles, which are caused by long-term changes in Earths solar orbit every 20,000, 40,000 and 100,000 years. These changes in turn cause polar ice sheets to grow or melt.
While sea levels were thought to have fluctuated by several tens of metres, up until now efforts to reconstruct the precise amplitude had been thwarted by difficulties due to Earth deformation processes and the incomplete nature of many of the cycles.
Our research used a well-established theoretical relationship between the size of the particles transported by waves on the continental shelf and the depth to the seabed. We then applied this method to 800 metres of drill core and outcrop, representing continuous sediment sequences that span a time period from 2.5 to 3.3 million years ago.
We show that during the Pliocene, global sea levels regularly fluctuated between five to 25 metres. We accounted for local tectonic land movements and regional sea-level changes caused by gravitational and crustal changes to determine the sea-level estimates, known as the PlioSeaNZ sea-level record. This provides an approximation of changes in global mean sea level.
Our study also shows that most of the sea-level rise during the Pliocene came from Antarctica’s ice sheets. During the warm Pliocene, the geography of Earth’s continents and oceans and the size of polar ice sheets were similar to today, with only a small ice sheet on Greenland during the warmest period. The melting of the Greenland ice sheet would have contributed at most five metres to the maximum 25 metres of global sea-level rise recorded at Whanganui Basin.
Of critical concern is that over 90% of the heat from global warming to date has gone into the ocean. Much of it has gone into the Southern Ocean, which bathes the margins of Antarctica’s ice sheet.
Already, we are observing warm circumpolar deep water upwelling and entering ice shelf cavities in several sites around Antarctica today. Along the Amundsen Sea coast of West Antarctica, where the ocean has been heating the most, the ice sheet is thinning and retreating the fastest. One third of Antarctica’s ice sheet — the equivalent to up to 20 metres of sea-level rise — is grounded below sea level and vulnerable to widespread collapse from ocean heating.
Our study has important implications for the stability and sensitivity of the Antarctic ice sheet and its potential to contribute to future sea levels. It supports the concept that a tipping point in the Antarctic ice sheet may be crossed if global temperatures are allowed to rise by more than 2℃. This could result in large parts of the ice sheet being committed to melt-down over the coming centuries, reshaping shorelines around the world.
Scientists have long been concerned about the potential collapse of the West Antarctic Ice Sheet and its contribution to global sea level rise. Much of West Antarctica’s ice lies below sea level, and warming ocean temperatures may lead to runaway ice sheet retreat.
This process, called marine ice sheet instability, has already been observed along parts of the Amundsen Sea region, where warming of the ocean has led to melting underneath the floating ice shelves that fringe the continent. As these ice shelves thin, the ice grounded on land flows more rapidly into the ocean and raises the sea level.
Although the Amundsen Sea region has shown the most rapid changes to date, more ice actually drains from West Antarctica via the Ross Ice Shelf than any other area. How this ice sheet responds to climate change in the Ross Sea region is therefore a key factor in Antarctica’s contribution to global sea level rise in the future.
Periods of past ice sheet retreat can give us insights into how sensitive the Ross Sea region is to changes in ocean and air temperatures. Our research, published today, argues that ocean warming was a key driver of glacial retreat since the last ice age in the Ross Sea. This suggests that the Ross Ice Shelf is highly sensitive to changes in the ocean.
Since the last ice age, the ice sheet retreated more than 1,000km in the Ross Sea region – more than any other region on the continent. But there is little consensus among the scientific community about how much climate and the ocean have contributed to this retreat.
Much of what we know about the past ice sheet retreat in the Ross Sea comes from rock samples found in the Transantarctic Mountains. Dating techniques allow scientists to determine when these rocks were exposed to the surface as the ice around them retreated. These rock samples, which were collected far from where the initial ice retreat took place, have generally led to interpretations in which the ice sheet retreat happened much later than, and independently of, the rise in air and ocean temperatures following the last ice age.
But radiocarbon ages from sediments in the Ross Sea suggest an earlier retreat, more in line with when climate began to warm from the last ice age.
To investigate how sensitive this region was to past changes, we developed a regional model of the Antarctic ice sheet. The model works by simulating the physics of the ice sheet and its response to changes in ocean and air temperatures. The simulations are then compared to geological records to check accuracy.
Our main findings are that warming of the ocean and atmosphere were the main causes of the major glacial retreat that took place in the Ross Sea region since the last ice age. But the dominance of these two controls in influencing the ice sheet evolved through time. Although air temperatures influenced the timing of the initial ice sheet retreat, ocean warming became the main driver due to melting of the Ross Ice Shelf from below, similar to what is currently observed in the Amundsen Sea.
The model also identifies key areas of uncertainty of past ice sheet behaviour. Obtaining sediment and rock samples and oceanographic data would help to improve modelling capabilities. The Siple Coast region of the Ross Ice Shelf is especially sensitive to changes in melt rates at the base of the ice shelf, and is therefore a critical region to sample.
Understanding processes that were important in the past allows us to improve and validate our model, which in turn gives us confidence in our future projections. Through its history, the ice sheet in the Ross Sea has been sensitive to changes in ocean and air temperatures. Currently, ocean warming underneath the Ross Ice Shelf is the main concern, given its potential to cause melting from below.
Challenges remain in determining exactly how ocean temperatures will change underneath the Ross Ice Shelf in the coming decades. This will depend on changes to patterns of ocean circulation, with complex interactions and feedback between sea ice, surface winds and melt water from the ice sheet.
Given the sensitivity of ice shelves to ocean warming, we need an integrated modelling approach that can accurately reproduce both the ocean circulation and dynamics of the ice sheet. But the computational cost is high.
Ultimately, these integrated projections of the Southern Ocean and Antarctic ice sheet will help policymakers and communities to develop meaningful adaptation strategies for cities and coastal infrastructure exposed to the risk of rising seas.
The ocean that surrounds Antarctica plays a crucial role in regulating the mass balance of the continent’s ice cover. We now know that the thinning of ice that affects nearly a quarter of the West Antarctic Ice Sheet is clearly linked to the ocean.
The connection between the Southern Ocean and Antarctica’s ice sheet lies in ice shelves – massive slabs of glacial ice, many hundreds of metres thick, that float on the ocean. Ice shelves grind against coastlines and islands and buttress the outflow of grounded ice. When the ocean erodes ice shelves from below, this buttressing action is reduced.
While some ice shelves are thinning rapidly, others remain stable, and the key to understanding these differences lies within the hidden oceans beneath ice shelves. Our recently published research explores the ocean processes that drive melting of the world’s largest ice shelf. It shows that a frequently overlooked process is driving rapid melting of a key part of the shelf.
Rapid ice loss from Antarctica is frequently linked to Circumpolar Deep Water (CDW). This relatively warm (+1C) and salty water mass, which is found at depths below 300 metres around Antarctica, can drive rapid melting. For example, in the south-east Pacific, along West Antarctica’s Amundsen Sea coast, CDW crosses the continental shelf in deep channels and enters ice shelf cavities, driving rapid melting and thinning.
Interestingly, not all ice shelves are melting quickly. The largest ice shelves, including the vast Ross and Filchner-Ronne ice shelves, appear close to equilibrium. They are largely isolated from CDW by the cold waters that surround them.
The contrasting effects of CDW and cold shelf waters, combined with their distribution, explain much of the variability in the melting we observe around Antarctica today. But despite ongoing efforts to probe the ice shelf cavities, these hidden seas remain among the least explored parts of Earth’s oceans.
It is within this context that our research explores a new and hard-won dataset of oceanographic observations and melt rates from the world’s largest ice shelf.
In 2011, we used a 260 metre deep borehole that had been melted through the north-western corner of the Ross Ice Shelf, seven kilometres from the open ocean, to deploy instruments that monitor ocean conditions and melt rates beneath the ice. The instruments remained in place for four years.
The observations showed that far from being a quiet back water, conditions beneath the ice shelf are constantly changing. Water temperature, salinity and currents follow a strong seasonal cycle, which suggests that warm surface water from north of the ice front is drawn southward into the cavity during summer.
Melt rates at the mooring site average 1.8 metres per year. While this rate is much lower than ice shelves impacted by warm CDW, it is ten times higher than the average rate for the Ross Ice Shelf. Strong seasonal variability in the melt rate suggests that this melting hotspot is linked to the summer inflow.
To assess the scale of this effect, we used a high-precision radar to map basal melt rates across a region of about 8,000 square kilometres around the mooring site. Careful observations at around 80 sites allowed us to measure the vertical movement of the ice base and internal layers within the ice shelf over a one-year interval. We could then determine how much of the thinning was caused by basal melting.
Melting was fastest near the ice front where we observed short-term melt rates of up to 15 centimetres per day – several orders of magnitude higher than the ice shelf average rate. Melt rates reduced with distance from the ice front, but rapid melting extended far beyond the mooring site. Melting from the survey region accounted for some 20% of the total from the entire ice shelf.
Why is this region of the shelf melting so much more quickly than elsewhere? As is so often the case in the ocean, it appears that winds play a key role.
During winter and spring, strong katabatic winds sweep across the western Ross Ice Shelf and drive sea ice from the coast. This leads to the formation of an area that is free of sea ice, a polynya, where the ocean is exposed to the atmosphere. During winter, this area of open ocean cools rapidly and sea ice grows. But during spring and summer, the dark ocean surface absorbs heat from the sun and warms, forming a warm surface pool with enough heat to drive the observed melting.
Although the melt rates we observe are far lower than those seen on ice shelves influenced by CDW, the observations suggest that for the Ross Ice Shelf, surface heat is important.
Given this heat is closely linked to surface climate, it is likely that the predicted reductions in sea ice within the coming century will increase basal melt rates. While the rapid melting we observed is currently balanced by ice inflow, glacier models show that this is a structurally critical region where the ice shelf is pinned against Ross Island. Any increase in melt rates could reduce buttressing from Ross Island, increasing the discharge of land-based ice, and ultimately add to sea levels.
While there is still much to learn about these processes, and further surprises are certain, one thing is clear. The ocean plays a key role in the dynamics of Antarctica’s ice sheet and to understand the stability of the ice sheet we must look to the ocean.
Craig Stewart, Marine Physicist, National Institute of Water and Atmospheric Research
A new report has warned that even if global warming is held at 1.5℃, we will still lose a third of the glaciers in the Hindu Kush-Himalaya (HKH) region. What does that mean for rivers that flow down these mountains, and the people who depend on them?
The HKH region is home to the tallest mountains on Earth, and also to the source of rivers that sustain close to 2 billion people. These rivers supply agriculture with water and with sediments that fertilise soils in valleys and the floodplain.
Some of these rivers are hugely culturally significant. The Ganges (or Ganga), for instance, which flows for more than 2,525km from the western Himalayas into the Bay of Bengal, is personified in Hinduism as the goddess Gaṅgā.
Before we get to the effect of melting glaciers on Himalayan rivers, we need to understand where they get their water.
For much of Himalayas, rain falls mostly during the monsoon active between June and September. The monsoon brings heavy rain and often causes devastating floods, such as in northern India in 2013, which forced the evacuation of more than 110,000 people.
But the summer monsoon is not the only culprit for devastating floods. Landslides can dam the river, and when this dam bursts it can cause dramatic, unpredictable flooding. Some of those events have been linked to folk stories of floods in many cultures around the world. In the Himalayas, a study tracking the 1,000-year history of large floods showed that heavy rainfall and landslide-dam burst are the main causes.
When they melt, glaciers can also create natural dams, which can then burst and send floods down the valley. In this way, the newly forecast melting poses an acute threat.
The potential problem is worsened still further by the Intergovernmental Panel on Climate Change’s prediction that the frequency of extreme rainfall events will also increase.
What will happen to Himalayan rivers when the taps are turned to high in this way? To answer this, we need to look into the past.
For tens of thousands of years, rivers have polished rocks and laid down sediments in the lower valleys of the mountain range. These sediments and rocks tell us the story of how the river behaves when the tap opens or closes.
Some experts propose that intense rain tends to trigger landslides, choking the river with sediments which are then dumped in the valleys. Others suggest that the supply of sediments to the river generally doesn’t change much even in extreme rainfall events, and that the main effect of the extra flow is that the river erodes further into its bed.
The most recent work supports the latter theory. It found that 25,000-35,000 years ago, when the monsoon was much weaker than today, sediments were filling up Himalayan valleys. But more recently (3,000-6,000 years ago), rock surfaces were exposed during a period of strong monsoon, illustrating how the river carved into its bed in response to higher rainfall.
So what does the past tell us about the future of Himalayan rivers? More frequent extreme rainfall events mean more floods, of course. But a stronger monsoon also means rivers will cut deeper into their beds, instead of fertilising Himalayan valleys and the Indo-Gangetic plain with sediments.
What about glaciers melting? For as long as there are glaciers, this will increase the amount of meltwater in the rivers each spring (until 2060, according the report, after which there won’t be any meltwater to talk about). So this too will contribute to rivers carving into their beds instead of distributing sediments. It will also increase the risk of flooding from outburst of glacial lake dams.
So what is at stake? The melting glaciers? No. Given thousands or millions of years, it seems likely that they will one day return. But on a more meaningful human timescale, what is really at stake is us – our own survival. Global warming is reducing our resources, and making life more perilous along the way. The rivers of the Himalayas are just one more example.
The link below is to an article reporting on the rapid ice melt in Greenland.