Skip to content

When you choose to publish with PLOS, your research makes an impact. Make your work accessible to all, without restrictions, and accelerate scientific discovery with options like preprints and published peer review that make your work more Open.

PLOS BLOGS All Models Are Wrong

The things we take for granted

The cracks start without you noticing. You hadn’t been paying attention: you’d been looking at problems elsewhere, at more immediate emergencies. You hadn’t realised the dynamics had shifted at home. Things had been gradually eroding beneath the surface.

Sure, there’ve been good things too – month by month, and year by year, you’ve been making new memories, and building things up: at least, on the surface. Now you see that you’ve been taking it all for granted. That your foundations are no longer solid, but vulnerable and unstable.

You knew such changes were a risk of life, a risk of commitment: but you believed you were immune. Now the good things do not always offset the bad. The future is unclear.

What does it feel like to suddenly realise you could lose everything?

Simple outline of the Antarctic ice sheet, showing labels for the smaller, left-hand part (West) and much larger, right-hand part (East). An outline of the United Kingdom is next to the West part and would fit comfortably inside.
West and East Antarctic ice sheets, with Britain and Ireland shown for scale. Black lines show the edge of the ice resting on bedrock; blue lines show the edge of floating ice shelves. Adapted from

Since 2014, we’d been paying close attention to the potential irreversible decline of the West Antarctic ice sheet, after satellites and computer models began to agree there were signs of unstable collapse. We know this western part of Antarctica disappeared in past warm periods, regrowing with the ice ages. Its vulnerability — both then and now — comes from resting on underwater bedrock that becomes deeper and deeper inland from the coast (from pale to darker blue in the image below). Warm ocean currents tickle the edge of the ice, and as it retreats down the slope, it speeds up.

We scientists have made predictions about the risk of this instability that ranged from fairly modest ice loss over the next two hundred years, coming mostly from two major glaciers in West Antarctica (pink in the image below: more details in the original blog post), to apocalyptic scenarios in which large parts of the region’s ice crumble into the ocean before the century is out. Either way, we focused our efforts on understanding — and preventing — this part of the climate story.

Probability of retreat by 2200 projected by Ritz and others (2015), overlaid on bedrock elevation map by Le Brocq and others (2010). Brown and blue areas show where bedrock lies above or below sea level. Pink shows a 2/3 or greater chance of the ice retreating by 2200, and other colours show lower probabilities. Source: Catherine Ritz.

Meanwhile, we largely ignored the rest: East Antarctica. Far, far bigger, holding 90% of Antarctica’s 58 metre potential to raise sea levels, we’ve nevertheless taken its stability for granted. It is by far the coldest place on Earth — a deep, deep freeze caused by the weak sunlight of the South pole, reinforced by the vast, vast whiteness that reflects much of that sunlight away. The East is much less exposed to the oceans, too, because only about a third of the bedrock is underwater.

And we’d seen little change there in recent decades. If anything, East Antarctica had grown slightly, because snowfall on the surface was building up the ice sheet — year by year, month by month — outweighing any losses at the coasts.

We did know that the three main regions with underwater bedrock (Aurora, Wilkes and Recovery: see image below) could be vulnerable, given enough warming, over a long enough time. Records going back millions of years showed that ice was absent or lost from these places in several warm periods, when levels of carbon dioxide in the air were similar to those of today or expected this century  (for example, 3 million years ago: see image below). We also noticed when two of East Antarctica’s glaciers — one in the Aurora basin, the other in Wilkes — started to thin. We called her a sleeping giant.

Simulation of the thickness of the Antarctic ice sheet around 3 million years ago, during the warm mid-Pliocene period. Most of present day West Antarctica is not present, and neither are large parts of the three basins with underwater bedrock (“subglacial basins”, SB): Recovery (RSB), Aurora (ASB) and Wilkes (WSB). Source: Figure 1b from the new review by Stokes and others (2022), showing a simulation from DeConto and others (2021).

But those changes in the distant past happened over hundreds to thousands of years – at least, we couldn’t measure if they were any faster, and could therefore threaten our near future. And as recently as 2013, computer models weren’t advanced enough to predict how the East Antarctic ice sheet would change with global warming. So the Intergovernmental Panel on Climate Change (IPCC) used simpler methods, extrapolating from the past to the future. This method predicted only a tiny East Antarctic contribution to sea level rise, of about a centimetre this century.

We knew changes were a risk. But we believed this ice sheet was immune, for now.

Over the past decade, it’s become clear that the dynamics in East Antarctica have shifted. That several glaciers are vulnerable and thinning, because they are being gradually eroded beneath the surface by the ocean, just as they are in the West. A new generation of computer models have predicted that annual snowfall may in future no longer offset the melting at the coasts. The contribution to sea level rise this century under high emissions could be ten centimetres, twenty, or more.

Tens of centimetres may not sound like much. But the total global sea level rise of twenty centimetres predicted under medium emissions by 2050 — from Antarctica and every other source — would increase the frequency of extreme sea levels currently seen once in a hundred years to more than twice in a decade.

And in the next few centuries, some predictions for East Antarctica turn from centimetres into metres. Now we’re not talking about new coastal defences, but new coastlines.

The edge of Vanderford Glacier, one of the major outlet glaciers that appears to be thinning and retreating in Wilkes Land, East Antarctica. Source: Richard Jones

We realised it was time to pause, to take stock. To inspect the foundations of East Antarctica in more detail, without being distracted by the more visible and advanced changes in the West.

So we’ve published a review of the science of East Antarctica’s history, present and future, summarising evidence from 183 research studies. It describes our understanding of the ice sheet from glaciology, geology, oceanography, observations, and reconstructions of the deep past, and the computer models that are grounded in all of these. A review means we didn’t make new observations or run computer models, but we did some number-crunching to extract and compare results from the studies. I was responsible for the future projections, working with Catherine Ritz: for example, I recalculated one of our previous studies on uncertainties in rapid Antarctic collapse to predict only the East Antarctic part, and, as an author of last year’s IPCC report, I extracted details that weren’t presented in our chapter covering sea level rise.

The key numbers are in the image below (kindly made by co-author Richard Jones, with input from the study’s lead author Chris Stokes and me). It shows ranges of predictions for the East Antarctic contribution to sea level rise at 2100, 2300 and 2500, for two different emissions scenarios: global warming kept below 2°C, consistent with the Paris Agreement, or a very high emissions scenario, which reaches 4-5°C of warming this century and an unimaginable 10°C by 2300. We are currently heading for the lower end of this range: 2-3°C of warming this century. But only if we keep to the climate promises and policies we’ve made.

Infographic showing key projections summarised by in the new review by Stokes and others (2022). Source: Richard Jones.

The different shades show the uncertainties: lighter blues are less likely to happen than darker. Low and High show the lower and upper ends of the IPCC assessment for 2100, which was based on our big study and another one, and of all existing studies for 2300 and 2500. (There were too few studies extending to those times for an IPCC assessment). These aren’t absolute minimum and maximum numbers — that’s much harder to do — but are plausible bounds of what we predict is likely to happen in the future. Mid simply shows the mid-point of Low and High.

The Upper values for the high scenario are less likely to happen, but can’t yet be ruled out. The IPCC describes this as having low confidence in the underlying studies, which are based on expert judgement and a model that predicts rapid collapse of Antarctica. (The two studies didn’t extend to 2500, so there is no Upper here). Our climate policies and promises mean these contributions should be even less likely, because they aim to avoid these very high emissions. But the higher our emissions are, the more likely we are to trigger faster ice losses, like these.

Under the very high emissions scenario, East Antarctica is predicted to contribute up to 20 cm (or perhaps nearly half a metre) to global sea levels by 2100, up to three metres (or perhaps nearly five) by 2300, and over five metres by 2500.

But if we keep global warming below the 2°C limit of the Paris Agreement, the contribution stays below about half a metre throughout the five centuries: significant, but far less to adapt to with coastal defences and retreat.

Those are the high-end numbers: a world in which we’re unlucky and find the ice sheet is very sensitive to climate change. If we’re lucky, it will have middling sensitivity, and contribute just a couple of centimetres of sea level rise by 2100. And almost all of the Low sea level projections are below zero, which means there is a chance that increasing snowfall will offset the ice losses.

If we’re lucky.

What does it feel like to realise we could change the world’s coastlines?

Now we’re paying close attention. Keeping an eye on the cracks in the ice, the stability of the foundations, the erosion from below. Taking even more care than we did before.

The future may be unclear, but that doesn’t mean it’s out of our control. The fate of the East Antarctic ice sheet is still in our hands.

Technical details from IPCC reports

levels of carbon dioxide…today or expected this century: IPCC Sixth Assessment Report (2021), Technical Summary, shows concentrations reaching around 600ppm and 850ppm by 2100 for medium scenarios SSP2-4.5 and SSP3-7.0 (Figure TS.4). Our review describes various warm periods including the mid-Miocene climatic optimum, 600-800ppm, and the mid-Pliocene, around 400pm (i.e. similar to today).

two of East Antarctica’s glaciers: IPCC Fifth Assessment Report (2013), Chapter 13, FAQ 13.2, “Field and satellite-based observations, however, indicate enhanced outflow—manifested as ice-surface lowering—in a few localized coastal regions… (Pine Island and Thwaites Glaciers in West Antarctica, and Totten and Cook Glaciers in East Antarctica)

computer models weren’t advanced enough…: IPCC Fifth Assessment Report (2013), Chapter 13, Section “No process-based modelling is available on which to base projections of EAIS glaciers currently losing mass, such as Totten and Cook Glaciers.”

about a centimetre this century: from our review, which shows results from the IPCC (2013) method updated for IPCC (2021) using the latest climate models (in text and first row of Figure 6a).

the dynamics in East Antarctica have shifted: IPCC Special Report on the Oceans and Cryosphere in a Changing Climate (2019), Chapter 4, Section “Dynamic ice loss driven by ocean changes have also been observed on the East Antarctic margin (Li et al., 2016; Shen et al., 2018). This is an important development, because East Antarctica contains much more ice than West Antarctica, so even minor changes there could make major contributions to sea level in the future.” – both cited papers are referring to Totten Glacier in the Aurora basin.

would increase the frequency of extreme sea levels: IPCC Sixth Assessment Report (2021), Chapter 9, Table 9.9: SSP2.45 has projected median 20 cm global mean sea level rise by 2050 relative to 1995-2014 (Table 9.9), and a frequency amplification factor of 22, i.e. an average of 22 of these events per century, or 2.2 per decade (Section

  1. A lovely reflection, Tamsin, built on a fantastic review and analysis. For me the bottom line is actually encouraging, seeing no sign of the calamitous projections by 2100 but lots of reasons to slow warming. The challenge is huge – doing the exercise of pulling back from what would be ideal to what is conceivable in the world we know today. Work toward that Paris numbers or better, but also get comfortable with the reality that new generations will face new coastlines and humanity has to take on the exciting – yes and unnerving – task of reinventing coastal living even while cutting risk by cutting emissions. More thoughts from me, A.R. Siders and Jeremy Bassis in the link to my Sustain What dispatch: A Foot of Sea Rise by 2050 Spurs Fresh Efforts to Connect Changes in Ice Sheets with Action on Main Street

Leave a Reply

Your email address will not be published. Required fields are marked *

Add your ORCID here. (e.g. 0000-0002-7299-680X)

Related Posts
Back to top