← Did You Know?
🧊Earth8 min read

Thwaites: The Doomsday Glacier

It is roughly the size of Great Britain, it sits on ground that slopes the wrong way, and it is holding back a great deal more ice than it contains. Scientists mostly hate the nickname. The geometry is the genuinely alarming part.

In plain English

Most of what makes Thwaites frightening is not its size. It is the shape of the rock underneath it.

Thwaites is a glacier in West Antarctica, around 120 kilometres wide where it meets the sea, and roughly the area of Great Britain. On its own it holds enough ice to raise global sea level by about 65 centimetres.

That figure is bad but survivable. The problem is what sits behind it.

Much of the West Antarctic Ice Sheet rests on bedrock that is below sea level. Thwaites acts as a plug at one of the main exits. Remove it and you expose a far larger volume of ice to the ocean — enough, across the whole ice sheet, for something in the region of three metres.

Now the geometry. Where a glacier stops resting on rock and starts floating on water is called the grounding line. At Thwaites, the bed beneath that line slopes downward as you go inland. Retrograde, in the jargon.

Think about what that means. As the grounding line retreats, it moves into deeper water. Deeper water means a thicker column of ice at the front, which means more of it floats, which pushes the grounding line further back — into deeper water again.

That is a feedback loop that does not need any further warming to keep running. It is called marine ice sheet instability, and it is the reason Thwaites gets studied harder than almost any other glacier on Earth.

Five things to file under "wait, what?"

  • "Doomsday Glacier" is a journalist's phrase, and the scientists studying it mostly dislike it. It implies a sudden event. What the research actually describes is a process running over centuries, which may already be committed regardless of what happens next. That is arguably more serious than a catastrophe, because you cannot change your mind about it later.

  • The warm water is arriving from below, not above. Circumpolar Deep Water — relatively warm, salty and dense — flows along the continental shelf and up into the cavity beneath the floating ice shelf. Thwaites is being melted from underneath by seawater that never sees the surface, which is why air temperature is not the whole story.

  • A robot has been under there. In 2019 and 2020 the Icefin submersible was lowered through a 600-metre borehole and driven to the grounding zone itself. It found warm water reaching the point where the ice meets rock — the worst place for it to be.

  • The melting is uneven in an unhelpful way. A 2023 study found that melt across the flat underside of the ice was slower than models assumed, which sounds like good news. But melt inside crevasses, cracks and stepped terraces was rapid. The ice is not thinning evenly; it is being carved along its weaknesses, which is harder to model and worse for structural integrity.

  • It has retreated faster in the past than it is retreating now. Sea-floor evidence published in 2022 shows ridges cut by the grounding line during a period when it was pulling back at around 2.1 kilometres per year — roughly twice the fastest modern rate. Whatever happened then, the system is capable of moving quicker than we have observed it moving.

The full story

Where it is and why it matters disproportionately

Antarctica is really two ice sheets. East Antarctica is enormous, sits mostly on rock above sea level, and is comparatively stable. West Antarctica is smaller, and much of it sits in a basin below sea level, which means the ocean has access to it.

Thwaites and its neighbour Pine Island Glacier drain a large share of that basin into the Amundsen Sea. Together they account for a substantial and growing fraction of Antarctica's contribution to sea level rise. Thwaites alone is currently responsible for something around 4 per cent of global sea level rise, which is a remarkable share for one glacier.

Its importance is structural rather than volumetric. It is a buttress. Ice shelves and glacier tongues push back against the ice behind them, slowing its flow to the sea. Remove the buttress and the ice behind accelerates.

Marine ice sheet instability, properly

The idea goes back to work by John Mercer in the 1970s and was formalised over the following decades. It is worth being precise about, because it is the whole argument.

A glacier grounded below sea level floats where the ice becomes thin enough for buoyancy to lift it. If the bed deepens inland, then any retreat of the grounding line puts it somewhere the ice is thicker and the water is deeper. Flux through the grounding line increases with ice thickness — steeply. More ice pours out, the glacier thins further, the line retreats again.

There is no requirement for the ocean to keep warming once this is under way. The trigger and the process are separate things, which is why researchers talk about retreat being "committed" rather than "happening".

What is genuinely uncertain is the timescale, and how much other physics — bed roughness, sediment, the shape of the coastline further inland — slows things down.

What the fieldwork found

The International Thwaites Glacier Collaboration, a roughly £40 million joint UK and US programme, was the largest cooperative Antarctic field project in decades. Getting people and equipment onto Thwaites is genuinely difficult: it is remote even by Antarctic standards, heavily crevassed, and the weather is hostile.

The results complicated the picture rather than simplifying it. Grounding line retreat of around 14 kilometres was documented between the early 1990s and 2011. Warm water was confirmed at the grounding zone. But the melt rates measured beneath flat sections of ice were lower than models had assumed, while melt in the fractures was fast.

The honest summary is that Thwaites is losing ice, the mechanism is understood in outline, and the rate over the coming century remains genuinely uncertain — with the uncertainty running in both directions.

Why the framing matters

Sea level rise of three metres would redraw the map. It is not a flood that arrives one afternoon; it is a slow, unstoppable relocation of coastlines, ports, cities and farmland, playing out over generations while everyone watches.

The reason to be careful with the word "doomsday" is not to soften the science. It is that a story about imminent catastrophe invites a response of either panic or dismissal, and the actual situation asks for something harder: sustained attention to a process slower than a news cycle and faster than most institutions can plan for.

The glacier does not care which. The geometry of the bedrock was set long before anyone was watching.

Go deeper

For the curious:

On YouTube:

Keep going

More things worth knowing

All topics →