How Melting Ice Triggers Earthquakes and Tsunamis

Uh oh, controversial topic. But hear me out, because there’s some real science behind what I’m about to discuss. This isn’t about collapsing volcanoes or megathrust subduction zones. It’s about ice. Specifically, how the retreat of massive ice sheets can quietly destabilise the Earth’s crust and, under the right conditions, trigger earthquakes and tsunamis in places like Greenland—regions we usually describe as tectonically stable. The study I’m drawing from even describes a plausible trans-Atlantic tsunami generated by the reactivation of an ancient fault line during glacial retreat. I’ll walk you through the mechanism that makes this possible, and then we’ll look closely at the ancient event the study focuses on.

When most people picture earthquakes and tsunamis, they imagine plates colliding, slabs diving into the mantle, and violent boundaries like Japan or Chile. Greenland doesn’t fit that picture at all. It sits far from active plate margins, there’s no subduction trench swallowing oceanic crust, and modern seismicity there is relatively low. But that calm appearance hides something important. Greenland is, and has been for tens of thousands of years, one of the most heavily loaded pieces of continental crust on Earth. Ice sheets don’t just sit on the surface like snow. When they grow to several kilometres thick, they behave like a planetary-scale weight pressing the lithosphere downward and squeezing it from all sides.

This pressure changes the stress state of the crust in a very specific way. Under a thick ice sheet, vertical stress increases dramatically simply due to the weight of the ice. At the same time, horizontal stresses rise as the crust is flexed and compressed. In this configuration, faults that might otherwise be capable of slipping are effectively locked shut. It’s one reason Greenland today has fewer large earthquakes than you might expect for a continent-sized landmass with a complex geological history. The ice acts like a suppressor.

The crucial point—and this is where things get unintuitive—is that when ice retreats, those stresses don’t relax evenly. Vertical stress drops quickly as the weight is removed. Horizontal stress, however, lags behind. The lithosphere doesn’t spring back instantly; it responds slowly because it’s underlain by a viscoelastic mantle that flows over thousands of years. That mismatch matters. It creates a transient stress imbalance where faults that were once clamped shut are suddenly much closer to failure. In geology, timing is everything, and this short window during deglaciation can be enough to reactivate ancient faults that haven’t moved in millions of years.

This process is known as glacially induced faulting, and it’s not theoretical. Northern Europe preserves clear surface ruptures from earthquakes that occurred during the retreat of the Scandinavian Ice Sheet after the last ice age. Some of those earthquakes were enormous by intraplate standards, potentially reaching magnitudes comparable to major plate-boundary events. For a long time, Greenland was missing from this discussion—not because the physics didn’t apply, but because the evidence hadn’t been examined in detail.

That’s what makes the study so interesting. The researchers set out to model how Greenland’s crust responded to ice loss during the early Holocene, when the ice sheet was shrinking rapidly. Using glacial isostatic adjustment models combined with fault mechanics, they calculated how stresses evolved over time and where faults would be pushed toward instability. Their results showed something striking. Around 10,600 years ago, southern Greenland crossed a threshold where faults offshore became mechanically unstable due to ice retreat alone

On its own, that might sound abstract, but the story becomes much more concrete when you look at sea level data. In the Nanortalik region of southern Greenland, scientists have some of the best relative sea-level records in the Arctic. These come from isolation basins—natural rock depressions that were once connected to the ocean and then cut off as land uplifted. The elevations of sediments in these basins tell us where sea level was at specific times in the past.

Here’s the problem. When researchers tried to match these sea-level records with even the most advanced ice and Earth models, something didn’t line up. The oldest data points sat far higher than predicted—by more than ten metres in some cases. Tweaking ice thickness, mantle viscosity, or Earth structure didn’t fix the mismatch. No matter how the models were tuned, the discrepancy remained.

That’s where the faulting hypothesis enters. The authors proposed that the sea-level records weren’t wrong, and the ice models weren’t fundamentally broken. Instead, the land itself had moved suddenly. If an offshore fault slipped during deglaciation, it could have uplifted the coastline and the isolation basins all at once, shifting the apparent sea-level record upward. When they incorporated a fault rupture into their model—timed to when the crust became unstable—the mismatch largely disappeared. The geology and the physics snapped into alignment.

From there, the tsunami question becomes unavoidable. When an earthquake causes vertical displacement of the seafloor, especially along a thrust fault, it pushes water upward and outward. That’s the fundamental mechanism behind earthquake-generated tsunamis everywhere on Earth. The difference here is location, not physics.

The study explored two end-member scenarios. In the first, stress was released in a single large earthquake, with a magnitude potentially exceeding 8. In that case, tsunami modelling showed waves radiating across the North Atlantic. Greenland’s southern coast would have experienced the largest waves, but energy would also have reached North America and Europe. After accounting for shoaling near coastlines, the authors calculated run-up heights of several metres along parts of the British Isles and eastern Canada

In the second scenario, the stress was released through a series of smaller earthquakes spread over decades or centuries. Each individual event would have produced only modest waves, unlikely to leave obvious geological traces far from Greenland. Importantly, the available evidence doesn’t allow us to distinguish cleanly between these scenarios. The sea-level data can accommodate either a single dramatic rupture or multiple smaller ones.

This is where scientific caution is essential. No confirmed tsunami deposits from this event have been identified. But the absence of evidence here is not evidence of absence. At the time, much of the surrounding coastline was ice-covered, sea level was lower than today, and any tsunami deposits would now lie offshore, buried or eroded by subsequent sea-level rise. The authors are careful to emphasise this uncertainty, and so should we.

So what does any of this have to do with today? This is where the topic becomes controversial—but only if it’s framed carelessly. The study does not claim that Greenland is on the brink of generating tsunamis right now. It explicitly notes that modern seismicity has not yet increased in response to recent ice loss. However, it does something more subtle and arguably more important. It demonstrates that ice retreat is capable of altering crustal stress fields enough to reactivate faults in regions we normally consider geologically quiet.

The early Holocene event happened during a period of rapid warming, but that warming unfolded over millennia. Today, Greenland is losing ice at a rate that is extraordinary by geological standards. The same physical mechanisms are in play—unloading, delayed rebound, stress reorganisation—but compressed into a much shorter timescale. Whether that leads to future fault reactivation depends on factors we don’t fully understand yet: fault orientation, friction, pore pressures, and the detailed structure of the crust beneath the ice.

This is why the study’s conclusion matters. It’s not a prediction; it’s a warning about possibility. If offshore faults exist in stress configurations favourable to slip, continued ice loss could eventually push them toward failure. And if that failure involves significant seafloor displacement, tsunami generation becomes a downstream consequence, not an extraordinary leap.

There’s also a broader implication that often gets overlooked. Relative sea-level records are used worldwide to reconstruct past ice sheets and calibrate climate models. If some of those records were shifted by faulting during deglaciation, then ice histories inferred from them could be subtly biased. In other words, earthquakes triggered by ice retreat don’t just pose hazards—they can quietly rewrite our interpretations of the past.

So when people hear a phrase like “global warming triggering tsunamis,” it’s easy to recoil. It sounds sensational, even irresponsible. But when you slow it down and follow the chain of evidence, what you find is a grounded, mechanical story. Ice sheets load the crust. Removing them changes stress. Stress changes can activate faults. Fault activation can displace the seafloor. Displaced seafloor can generate tsunamis. None of those steps violates known physics. The uncertainty lies in how often, how fast, and how large.

That’s why Greenland is such a compelling case study. It strips away the familiar tectonic excuses and leaves us with a simpler, more unsettling reality. Sometimes, the trigger isn’t plates colliding at all. Sometimes, it’s the planet taking weight off its own crust and responding in ways that are anything but gentle.

If nothing else, the ancient event examined in this study reminds us that Earth’s systems are deeply interconnected. Climate doesn’t just reshape landscapes through erosion and melting. Under the right conditions, it can reach deep into the lithosphere, waking faults we thought were long asleep—and, occasionally, sending waves across entire oceans.

Here's the video we made on this on the OzGeology YouTube Channel:

Link to the study used to construct this article:

Early Holocene Greenland-ice mass loss likely triggered earthquakes and tsunami