Mars’s giant canyon Valles Marineris is still cracking, a marsquake shows

The largest canyon in the solar system is not a dead landscape. A marsquake recorded by NASA’s InSight lander came from a fault slipping deep beneath Valles Marineris, the canyon system that slashes across Mars for nearly four thousand kilometres, which means faults opened there billions of years ago can still break under the stresses acting on the planet now.

That matters because Mars has no moving tectonic plates. On Earth, plates grinding past each other keep old faults alive. Mars’s outer shell behaves like a single, slowly cooling lid, and faults in that kind of crust can heal and strengthen over time. The new analysis shows that at least one deep fault under the canyon has not healed shut, some twenty to forty kilometres below the surface.

The study, published in Nature Communications, was led by Jung-Hun Song of Imperial College London with Doyeon Kim, also at Imperial, Folarin Kolawole of Columbia University and colleagues. It rests on a single event, catalogued as S0976a, the only marsquake InSight ever located in the planet’s western hemisphere.

“Valles Marineris is one of the most spectacular geological structures anywhere in the Solar System, but until now we didn’t know for certain that its faults are still active,” Song said. “What we see is that this is not just a ‘dead’ fossil landscape, as faults deep beneath the canyon can still slip and generate marsquakes, long after the geological processes that originally formed them have ceased.”

Reading a fault from one seismometer on the far side of Mars

On Earth, seismologists work out how a fault moved by comparing the first jolt recorded at dozens of stations around the quake. InSight had one seismometer, and S0976a struck almost on the opposite side of the planet, about 146 degrees of arc away, or roughly 8,600 kilometres across the surface. The direct waves never reached the lander in usable form. What arrived were PP and SS waves, pressure and shear waves that bounced once off the Martian surface halfway along the trip.

The team modelled those two wave trains with computer-simulated seismograms for several models of the Martian interior, testing every combination of fault orientation, slip direction and depth down to 50 kilometres. To check that a lone station could pull this off, they ran the same method on a magnitude 6.5 earthquake in Nevada recorded by a single seismometer in Australia at almost the same distance. It recovered the correct fault.

For S0976a, the answer was a strike-slip rupture, in which two blocks of crust slide horizontally past each other along a steep fault, like California’s San Andreas. Of all the solutions that fit the waves acceptably, 82.2% belong to that family. The rupture sat about 20 to 40 kilometres down, in the middle to lower crust, above the base of the crust at 55 to 65 kilometres. Combining the wave fit with maps of known faults and the direction the high-frequency waves came from places the source near Ius Chasma, in the western part of the canyon.

Ruling out a landslide or a meteorite

A seismic signal on Mars is not automatically a fault. Meteorite impacts and landslides also shake the planet, and Valles Marineris is lined with giant landslide deposits. The team tested those sources directly. Impact-like explosions and force-based landslide models could not reproduce the polarities, timing or relative strength of the waves, and they could not produce the strong horizontal shaking in the data. Confirmed impacts on Mars also tend to produce high-frequency signals, while S0976a was a low-frequency event.

The researchers then searched every image of the source region taken by the Context Camera on NASA’s Mars Reconnaissance Orbiter after the quake. A landslide big enough to make a signal like this would need to move at least 10 million cubic metres of rock and leave a fresh deposit several hundred metres to a kilometre across. There was none, and no fresh crater either. By contrast, the only later distant event in InSight’s catalogue, S1000a, turned out to be a meteoroid impact with a new crater spotted from orbit.

“What makes the result compelling is that several completely different types of evidence point in the same direction,” Song said. “The seismic waves tell us that there was a sideways sliding motion on a fault well below the surface, while the geology seen from orbit shows a region crossed by long-lived fault systems.”

A canyon nine times longer than the Grand Canyon

Valles Marineris stretches nearly 3,900 kilometres, reaches up to 600 kilometres wide and plunges more than nine kilometres deep, according to Imperial. That makes it nearly nine times longer, more than 20 times wider and about five times deeper than the Grand Canyon at their maximum dimensions. It formed alongside Tharsis, the vast volcanic bulge that holds Olympus Mons, and geologists have long argued over whether it opened as a rift, collapsed under its own weight or was torn open at an angle.

The marsquake weighs in on that debate. The stress pattern it records does not fit models in which the canyon formed by gravitational collapse of thickened crust. It fits a canyon torn open obliquely, with sideways shear as well as stretching. The quake’s direction of maximum squeeze, roughly west-northwest, also matches the compression inferred for the canyon’s early deformation in the Hesperian, an early period of Martian history. The same stress direction appears to be operating on the same fault fabric billions of years later.

The authors point to three present-day forces that could load those faults: the slow contraction of a cooling planet, the weight of the Tharsis bulge, and the unloading caused by erosion carving out the canyon. With S0976a, InSight has now revealed three different styles of active faulting on Mars: pulling apart beneath the Cerberus Fossae fissures, squeezing near the crustal boundary in southern Elysium Planitia, where the mission’s largest marsquake, S1222a, struck, and sliding beneath Valles Marineris.

“It shows that on a planet without plate tectonics, structures created billions of years ago can remain important and still control where the crust breaks today,” Kolawole said.

What it doesn’t settle

Everything rests on one marsquake recorded by one instrument. With a single station, the method finds two possible fault planes at right angles and cannot say which one actually slipped, and the estimated orientations do not line up with a single dominant fault set mapped at the surface. The location also leans partly on the density of mapped faults, which the authors call an approximate guide to where deep slip occurs, not a direct measurement.

No aftershocks were recorded. The team argues that this is expected: in their simulations the chance of InSight detecting even one sizeable aftershock from that distance within 50 Martian days was only 0.1%, because the planet’s core and deep interior block and weaken waves from the far side. The silence does not count against the result, but it also leaves no second event to confirm it.

The study cannot say how often the canyon breaks, or how large its quakes can get. S0976a was a moderate event, consistent with limited slip on a small patch of an existing fault, not a new fracture. The authors also cannot separate how much of today’s stress comes from the bending of the crust under Tharsis and how much from deeper support, and they note that the small number of well-located marsquakes may reflect InSight’s short, single-station record as much as a quiet planet.

Common questions about the Valles Marineris marsquake

Is Mars still tectonically active?

Yes, at least in places. InSight recorded marsquakes during its mission, and analyses of individual events show active faulting beneath the Cerberus Fossae fissures, southern Elysium Planitia and now Valles Marineris. Mars has no moving plates, so the activity comes from stresses within a single rigid shell.

How big is Valles Marineris compared with the Grand Canyon?

It runs for nearly 3,900 kilometres, up to 600 kilometres wide and more than nine kilometres deep. That is nearly nine times longer, more than 20 times wider and about five times deeper than the Grand Canyon.

What caused the marsquake under Valles Marineris?

According to the study, a strike-slip rupture on an old fault 20 to 40 kilometres below the western canyon, where blocks of crust slid horizontally past each other. Meteorite impacts and landslides were ruled out because they could not reproduce the recorded waves and no fresh crater or slide appeared in orbital images.

Does Mars have plate tectonics?

No. Its outer shell acts as a single stagnant lid that has been cooling and contracting for billions of years. This study shows that faults inside that lid can still slip under today’s stresses.

InSight recorded S0976a on 25 August 2021, its 976th Martian day on the surface, and the mission ended in December 2022. The paper appeared on 15 October 2026. With no seismometer now operating on Mars, the existing InSight catalogue is the only seismic record of the planet, and the team says this result shows how much information those recordings still hold.

The same quake has already been mined once: its waves passed through the Martian core and helped map the planet’s deepest layer. “From a single marsquake, we can learn both about the centre of the planet and about an active fault beneath one of its most remarkable landscapes,” said Kim, the study’s senior author.

Reference: Song et al., “Active crustal straining and fault reactivation at Valles Marineris,” Nature Communications, 2026. DOI: 10.1038/s41467-026-78193-6

Tags: astronomy, nasa, Mars, InSight, marsquake, planetary geology