The Earthquakes That Outrun Their Own Waves
A jet crossing the sound barrier does not gently speed up past its own noise. It compresses the pressure waves it is making into a single sharp front, and the boom arrives all at once. Some earthquakes, it turns out, do something close to the same thing underground.
Most ruptures crawl along a fault more slowly than the shear waves they radiate. A supershear rupture outpaces them. The tear travels faster than its own shear waves can carry the shaking away, and the waves pile up into a Mach cone, described in a new study as "an intense shock front similar to the sonic boom of a supersonic jet." Once treated as a rare curiosity, supershear behaviour has been identified in roughly 36% of magnitude-7-and-larger strike-slip earthquakes over the past 15 years, a share large enough that seismologists can no longer file it under exceptions.
The question that follows is what a rupture like that does to the ground, and to the buildings standing on it. That is what Mohamed Abdelmeguid of Caltech and colleagues set out to quantify in a study published 21 July 2026 in the Bulletin of the Seismological Society of America, summarized in a release from the Seismological Society of America. Rather than untangle the peculiarities of any one disaster, they ran idealized ruptures, stripped-down simulations designed to isolate the supershear signature from everything else a real fault brings with it.
Three effects stood out. The first is reach. Peak ground velocity, a measure of how violently the earth moves, stayed elevated as far as 20 kilometres from the fault, farther out than existing ground-motion models predict. The Mach front carries its punch well beyond the immediate rupture zone.
The second is duration, close in. Within about 7 kilometres of the fault, supershear ruptures shook the ground longer than ordinary ones. The authors trace it to a "double punch": the supershear pulse strikes first, then a slower subshear pulse follows behind it, so a site near the fault takes two hits instead of one.
The third is the part that reaches straight into engineering. Supershear ruptures produced larger spectral accelerations in the frequency band that matters most to buildings of about three to five storeys. That is not an abstract detail. It means the shaking concentrates its energy in exactly the range most likely to threaten a very common class of mid-rise structure, and it does so at levels beyond what standard models feed into building codes.
These are idealized ruptures, not reconstructions of specific earthquakes, so they isolate a signature rather than forecast a particular event. And the supershear effect is not a switch that stays on. The team notes that "short or episodic supershear bursts produce ground-motion characteristics that look quite similar to subshear cases." Sustained supershear is what carries the distinct, heightened signature; brief flickers of it largely wash out.
Even bounded that way, the direction of the result is what unsettles. Supershear ruptures are common among large strike-slip earthquakes, they now appear to shake harder and longer than the codes assume, and recent disasters have carried their fingerprint. The 2023 earthquake in Turkey and the 2025 Myanmar earthquake, which damaged the Great Wall Hotel in Mandalay, both show supershear behaviour. As identifications like these accumulate, the case grows for ground-motion models, and the building rules that lean on them, to reckon with the earthquakes fast enough to outrun their own waves.