George Zandt's 1979 anomaly under Randolph, Utah, was dismissed because continents were not supposed to quake that deep. A new study, anchored by a 2025 Maeser event, has confirmed a whole catalog of them.
When George Zandt flagged a magnitude 3.8 earthquake beneath Randolph, Utah in 1979, the seismological response was reasonable, not foolish: the focal depth sat near 90 kilometers below sea level, deep into the upper mantle, in a tectonic setting where no one expected earthquakes. The reading was published only as an abstract, and the case was treated as a curiosity that probably reflected an instrument glitch.
Forty-seven years later, a University of Utah team has confirmed that the 1979 event was real, and the real story is bigger than one anomaly. The same study documents a catalog of nine deep events in northern Utah and southwestern Wyoming, all of them tens of kilometers below the crust, all of them consistent with a category seismology used to call implausible under continents: continental mantle earthquakes, or CMEs.
The keystone is the September 10, 2025 Maeser earthquake, a magnitude 4.1 event in Utah's Uinta Basin. At 68 kilometers depth, more than 20 kilometers below the Moho, Maeser is the cleanest example yet of an "archetypal continental mantle event," the authors write in The Seismic Record. What that phrase really means is that the rock failed in sudden brittle fashion in a layer that, on long timescales, behaves like warm taffy. The mechanism is not settled, and the authors are clear about that. What is settled is that the failure mode exists, and that it is recurrent.
Zandt, by then retired from the University of Arizona, came out of retirement to co-author the University of Utah announcement of the result. The lead author is Keith Koper, director of the U of U Seismograph Stations. Graduate student Sean Hutchings re-ran the waveform analysis through the UUSS archive, and that reanalysis is what turned a single oddity from 1979 into a class.
The geography is part of the reason the class was hidden. The events cluster near the edge of the Wyoming Craton's lithospheric keel, the thick, cold root of old continental crust that runs beneath most of Wyoming and parts of Utah. A craton is, by definition, the part of a continent that has been tectonically quiet for the better part of a billion years. Textbook seismology has long treated cratonic interiors as seismically quiet down to the Moho and below. The University of Utah press release shows the events plotted as a constellation of stars along the craton's margin: the 1979 Randolph event, the 2025 Maeser event, six newly identified CMEs from 2007 to 2010, and four more suspected CMEs located by UUSS in 2025.
The honest reading of 1979 is not that the skeptics were wrong to be skeptical. The instruments available could not resolve a small, deep event from noise the way modern broadband networks can, and the prior of "impossible under a craton" was based on real geological reasoning. What changed was the data: denser station coverage, better waveform-processing methods, and an archive that goes back far enough to let a student in 2025 reconstruct what a 1979 seismologist had to take on faith.
The hazard implication is the part that does not have a clean answer. Koper's own framing, as quoted in the U of U release, is that with no mappable surface faults, the maximum possible magnitude of a continental mantle event is unknown, and that this unknown sits outside the crustal hazard frameworks that state and federal models use to set building codes. That is a real, specific, and properly attributed point. It is not a prediction of a big surface earthquake, because the events documented here are, in the relevant sense, the deep event catalog of a craton: small, isolated, with no foreshocks and no aftershocks, and never felt at the surface in any of the cases reported.
Two things to watch. The first is whether the mechanism question gets traction: in-situ brittle failure of nominally ductile mantle rock, dehydration embrittlement at the keel, or some combination specific to the Wyoming Craton's thermal history, are all on the table in the literature, and The Seismic Record paper is explicit that the question is open. The second is whether the catalog keeps growing. UUSS has been quietly filling in the suspected events from 2025, and a denser network under the Uinta Basin, combined with continued reprocessing of older archives, is the only honest way to test whether the nine events the new study documents are the start of a class or roughly all of it.
Either way, the 1979 anomaly is no longer a curiosity. It is the founding member of a catalog that seismology spent nearly half a century arguing could not exist, and the new analysis makes the case in instruments, not in rhetoric.