The coastal region of Oaxaca, Mexico, has produced more than a dozen M ≥ 7 earthquakes in the last 120 years, and several of these events have struck in almost exactly the same spots decades apart. Figuring out whether these are true "repeaters" that rupture the same patch of the megathrust, or whether they simply happen to be nearby but not really overlapping, is not just an academic question: it tells us how persistent the frictional properties of a fault segment are through time and, ultimately, how confident we can be about forecasting the location and size of the next big earthquake in the region.

Prior work suggested that at least three pairs of large earthquakes along Oaxaca (1928/1968/2018 near Pinotepa Nacional, and 1928/1965/2020 near La Crucecita) are repeaters, based on how similar their seismograms looked at just one or two seismic stations. But is a match at a single station enough to say two earthquakes ruptured the same patch of fault? A similar-looking waveform recorded from only one direction could easily be hiding real differences in the rupture.

To test this more rigorously, we compared recordings of these earthquake pairs from stations spread broadly in azimuth all around the globe, the seismological equivalent of listening to the same pair of events from many different rooms instead of just one. The catch is that the 1965 and 1968 earthquakes were only ever recorded on old photographic paper rolls from the historic World-Wide Standardized Seismograph Network, so before we could compare anything, we had to track down the original film records at the Earthquake Research Institute in Tokyo, scan them, and manually digitize and calibrate the waveforms before processing them the same way as the modern digital recordings of the 2018 and 2020 earthquakes.

We found remarkably high waveform correlations (mostly above 0.9) between the 1968 and 2018 Pinotepa Nacional earthquakes, and between the 1965 and 2020 La Crucecita earthquakes, across dozens of station pairs at many different azimuths. This azimuthally broad agreement is much stronger evidence of true repeating ruptures than a single-station comparison, and it shows that the same asperities, or locked patches, on the megathrust have failed repeatedly, to within about 30 km, over at least three earthquake cycles in each of these two regions.

Interestingly, the stretch of coast in between, central Oaxaca, has not broken since a pair of large earthquakes in 1928 and 1978, roughly 45-50 years ago, which is about how long it usually takes for stress to rebuild on these patches. Combined with the high geodetic coupling and the slow-slip events recently documented on the deeper part of the plate boundary there, this makes central Oaxaca a strong candidate for the region's next major earthquake.

FOR MORE DETAILS ON THIS TOPIC PLEASE CLICK HERE