The 2025 M 8.8 Kamchatka Earthquake, an Event at the Parkfield of M 9s

Today there was a M 8.8 earthquake on the Kuril subduction zone off of Kamchatka, Russia. Other than this being a simply massive earthquake, but otherwise a typical subduction megathrust event, not much is known about it. Surely it is special in its own way.

However, the site of the event is notable and I think not widely recognized. As is known among those who care about great subduction zone earthquakes, there was a M ~9 event in 1952. Wikipedia calls it the 1952 Severo-Kurilsk earthquake and ascribes it a work magnitude M_w 9.0. Johnson and Satake (1999) invert tsunami tide gauge data and give it M_w 8.8, similar to today's assessment by the USGS (after being upgraded from an 8.0).

But what is less well known, and until today of lesser significance, was that the area also had earthquakes of similar size in 1737 and 1841. Now, I don't know any more about those historical events than what is on the Wikipedia pages, but we should obviously give the magnitudes large uncertainties, perhaps with a bias towards the magnitudes being lower (as the magnitudes of historical earthquakes are often overestimated). Nonetheless, in the 1841 earthquake, a tsunami of 15 m was apparently recorded in Hilo, Hawaii. So it wasn't a M 7.0 event.

There are also reports of M 8.0 events in 1904, 1923, 1959, and a 7.8 in 1997; Johnson and Satake (1999) and Bürgmann et al. (2005) show maps derived from earlier Russian sources indicating that most of these events were on the margins of the 1952 event (and that the 1841 event may have been more compact).

But 2025, 1952, 1841, 1737. M 8.8 or 9.0 events.

Just using the integers of the years, that gives a mean recurrence interval of 96.0 +/- 16.5 years, for Magnitude F'in 9 earthquakes, in what really looks like the same location.

I checked the locking results from that part of the Kuril subduction zone from my global block model. I get about 50 mm/yr of dip-slip strain accumulation in the upper ~20 km and 30 mm/yr in the lower. (I will try to make a figure tomorrow.) Bürgmann et al. (2005), using worse data but focusing on that part of the megathrust specifically, find asperities up to about 80 mm/yr but dropping smoothly down to zero for the rest.

Locking estimates from Bürgmann et al. (2005).

Either of these values accumulating for 100 years would yield 5-8 m of slip available in the shallower part of the zone, which might be about right for a M 8.8.

I think this is fascinating. So much has been done at Parkfield, which hosts similar M 6.0 events at 24.5 +/- 8.4 years on average since the 1850s. The fact that the sequence was so periodic, until it wasn't, is still very interesting. But Parkfield is kind of a pathological spot at the transition between the locked and creeping parts of the San Andreas. And M 6.0 earthquakes are notable and can be deadly in the wrong circumstances, but they are not in general the earthquakes to really worry about unless you build your unreinforced masonry building right on the fault.

This site hosts several of the largest earthquakes ever recorded, and seems to have been kicking them off somewhat regularly since European colonization (perhaps there are Japanese records before that?). Although the political situation right now is not as favorable as in the past, it would still be excellent if the Russians can dig into the situation, so to speak, and invite international scientists too. Surely the paleoseismic record is rich.

A final note is that the Kuril subduction zone is notable for transitioning from ocean-continent subduction at Kamchatka to ocean-ocean to the south until Hokkaido. The ocean-ocean section hosts M 8 events, such as an M 8.3 on 2005-11-26, but do we know anything about M 9s there? I don't know of much evidence for M 9 events on ocean-ocean subduction zones. I guess if the zone can host an 8 and there are no geometric limitations, a 9 should be plausible too. But is there any record of such events, globally?