If the 1966 Parkfield earthquake really broke the shallow patch Custodio claims, none of the last three events had similar patterns of fault rupture. We already know from Segall's work that the 1966 event rupture extended further to the south than the 1934 one.

And if the 1966 and 2004 events broke largely separate sections of the fault, and 1934 broke a small fault plane, why in the world should there be a nearly steady 22-yr periodicity to earthquakes?

However, mixing geodetic and strong motion data, as has been used in the various studies, has well-known problems with resolving depth of rupture. If the 1966 rupture was really deeper, and broke a similar section of fault to the 2004 event, at least we might be seeing the breakage of the same asperity as part of the last three Parkfield earthquakes, which I had the impression that we were.

I've actually asked 4 or 5 people who study this sort of thing if the shallow 1966 fault plane is well-resolved, and they seemed surprised to hear the claim. Given that the Parkfield experiment was predicated on a repeat of the 1966 earthquake 22 plus or minus just a couple of years later, if the shallow 1966 fault rupture model is right, the characteristic earthquake model on which the Parkfield experiment was based has gone down in flames (at least for the case of this section of fault). Yet people continue to claim the 2004 event was the expected one.

Am I missing something here? Didn't someone give Custodio a grilling about such a revolutionary claim?