Nice article.

This was the first documented & researched example of a megathrust rebound/long duration mercalli X (r/lx) event. In this type of quake the locked megathrust finally lets go aft & rebounds up the trench. The result is decompression of the overriding plate & also a lurch of the previously held up descending plate. The result of the displacement of the ocean floor in the trench is a bloody great big tsunami.

Gordon Pflaker was the first to research the effect, and as a result he became the first to adequately describe r/lx events in detail.

As for Alaska, the quake had an Ms magnitude of 8.6 (big enough to produce land surface waves, which was bigger then the 8.4 in Tohoku or the 8.2 in Aceh). The 9.2 calculation was however the sum of the Ms waves over the entire length of the runout. this moment of 9.2 was the result of a variance of Ms waves over the length of the fault (which was biggest at the epicentre). The runout was also biggest at the epicentre & decreased toward the terminus (Kodiak). With Alaska also the Mp magnitude was approx 6.5 (a mite low). The reason for the difference between this and a normal thrust quake was the smooth lift off when the megathrust broke, with shaking as a result of the overriding plate rolling over humpy irregularities in the thrust plane. The result on the surface is an effect known in Indonesia for legendary ages as the "magic carpet ride" due to the large wide ripples on the ground surface & the seasick feeling.

1964 (Alaska [9.2]) was an arc runout with a very strong megathrust close to the Denali range margin with about 14m runout in the north and ~10m in the south.

Nowadays Since the 1964 r/lx there have been several others. 1965 Rat islands (Aleutians [8.8]) with an arc runout of around 10m.

2004 (South Andaman [9.0 since upgraded to 9.25 due to a silent event in the N. part of the runout]), with a backarc spreading driven runout averaging 25m with very little plate lurch from the descending plate backed up against the Himalayas.

2006 (Central Kuril islands [8.4]) A small short segment megathrust event with an arc runout of about 11m in the Kuril-Kamchatka backarc basin.

2010 (N.Central Chile [8.8]) An arc runout of about 12m in the segment N. of the 1960 9.45 event. This was smaller then the r/lx in the same segment described by Darwin as it did not include the Santiago segment near the Rioja range (still locked).

2011 (Tohoku [9.0]) A rare zygomorph event resulting from an overcompressed arc, with a megathrust developed on BOTH sides of the arc. In this case the E. megathrust let go aft with ~50m runout near the epicentre and an arc runout @ the S edge of about 11m and a medium runout @ Sanriku in the N. part of about 18m.
For the record the N. Tohoku Junction box with Kuril-Kamchatka let go in 1896 with a SE vector and a very similar sized tsunami @ sanriku & a runout of about 20m. The megathrust of the 1896 segment was crossed by the N part of the E. directed 2011 quake decompressing the forearc again.

Pflaker's 1965 paper describes the process adequately but was (rather critically) ignored in Japan, where the research extended only to thrusts in the forearc ABOVE the locked megathrust.

Megathrust events are if not clockwork, at least regular enough to be predictable via the conditions in the forearc & also the arc volcanism. The paper on arc volcanism connections to megathrusts has yet to be written (will be pending once the house is sold if I get bored with my spare time).

Have a nice day: Ag