[jwalsh07]

Calculations performed by Richard Gross of NASA's Jet Propulsion Laboratory in California show that the quake sped up the rotation of the Earth and enlarged its wobble, causing the length of a day to shrink permanently by 3 millionths of a second.

FWIW, this is wrong. The rotation speed of the Earth is not a constant. It changes, mostly it slows down. But have no fear, Angular momentum is conserved even in the Meadowlands.

[VeritatisSplendor]

Angular momentum is conserved, but that doesn't mean the quake didn't change the rotation speed -- if the result was a "settling" of the ocean floor the moment of inertia would shrink and the rotation would speed up.

Of course the tidal friction with the moon changes the earth's rotation speed because it is the angular momentum of the entire earth-moon system that is conserved -- although the rotation slows down, the moon gets correspondingly further away (3 millimeters further away every month). At the time of the first dinosaurs, 250 million years ago, the moon was 5 or 6 thousand miles closer (which is still a small change compared to the current variation in the moon's distance from earth -- annular eclipses were still possible then, and total eclipses will still be possible for another 500 million years).

[staytrue]

I believe the rotation of earth could speed up if the high points on earth slid into low points. It would be like a skater who draws in his arms to accelerate the spin.

[KC Burke]

I was given to understand that the movement of one plate under the other in the subduction fault had actually pulled the diameter of the sphere in and much like a spinning skater pulling his arms in, the slightly tighter globe had increased its speed by three micro-seconds. If that is the explanation, only a release to greater size by a corresponding incident would lengthen it, in my meager conception.