EQF,

The only curve that is trivial for me to check is wrong.

Change in Earth acceleration should be the derivative of Earth acceleration. This means the peaks in one curve should line up with the zero-crossings in the other curve. Unless you're doing something weird to collapse the actual vectors to the plotted scalars, at least one of the two curves is wrong.

Your footnotes explicitly say the plotted scalar change in acceleration is derived from the scalar acceleration, albeit with an impossible finite difference scheme using a one-day time step.

And, as I noted before, the velocity of the Earth is unlikely to affect earthquakes.

John