I've read that peak lateral ground acceleration for magnitude 7 - 8 earthquakes is ~1 g at the epicenter (is that correct?), but I'm wondering if it might be significantly higher right at the rupturing fault itself. This paper compares theoretical models of this motion but, in scanning through it, I didn't find any empirical measurements:
What is the approximate range of lateral ground acceleration (and peak velocity), as a function of earthquake magnitude, right at the fault, for strike-slip earthquakes?