That handles distance, but not magnitude, which I assume is based on some sort of amplitude.

Incidentally, from what you said above, I'm hugely puzzled to try and imagine what sort of physical cause could operate in such a way. Clearly it makes no sense to argue that the crust emits a signal before an earthquake the duration of which is proportional to the distance to Santa Clarita. So one presumes the signal is not always the same strength, and for a more distant earthquake you only pick up the strongest part (at the end/beginning/middle?) at that range. If that is the case, for a close event you should see a stronger signal for part of the duration corresponding to what you would record if the eq was more distant. Is that the case? Or can't you tell with limited observing time? You really do need a way of making recordings.

Roger