Don;

I'm doing something very similar in the way quakes are chosen as possible hits.

Since I don't have fixed limits I use a multiple of the predicted ranges to determine what might poissibly fit, then select the best ones from those.

Roger