Roger:

Try this logic:

1) Increase radius by 10 %
2) If the number of quakes to evaluate within this new radius exceeds some number (say 10 ?) then go back and repeat step one but use only one-half of the previous factor (e.g. 1/2* 10 % = 5%)
3) Repeat step 2 as many times as needed (to constrain how many quake probabilities are actually calculated and save run-time)....

Hope I correctly guestimated what the program is doing....

Just my $.02 worth