Chris;

Granted. What we see is not probability of occurrence so much as probability of reporting.

One possible approach is to use a low-end cutoff based on the reporting threshold of the system being used. I'm currently using a 5 year chunk of the CNSS catalog which seems to go down somewhere below 3.

I know 5 years is too small but it's 323929 quakes globally and adding more makes it take too long to run.

Roger