Roger,

A sensible next step is to compare the true pattern with a pattern generated from the catalog after randomizing the times. This would ensure that any peaks don't just come from, for example, the distance between Tonga seismicity and Indonesian seismicity.

John