I'd suggest the next step would be to check maybe 20 of the closest pairs in space and time to see if there is an obvious connection explaining the surplus of events that we've forgotten.