EBEL, JOHN E. and ALAN L. KAFKA, Weston Observatory, Boston College, 381 Concord Rd., Weston, MA 02493, USA,,

An analysis of the seismicity of New England and nearby areas from 1975 to 1998 indicates that some aspects of the network seismicity of the region behave in a non-Poissonian manner. Specifically, the rate of occurrence of two felt earthquakes within 10 days of each other somewhere in the New England region is higher than expected from a Poisson distribution based on the mean rate of earthquakes of magnitude 2.7 and above. Furthermore, chances of a felt earthquake somewhere in New England within a week or so of an M*3.5 event in the region is enhanced still further. Analyses suggest that this increased activity rate within a week or so of an M*2.7 event is probably not due to aftershocks in the earthquake catalog. For anywhere within New England, the Poisson probability of a felt earthquake of M*2.7 is 11% during any 7-day period based on the rate of earthquakes observed by the network since 1975. However, in the network catalog we have observed that 22% of the M*2.7 events are followed by another M*2.7 event somewhere in New England within 7 days. Furthermore, 35% of the M*3.5 events in the network catalog are followed within 7 days by an M*2.7 event somewhere in New England. In the 1975-1998 dataset there are observed 13 temporal "clusters" of seismicity, defined as 4 or more events that all occur at less than the mean earthquake repeat time from one event to the next. The typical cluster averages 2-3 months in duration. While these clusters encompass only 16% of the time duration of the catalog analyzed, 60% of the M*3.5 earthquakes in the dataset took place in the clusters. Long periods of quiescence with no felt earthquakes are followed by almost no M*3.5 events. These observations suggest that the chances of a more widely felt earthquake in the New England region may be enhanced when the region as a whole is more seismically active.