Prior  Click here
Next   Click here
Equally dismaying. To
be honest, the
expectation was that
autocorrelation within
the clusters would
increase the correlation
score. We tried
including events from
Richter 4.5 to 4.9. It
does NOT make a
dramatic difference. For
Japan in 2011 (so far)
we can safely say if
aftershocks have
meaning then some
other measure(s) are
The statistical effects we had hoped to see in the recent Japanese events were obtained
when clustering Chilean events of 2010 and Indonesian events of very late 2004 and
2005. Unfortunately, super-quakes in the very Richter 8s and 9s are very rare. It is
certainly tempting to remark that the Japanese events are untypical of the last century
or so. Small sample size myopia has a tendency to distort analysis of recent events.