This is quite interesting.  Statistics will be necessary to make any claims of
course.  And doing that will also be tricky.  I am wondering if the precursor
signals always have the same form.  If so, then a waveform that models the
precursor can be "idealized" in a time series and then correlated with a long
time record of real data.  Any wave forms that are similar will then show up
in the resultant times series as spikes, with a value of 1.0 being an exact
correlation.  This would then have to be correlated with some function showing
the times of occurences of earthquakes for different parameters such as a time
window before the onset of a quake, magnitude, depth, etc..  The results could
then tell you what the statistical significance of the correlations are...this
is a messy business!

Good luck!

******************************************************************************
John Hernlund
Department of Geological Sciences
Arizona State University
E-mail: hernlund@.......
WWW: http://www.public.asu.edu/~hernlund/
******************************************************************************

__________________________________________________________

Public Seismic Network Mailing List (PSN-L)