Hi Everyone,

I received the following from Dr. Randall Peters. With his permission I am forwarding 
it to the list.

Regards,
Larry Cochrane
Redwood City, PSN


-------- Original Message --------
Subject: hurricane Dennis study
Date: Sun, 17 Jul 2005 12:11:05 -0400
From: Randall Peters 
To: lcochrane@..............

Hi Larry.
    I have posted a paper to our server here at Mercer University Physics
that might be interesting to the amateur seismology community.  It is
titled, "Correlation measurements of atmospheric pressure variations and
seismicity during hurricane Dennis", online at
http://physics.mercer.edu/hpage/corr.html
   The seismometer used for this study is an old Sprengnether that was
once part of the WWSSN.  It was a gift from emeritus Prof. Deskin
Shurbet of Texas Tech University (my academic home before coming to
Mercer).  I've noticed some discussions on psn concerning the viability
of old surplused instruments like this, which use the LaCoste
zero-length spring.  My opinion is that they can be made to function
very well indeed, with a modernized sensor.  In my case, the instrument
works with an array form of the SDC sensor.  As opposed to the
gap-varying mode of conventional commercial seismometers used by
professionals, the area-varying sensor has many advantages.  For
example, I am presently operating without any force feedback, although
for the previous hurricane Charley study
http://arxiv.org/html/physics/0506162
I used the original coil/magnet (sensor/damper) subsystem as an actuator
to provide 'weak' force feedback.  The feedback involved only a
long-period opamp integrator.  The use of total force-balance (PID, or
proportional integral derivative network) is a cause for serious
performance degradation at low frequencies.  I would like to encourage
the amateur community to get involved with me in studying the mHz
frequency regime (where the earth hums), by opening up instrument
performance at low frequencies.  To do so requires a paradigm shift, and
I suspect that the amateurs are more open to this than are the
professionals.  One of the first things necessary is to depart (heresy
of heresies) from insisting on velocity detection.  To take the
derivative is to ruin low frequency sensitivity, since as every student
that I teach knows-- it pulls out an 'omega' by the chain rule of
calculus--causing a 20 dB/decade falloff below the lower frequency
cutoff.
     If you look at the unfiltered earthquake record in the above
referenced paper concerning hurricane Dennis, you will see that the
output from the sensor is displacement, rather than velocity.  For
anybody insisting on 'doing business as usual', for better identifying P
and S body waves; then it is a simple matter to do post-processor
numerical differentiation of the filtered waveform.  Incidently, for
those who may be interested, I've posted a paper that shows how to do
filtering with excel:
http://physics.mercer.edu/hpage/filter/gauss.html
    The beauty of better low frequency performance is that you can study
the long-period Rayleigh and Love waves also with the same instrument.
And although many are suspicious of my claims, I am confident that my
Sprengnether is picking up eigenmodes (free oscillations of the earth).
  Many of my own successes have been the indirect consequence of the
amateur community.  For example, were it not for John Lahr and his
interaction with a broad group of individuals to include the amateurs, I
would not have learned of the Dataq software that I have found to be so
powerful.  Also, Chris Chapman, who contributes regularly to psn
discussions is a friend whose expertise I admire and whose suggestions
(and corrections) directed toward various parts of my work have been
greatly appreciated.  Additionally, Allan Coleman has just built an
interesting broadband vertical seismometer using a gap-varying form of
my SDC sensor with a Willmore spring.
http://physics.mercer.edu/petepag/mkxx.pdf
Bottom line--don't ever underestimate the significance of what you've
about!
    I have been reluctant to get involved with psn before now for reasons
expressed in one of Chris' writings sometime back.  He noted that
academics like myself get caught up so thoroughly in bureaucratic issues
that they barely have time to do the research that is expected of them.
    In the event that I get more feedback (emails) from this
communication than I can completely respond to personally , please don't
be offended.  I despise the treatment of 'kill by silence' that has been
directed at me because of my unconventional ideas.
     Perhaps folks will be interested to know about the 'affordable
earthquake detector' that I am trying with some partners to bring to
market.  There was an article about this in Popular Science (April
issue, page 104).  I have developed a truly inexpensive autozeroeing
electronics package to work with the SDC sensors.  Used it, in fact, to
operate the pressure sensor in the Dennis study.   The board has only
three chips, each of which is less than $1 apiece.  Doubt that this
scheme, which works with diodes, is able to do the low frequency
measurements I've mentioned; but it certainly would be an interesting
possibility for a bare-bones vertical instrument with outstanding
performance.  Maybe there are some individuals who would like to try and
build something using this electronics with my compound vertical
seismometer, described at
http://www.iris.edu/stations/seisWorkshop04/PDF/CompoundVerticalSeismometer.pdf

     Would like to finally mention that Analog Devices has just marketed
a 24-bit ADC that should work with my sensor directly and which sells
for less than $10.
http://www.analog.com/en/prod/0,2877,AD7745,00.html (Tim Long at Georgia
Tech, just-retired seismologist alerted me to this chip; Tim is
interested in putting cheap but functional instruments in the hands of
science teachers; a great idea for which I hope we all can help!)
Only problem is that this chip requires a microcontroller to work.
Anybody with dsp experience interested in developing the necessary
software tools to make this happen with a cheap microcontroller?  I can
envision something that might make the professionals take notice even if
the price were a hundred times greater!!
    Keep up the good work,
Randall Peters, PhD
Professor and Chairman
Dept. of Physics
Mercer University
Macon, Georgia 31207


__________________________________________________________

Public Seismic Network Mailing List (PSN-L)