PSN-L Email List Message

Subject: Re: Why we like verticals - a rethink
From: chrisatupw@.......
Date: Mon, 14 Oct 2013 05:58:17 -0400 (EDT)


From: Brett Nordgren 
To: psnlist 
Sent: Wed, 9 Oct 2013 22:24
Subject: Re: Why we like verticals

    From what I can estimate now, the entire trace represented
about one hour.
See http://bnordgren.org/seismo/T120ZNE.JPG

Dave later reported that, contrary to what I'd thought, it was not
unusually windy at the time and the weather was quite typical with
microseisms relatively low.  I can see similar traces any time I
look.  Also the vault was relatively stable at the time with nothing
out of the ordinary, and bear in mind that the three traces are
derived from three identical Galperin-oriented sensors, so you would
expect that many possible noise sources would affect all three axes
in a similar way.  The motion we were seeing was truly either
horizontal or tilting of the ground.  I am currently suspecting that
tilt noise is mainly due to large-scale atmospheric pressure changes
deforming the ground, though I think it's a subject that's still being
studied.
****Lets consider both what is in the traces and what is not there.
The vertical trace is almost dead flat with only the quake signal on
it - there seems to be NO evidence of the very large LONG period
signals SEEN on the horizontal traces - so the three sensors and
the electronics all seem to be working fine. I conclude that the
seismometer is NOT moving vertically under long period noise to
any significant extent.
    Do you observe any repeatable time differences in the noise
amplitude over 24 hours, please ?
    There is NORMALLY about a 20 dB difference in the longer period
noise between the horizontal and the vertical traces =3D ~x10. But if you
reduce the amplitude of the horizontal signals by 10, you would still
expect to see appreciable signals on the vertical. Something does not
quite seem to 'add up' ! I would not expect this result from simple air
pressure action on the earth - I would expect to see equally large or
even larger vertical signals.
    While the circuits resolving the three components seems to be
working excellently, would it be worth checking on the output amplifier
gains for the three resolved signals ?
    What is the soil / rock under Dave's seismometers, please ? One
might expect signals like these if the seismometer was mounted near
the edge of a large concrete slab sitting on clay / highly compacted
soil / water saturated soil with a void underneath. This could react
strongly to changes in air pressure and give large tilt signals. But I
don't see any correlating pressure changes on the barograph trace ?
I note that at least one noise trace seems to be a differentiated 5
minute signal.
    Which leaves the question as to what / how the almost purely
horizontal signals could be generated ? How close is Dave to the
Pacific Ocean ? I was wondering if he could observe any ocean
surges / wave / tidal noise ? I don't know what these would look like
on a seismometer trace ? Do you have a time plot of the temperature
and solar intensity variations, please ?
    What is the structure / composition of the ground vertically under
Dave's seismometers, please ?

I uploaded a 1Hr barometric pressure trace recorded at the same time
as the ZNE traces.  At the moment I don't have the sensitivity value
available, but it is extremely high in that recording.  At times we
have seen 10-minute pressure oscillations that could be 10x what you
see there, possibly greater.  The barograph has flat response down to
about 3600 seconds.  I don't see any obvious correlations.
 http://bnordgren.org/seismo/131106Baro.JPG
****Barographs usually have responses down to DC, not 3,600 seconds.
Are yours different ?

Microseisms, which appear to have been small at the time, had already
been removed with a 17-sec low-pass filter, and the instrument
response is otherwise flat down to 120 seconds.

     Regards,

     Chris Chapman






__________________________________________________________

Public Seismic Network Mailing List (PSNLIST)

To leave this list email PSNLIST-REQUEST@.............. with 
the body of the message (first line only): unsubscribe
See http://www.seismicnet.com/maillist.html for more information.

[ Top ] [ Back ] [ Home Page ]