Charles,

I have seen such arrangements for exact repositioning in kinematic
mounts for telescope objective instruments and some differential
GPS monuments. But I was being a bit hyperbolic in suggesting that
a quake would slide an LP seis (but speaking from experience in noting
that > I < might shove it off the glass plates). We had an Mb 8.6 at Adak
in 1986, 30 km from the main vault that indicated almost half of g
on the seismoscope; the LP seismometers slid a couple of mm, as indicated
by tracks in the dust on the glass plates, and remained functional but
tilted off center by several mm.

Regarding the carpet discussion:
I would not follow the advice that an LP horizontal with sufficient
sensitivity can be operated on top of a carpet, especially a horizontal
with its very high tilt sensitivity. By "sufficient sensitivity" I mean
the ability to routinely record the 6-second microseism background with
at least a 10:1 signal to noise ratio. These typically run from 2 to 10
(during a storm) microns peak-to-peak. Magnitude formulas want the wave
amplitude entered in milli-microns (nanometers), so smaller events have
to be read over the mean of the microseisms.  It is not a matter of spectral
sensitivity, since all the seismic waves of interest are much larger than
the instrument (a 1 second surface wave is 2+ km long). But the P-wave
of a M 5.7 near Mexico city is only a few times larger in amplitude than
the microseisms at St. Louis, and wouldn't be seen on a less sensitive
system. If I don't see the microseisms I try to find out what went wrong.
Seismologists have not gone to all the effort of piers, vaults, proximity 
to bedrock, etc, because they have too much money (fat chance), but to
reduce the noise of the instrument environment.

Regards,
Sean-Thomas
__________________________________________________________

Public Seismic Network Mailing List (PSN-L)