Re microseisms from hurricane Dennis:

Here at St. Louis the level of the 6-second microseisms as recorded
from the STM8 broadband in the basement had increased by a factor of 10 over
the previous week. THe peak level from the VBB was about 20 millivolts.
With a broadband sensitivity of 5.29 mv/micron/second, this indicates
a peak velocity of 3.4 microns/second. At a period T of 6 seconds, 
the angular frequency is w =2*pi/T, or 1.05*sec^-1. The velocity is
divided by w to give a peak displacement of 3.6 microns.

Here are some notes I have previously posted:

Re: 6-second microseisms and hurricanes: 

An interesting phenomenon in the eastern USA is that when a large
wind storm due to a deep barometric low occurs off the east coast, 
particularly off of New England, the wave energy pounding on the 
continental shelf propagates through the lithosphere under the 
eastern USA as a 6-second surface wave, at about half the period of
the oceanic wave. This is also true of any storm off the east 
coast, except that the period will be different, varying from 4 to
10 seconds.  Storms off the southwest coast cause a longer storm wave,
but the predominant energy is still from 5 to 8 seconds. 

This storm microseism noise is the major reason that the original 
global seismograph stations were configured as two separate instruments:
a SP or short-period, peaking in gain at about 1 second, at 50k to 100k,
and the LP, or long-period, peaking at 15 to 30 seconds, with magnifications
from 1.5k to 6k. During a microseism storm, even the LP record could be
a scribbled mess; usually the storm peak would only last for a several
hours, so we wouldn't interrupt the recording.

With the advent of wide dynamic range digital recording and VBB response
instruments, we deal with the microseisms in later digital processing.
However, for a visible monitor record on a drum recorder, we usually
use a "twin-T" notch filter to reduce the microseisms by 40db (1/100).
The twin-T is passive and easy to make with 3 resistors and 4 capacitors.
Even with the filter, a strong storm can fill the record with noise, which
is easily seen as 10 waves per minute.

A historical note on tracking hurricanes before the use of
satellites and aircraft:  small seismograph arrays were used.
At our old station FLO outside St. Louis, two additional small
vaults were built 1/4 mile west and north of the main vault.
THen intermediate period (10 second) seismometers were used,
connected to 6-second galvanometers for photo recording with a 
peak response at 6-seconds.
THe relative phase of the microseisms crossing the L shaped
array  provided a vector pointing toward the origin if the waves.
It was much more accurate than 3-component particle motion analysis. 
With several such stations (one was at Spring Hill, AL, another at
Rochester, NY) providing a pointer, the location of the hurricane
could be estimated, even when hundreds of km from shore.

There are some examples of the spectral noise from microseisms on
my web page featuring figures and data; there is even an example of
a storm peak from a hurricane last year.
http://www.eas.slu.edu/People/STMorrissey/index.html

Regards,
Sean-Thomas

_____________________________________________________________________

Public Seismic Network Mailing List (PSN-L)