PSN-L Email List Message
Subject: Re: Detecting micro-earthquakes
From: ChrisAtUpw@.......
Date: Fri, 10 Feb 2006 13:28:54 EST
In a message dated 10/02/2006, rsparks@.......... writes:
While I have become familiar with my own seismometer and capabilities, I am
looking at a different world when looking at your files. First is the 200
samples per second which provides resolution of events as high as 100 Hz. That
is very good.
Hi Jon,
Your amplifiers have a cut-off at ~20 Hz. The minimum sample rate is 40
sps. 200 sps is not necessary and could use a lot of memory. Can you use 60
to 100 sps?
Next is the 4.5 Hz geophone which has a peak response at about 4.5 Hz. I
expect it to detect local quakes very well. The expectations for large distant
quakes are less.
It will not have a peaked response if it is correctly loaded. There
should be a ~2.7 K Ohm resistor soldered across the output terminals of each
geophone.
What surprised me was the large amount of energy found at 10 Hz, declining
up to about 20 Hz. I did not expect to see that when I applied the FFT
algorithm from Winquake to each file, and for the most part, to each subdivided time
period within each file. What I expected to see was the ocean wave noise at
0.2 Hz but at a much reduced peak compared to my own seismometer which has a
natural period of about 3 1/5 to 4 seconds.
You don't see microseisms at about 6 sec with a 4.5 Hz geophone. The
response will be down by a factor of over 700. If you are on the coast you will
likely see higher frequency noise associated with breaking waves.
I am curious to know if you can explain the 10 Hz energy source? Also ocean
waves? Maybe a nearby flag pole and flag waving in the wind?
The Strouhal number is between 0.2 and 0.4, depending on the 'bluntness'
of the object; S=fxD/v, where f is the vortex shedding frequency, D is the
width of the object and v is the velocity. Vortices are shed alternately from
either side of the object. The force on an object is at f/2 Hz, so a strong
20 m/sec, 76 kph wind will give D = 20 / 100, D ~ 0.2 m.
I suspect that the 10 Hz could be resonances in air ducts in a central
heating system? Maybe a fan? The velocity of sound is about 330 m/s, so 10 Hz
has a wavelength of about 33 m. Do you have any ducts or corridors about 33,
16.5 or 8.25 m long?
To see the response, choose a section of recording with no quake and
click on the FFT tab in the top tool bar. You will see a plot of the amplitude
against the period.
The second surprise was the large noise count, up in the +/- 300 unit range.
It is nothing to be alarmed about, only different from my system. I wonder
what the noise floor is when the seismometer is disconnected from the
electronics?
You need to substitute 390 Ohm metal film resistors for the geophones to do
this test. Can you buy electronic components easily?
Regards,
Chris Chapman
In a message dated 10/02/2006, rsparks@.......... writes:
<=
FONT=20
style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>While I=20
have become familiar with my own seismometer and capabilities, I am lookin=
g at=20
a different world when looking at your files. First is the 200 sample=
s=20
per second which provides resolution of events as high as 100 Hz. Tha=
t is=20
very good.
Hi Jon,
Your amplifiers have a cut-off at ~20 Hz. The=20
minimum sample rate is 40 sps. 200 sps is not necessary and could use a lot=20=
of=20
memory. Can you use 60 to 100 sps?
<=
FONT=20
style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>Next is=20
the 4.5 Hz geophone which has a peak response at about 4.5 Hz. I expe=
ct=20
it to detect local quakes very well. The expectations for large distant qu=
akes=20
are less.
It will not have a peaked response if it is=20
correctly loaded. There should be a ~2.7 K Ohm resistor soldered across the=20
output terminals of each geophone.
<=
FONT=20
style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>What=20
surprised me was the large amount of energy found at 10 Hz, declining=
up=20
to about 20 Hz. I did not expect to see that when I applied the FFT algori=
thm=20
from Winquake to each file, and for the most part, to each subdivided time=
=20
period within each file. What I expected to see was the ocean wave no=
ise=20
at 0.2 Hz but at a much reduced peak compared to my own seismometer which=20=
has=20
a natural period of about 3 1/5 to 4 seconds.
You don't see microseisms at about 6 sec with a=
4.5=20
Hz geophone. The response will be down by a factor of over 700. If you=20=
are=20
on the coast you will likely see higher frequency noise associated with brea=
king=20
waves.
<=
FONT=20
style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>I am=20
curious to know if you can explain the 10 Hz energy source? Also ocean wav=
es?=20
Maybe a nearby flag pole and flag waving in the wind?
The Strouhal number is between 0.2 and 0.4=
,=20
depending on the 'bluntness' of the object; S=3DfxD/v, where f is the vortex=
=20
shedding frequency, D is the width of the object and v is the velocity. Vort=
ices=20
are shed alternately from either side of the object. The force on an object=20=
is=20
at f/2 Hz, so a strong 20 m/sec, 76 kph wind will give D =3D 20 /=20=
100, D=20
~ 0.2 m.
I suspect that the 10 Hz could be resonanc=
es=20
in air ducts in a central heating system? Maybe a fan? The velocity of=20
sound is about 330 m/s, so 10 Hz has a wavelength of about 33 m. Do you have=
any=20
ducts or corridors about 33, 16.5 or 8.25 m long?
To see the response, choose a section of=20
recording with no quake and click on the FFT tab in the top t=
ool=20
bar. You will see a plot of the amplitude against the period.
<=
FONT=20
style=3D"BACKGROUND-COLOR: transparent" face=3DArial color=3D#000000 size=
=3D2>The=20
second surprise was the large noise count, up in the +/- 300 unit ran=
ge.=20
It is nothing to be alarmed about, only different from my system. I wonder=
=20
what the noise floor is when the seismometer is disconnected from the=20
electronics?
You need to substitute 390 Ohm metal film resis=
tors=20
for the geophones to do this test. Can you buy electronic components=20
easily?
Regards,
Chris Chapman
[ Top ]
[ Back ]
[ Home Page ]