Frequencies

[John Hernlund]

Hi,
I have just recently gotten my own seismograph running and had a few
questions about frequencies. What frequency range is the most usual for
seismic events? Where in the spectrum should I be looking?

******************************************************************************

John Hernlund
E-mail: hern...@asu.edu
WWW: http://www.mc.maricopa.edu/academic/phy_sci/Geology/hernlund/

******************************************************************************

[Susan Hough]

John Hernlund (hern...@imap3.asu.edu) wrote:
: Hi,

: I have just recently gotten my own seismograph running and had a few
: questions about frequencies. What frequency range is the most usual for
: seismic events? Where in the spectrum should I be looking?

Depends on the magnitude. Earthquakes radiate energy predominantly at
a frequency that is inversely proportional to the rupture length. For
very small earthquakes, the high frequencies may be effectively limited
by attenuation through the upper few 100 meters/few kilometers directly
under your site. That is, if you're looking to record M0-3 events,
you might tend to see mostly 10-30 Hz energy for everything. If your
seismometer is on very hard rock, you could see higher frequencies,
maybe up to 100 Hz for small events. For comparison, a M5 event might
have dominant energy at 1Hz; a teleseismic M8 event would give you
data at ~0.05 Hz (20 sec) and below.

Sue, speaking for myself

[see...@nospam.wolfenet.com]

Sue, in the recent quake(s) here in Seattle, I noticed a back
and forth motion about 2 hz, (2nd story of an apt) and was
4.9. (Two years ago a 5.1). The freq seems consistant,
and we commonly have subduction quakes here. Is this (freq)
typical given with what mentioned?

Reply to: see...@NOSPAM.wolfenet.com, and delete "NOSPAM".

[Susan Hough]

bitb...@wolfenet.com (see...@NOSPAM.wolfenet.com) wrote:

: On 2 Jul 1997, Susan Hough wrote:
:
: > John Hernlund (hern...@imap3.asu.edu) wrote:
: > : Hi,
: > : I have just recently gotten my own seismograph running and had a few
: > : questions about frequencies. What frequency range is the most usual for
: > : seismic events? Where in the spectrum should I be looking?
: >
: > Depends on the magnitude. Earthquakes radiate energy predominantly at
: > a frequency that is inversely proportional to the rupture length. For
: > very small earthquakes, the high frequencies may be effectively limited
: > by attenuation through the upper few 100 meters/few kilometers directly
: > under your site. That is, if you're looking to record M0-3 events,
: > you might tend to see mostly 10-30 Hz energy for everything. If your
: > seismometer is on very hard rock, you could see higher frequencies,
: > maybe up to 100 Hz for small events. For comparison, a M5 event might
: > have dominant energy at 1Hz; a teleseismic M8 event would give you
: > data at ~0.05 Hz (20 sec) and below.
:

: Sue, in the recent quake(s) here in Seattle, I noticed a back

: and forth motion about 2 hz, (2nd story of an apt) and was
: 4.9. (Two years ago a 5.1). The freq seems consistant,

2 Hz sounds about write for M5, but you also have building response t
consider. The rule of thumb is that buildings resonate at a period
of 0.1*number of stories, so that 20 stories=2 sec (=0.5 Hz). A 4 story
building, meanwhile, would be ~0.4 sec, or 2.5 Hz. Because of the
resonance effect, a building can have dominant motion at a frequency
that is different than the dominant frequency of shaking on the ground,
with the worst damage coming when the two correspond (e.g., the 20
story buildings in Mexico City back in '85).

: and we commonly have subduction quakes here. Is this (freq)

: typical given with what mentioned?

The Pac NW has the potential for huge subduction quakes with very
long period energy (plus plenty at higher frequency; earthquakes
radiate over spectrum). Most of the moderate earthquakes up there, at least
in recent years, have not been subduction events.

Sue, speaking for myself

[Kevin J. Miller]

In article <5prce9$9...@gap.cco.caltech.edu>,

Susan Hough <ho...@amelia.gps.caltech.edu> wrote:
>: > Depends on the magnitude. Earthquakes radiate energy predominantly at
>: > a frequency that is inversely proportional to the rupture length. For
>: > very small earthquakes, the high frequencies may be effectively limited
>: > by attenuation through the upper few 100 meters/few kilometers directly
>: > under your site. That is, if you're looking to record M0-3 events,
>: > you might tend to see mostly 10-30 Hz energy for everything. If your
>: > seismometer is on very hard rock, you could see higher frequencies,
>: > maybe up to 100 Hz for small events. For comparison, a M5 event might
>: > have dominant energy at 1Hz; a teleseismic M8 event would give you
>: > data at ~0.05 Hz (20 sec) and below.
>:

Does this say something about how faults rupture? I was thinking about the talk
a few years back (it may be still current)
about a "zipper" effect, where faults
rupture along a small zone that travels the length of the fault. If the
frequency range is related to fault length, does that imply that the whole
fault is involved in the rupture more or less simultaneously? Or does the
fault zone resonate somehow, even though only a small part of it is
actually shearing at any given window of time? Or maybe the fault line acts as
an antenna, with only frequencies related to fault length being
constructively propagated? Gee, maybe this is not so simple after all...

(Glad to actually hear something about seismology on this group...)
- Kevin

[Susan Hough]

Kevin J. Miller (kjmi...@alumni.caltech.edu) wrote:
: In article <5prce9$9...@gap.cco.caltech.edu>,

: Susan Hough <ho...@amelia.gps.caltech.edu> wrote:
: >: > Depends on the magnitude. Earthquakes radiate energy predominantly at
: >: > a frequency that is inversely proportional to the rupture length. For
: >: > very small earthquakes, the high frequencies may be effectively limited
: >: > by attenuation through the upper few 100 meters/few kilometers directly
: >: > under your site. That is, if you're looking to record M0-3 events,
: >: > you might tend to see mostly 10-30 Hz energy for everything. If your
: >: > seismometer is on very hard rock, you could see higher frequencies,
: >: > maybe up to 100 Hz for small events. For comparison, a M5 event might
: >: > have dominant energy at 1Hz; a teleseismic M8 event would give you
: >: > data at ~0.05 Hz (20 sec) and below.
:
: Does this say something about how faults rupture? I was thinking about the talk

Yes and no. At teleseismic distances, you get a lot of energy in surface
waves with a period near 20 seconds...this is more a function of the generic
velocity structure of the earth than details of the source. You can think
in terms of ringing a bell: you can hit it any way you want...you might not
always get a sound out, but, if you do, it'll be the same one.

: a few years back (it may be still current)

: about a "zipper" effect, where faults
: rupture along a small zone that travels the length of the fault. If the

How energy is released from very large earthquakes is still a matter of
some debate. You're right, that it's an important difference, whether or
not slip occurs within a narrow pulse or over the entire fault at once.
It seems that large earthquakes do rupture as pulses, but there just isn't
a lot of local/regional data for M8+ events, and it's not understood, for
example, how the length of the slip pulse depends (or not) on magnitude.

: constructively propagated? Gee, maybe this is not so simple after all...

Nope...small-to-moderate earthquakes are a lot easier to understand.

: (Glad to actually hear something about seismology on this group...)

We try to sneak something in once in a while...

Sue, speaking for myself

[see...@nospam.wolfenet.com]

On 7 Jul 1997, Susan Hough wrote:

> bitb...@wolfenet.com (see...@NOSPAM.wolfenet.com) wrote:
> : On 2 Jul 1997, Susan Hough wrote:
> :
> : > John Hernlund (hern...@imap3.asu.edu) wrote:
> : > : Hi,
> : > : I have just recently gotten my own seismograph running and had a few
> : > : questions about frequencies. What frequency range is the most usual for
> : > : seismic events? Where in the spectrum should I be looking?
> : >

> : > Depends on the magnitude. Earthquakes radiate energy predominantly at
> : > a frequency that is inversely proportional to the rupture length. For
> : > very small earthquakes, the high frequencies may be effectively limited
> : > by attenuation through the upper few 100 meters/few kilometers directly
> : > under your site. That is, if you're looking to record M0-3 events,
> : > you might tend to see mostly 10-30 Hz energy for everything. If your
> : > seismometer is on very hard rock, you could see higher frequencies,
> : > maybe up to 100 Hz for small events. For comparison, a M5 event might
> : > have dominant energy at 1Hz; a teleseismic M8 event would give you
> : > data at ~0.05 Hz (20 sec) and below.
> :

> : Sue, in the recent quake(s) here in Seattle, I noticed a back
> : and forth motion about 2 hz, (2nd story of an apt) and was
> : 4.9. (Two years ago a 5.1). The freq seems consistant,
>
> 2 Hz sounds about write for M5, but you also have building response t
> consider. The rule of thumb is that buildings resonate at a period
> of 0.1*number of stories, so that 20 stories=2 sec (=0.5 Hz). A 4 story
> building, meanwhile, would be ~0.4 sec, or 2.5 Hz. Because of the
> resonance effect, a building can have dominant motion at a frequency
> that is different than the dominant frequency of shaking on the ground,
> with the worst damage coming when the two correspond (e.g., the 20
> story buildings in Mexico City back in '85).
>
> : and we commonly have subduction quakes here. Is this (freq)
> : typical given with what mentioned?
>
> The Pac NW has the potential for huge subduction quakes with very
> long period energy (plus plenty at higher frequency; earthquakes
> radiate over spectrum). Most of the moderate earthquakes up there, at least
> in recent years, have not been subduction events.
>
> Sue, speaking for myself

I have always heard that virtually everything here is/had been
subduction quakes. The recent events have exposed some newer
unknown faults, notwithstanding not just 1 but THREE faults
now known to go RIGHT THROUGH SEATTLE east & west up to the
local Cascade Mtn range. All the motion in the recent stuff
though has been north - south motion, which if I am not wrong
are new faults, not existing ones.

I felt the two big recent ones easily, and thought I felt
something gently here in Seattle, during the Okanogan shake,
(250 mi away), and the Vancoouver Isl shake, (200 mi away).
The latter being very soft gentle detection.
(The time frames fit later, if I am not wrong). I DO see what
you mean about building ampflication. My apt is a 50 yr
old wooden structure, that actually has some "give",
as opposed to (as an example) brick structures in the area.

Remove "bitbucket" & "NOSPAM", to write to me.

[Bill Smith]

Sue, perhaps this has been addressed before, but you mention that
earthquakes radiate at a frequency that is inversely proportional
to the rupture length. Does the nature of the break have any
bearing on this? For example, does a fault start at a pinpoint,
or several points and break in a cascading increase, or does a
fault somehow almost spontaneously move along its entire rupture
length? Do wave recordings show evidence of one or both
phenominia, especially with respect to secondary fault releases?
I suspect the fault plane of a paticular event, or rather the
stress vector that caused the fault to break would have a great
effect on how a rupture would initiate and might lend itself
to description via analysis of wave recordings. Is this
discussed in the literature?

Thank you for your energy to contribute in this newsgroup!

73 de Bill, AB6MT
bils...@crl.com

Susan Hough wrote:
>
> John Hernlund (hern...@imap3.asu.edu) wrote:
> : Hi,
> : I have just recently gotten my own seismograph running and had a few
> : questions about frequencies. What frequency range is the most usual for
> : seismic events? Where in the spectrum should I be looking?
>
> Depends on the magnitude. Earthquakes radiate energy predominantly at
> a frequency that is inversely proportional to the rupture length. For
> very small earthquakes, the high frequencies may be effectively limited
> by attenuation through the upper few 100 meters/few kilometers directly
> under your site. That is, if you're looking to record M0-3 events,
> you might tend to see mostly 10-30 Hz energy for everything. If your
> seismometer is on very hard rock, you could see higher frequencies,
> maybe up to 100 Hz for small events. For comparison, a M5 event might
> have dominant energy at 1Hz; a teleseismic M8 event would give you
> data at ~0.05 Hz (20 sec) and below.
>

> Sue, speaking for myself

[Susan Hough]

Bill Smith (bils...@crl.com) wrote:
: Sue, perhaps this has been addressed before, but you mention that

: earthquakes radiate at a frequency that is inversely proportional
: to the rupture length. Does the nature of the break have any
: bearing on this? For example, does a fault start at a pinpoint,
: or several points and break in a cascading increase, or does a
: fault somehow almost spontaneously move along its entire rupture
: length? Do wave recordings show evidence of one or both

This was touched on in follow-up messages in this same thread. Yes,
there is an issue of what happens for large earthquakes, with 'large'
generally meaning big enough to rupture the full depth extent of
the seismogenic layer. Small earthquakes stop at a point and rupture
with a certain velocity...it isn't the entire fault slipping at once,
but the beginning doesn't stop slipping way ahead of the end. Big
earthquakes are different: recent results say they also start from
a point, but then rupture in fairly narrow pulses that travel the
length of the fault. If rupture is triggered on secondary faults,
that's yet another complication: slip tends to scale with rupture
length, so that the waveforms from two 'conjugate' 15-km faults
that ruptured in sequence would look different from a single 30-km
fault rupture.

: I suspect the fault plane of a paticular event, or rather the

: stress vector that caused the fault to break would have a great
: effect on how a rupture would initiate and might lend itself
: to description via analysis of wave recordings.

Earthquakes look surprisingly similar to each other, in a lot of
ways (once they get going). 'Stress drop' is the basic parameter
that differentiates the 'character' of an earthquake: whether it had
high or low slip & energy release for its rupture length. For big
earthquakes, stress drop can be reasonably well estimated...and
systematically seems to vary only between about 10 and maybe 80 bars.
There is compelling evidence for some higher stress drop small/moderate
events, but bigger earthquakes are usually (almost by definition) on
faults that well-enough developed to generate big earthquakes...if
you assume that it's fault structure/geometric complexity that governs
the character of a rupture, this fact provides something of a moderating
influence.

Sue, speaking for myself