Register to reply

Seismic T-waves

by davenn
Tags: seismic, twaves
Share this thread:
davenn
#1
Dec18-13, 04:57 AM
Sci Advisor
PF Gold
davenn's Avatar
P: 2,599
Greetings all

Over the last couple of days I have been delving into a new personal discovery about seismic waves. And with the talk recently in several threads in the general and classic physics sections about sound wave propagation in water, I thought this was appropriate to share.

Many of us will be familiar with the more well known P, S and Surface waves generated by an earthquake. One that up until a couple of days ago that I was unaware of was the T wave.
On the 16th Dec 2013 there was a Mb 6.2 ( Mw 5.8) earthquake ~ 30 km offshore of the SW tip of the South Island of New Zealand.

Tectonics of the Region
The regional tectonic setting is a steeply dipping subduction zone boundary between the dipping Indo-Australian Plate and the over-riding Pacific Plate. This is part of the main plate boundary between the Indo-Australian and the Pacific Plate. Further north and running down the western side of the South Island is the oblique slipping Alpine Fault> At its southern end, in the Fiordland region, it transforms into the steeply dipping subduction zone which continues further south to Macquarie Island and beyond. The Fiordland Subduction Zone, has regular large earthquakes ands averages a M7+ event ~ every 15 years. The last large event was a M 7.8 in 2009 ~ 80 km to the north of Monday's M 6.2 event.

OK, enough background lets get to the fun stuff

On the long period seismometer the M 6.2 looked like any other quake of that size at that distance ( ~ 1850 km) from my Sydney, Australia location.



This sensor is bandpassed between 0.05 Hz and 0.1 Hz, the 10 to 50 second period is what I'm primarily interested in, tho there are some really interesting very long period, ~ 100 second, earth modes that are seen after the really large events. The earth rings ( resonates) like a bell for many many hours after the big quakes.

I happened to look at my short period sensor channels. These are 4.5 Hz geophones buried in the ground under the house. These geophones are primarily for recording local and regional events around eastern Australia. The phones recorded the P arrival and a little bit of the S arrival and as to be expected none of the of the low frequency surface waves.
The suprise was the huge burst of signal ~ 18 minutes after the arrival of the P wave on the geophone that didnt show up on the long period trace as shown above. Initially I thought that I had recorded a local event at the same time as the New Zealand event. Looking around some of the other online seismograms from around eastern Australia showed that this burst of signal was evident on them as well.
It was one of my fellow amateur seismo friends that pointed out that this was possibly a T wave phase. See the seismogram below from the geophone....



As can be seen, the T wave amplitude was very large compared to the P wave arrival.

NOTE the difference in recording time scale between the 2 grams

I have included a link to a paper on T wave generation. For those that dont want to read too deeply. What I have learned so far is that the T wave is generated at the water / rock boundary of the seafloor. The seismic waves coming up to the seafloor from below, generate an acoustic wave in the water which then radiate outwards from that point. In the case of the signal I recorded in Australia, the T wave traversed the Tasman Sea to where it encountered the edge of the Australian Continental Shelf. At this point some of that impinging wave was converted back into a seismic wave as it entered the sea / shelf interface and then it continued on to be recorded by my and other sensors on land in Australia.

I hope others find this as interesting as I have.

cheers
Dave

http://www.earth.northwestern.edu/do...ons/EAO169.pdf
Attached Thumbnails
131216.120708.zhi.gif   131216.120806.sydz.gif  
Phys.Org News Partner Earth sciences news on Phys.org
billiards
#2
Dec22-13, 02:07 AM
P: 747
Hi Dave, awesome observation!

So based on the little reading I did, I guess the quake would likely have been a "fast" one? (which I guess has something to do with a high acceleration of ground motion at source)

I first became aware of the so-called "T-wave" at this year's fall AGU conference where I noticed a few presentations about them. They seem to be picking up in fashion. Perhaps it is worth clarifying that T-waves are basically P-waves (compressional waves), except they take a special route through the ocean which makes them (a) less attenuated, and (b) slower moving (and thus later arriving) than "normal" P-waves.

So what can you learn about the Earth/ocean from your cool observation?
davenn
#3
Dec23-13, 02:18 AM
Sci Advisor
PF Gold
davenn's Avatar
P: 2,599
hi Billiards

Thanks for responding

Since posting the OP, I have discovered that one of my fellow seismic recording guys also recorded a T wave from a similar sized event ( ~ 0.5 of a magnitude less) from the same area back in 2009.

What I would like to do is be able to work out some approximate velocities.
1) There is ~ 1750 km of ocean path before the acoustic wave impinges on the Australia Continental Shelf

What speed did it travel through the ocean ... taking into account, ocean depth, and temperature ? ---- there are some articles that deal with sound waves at depth

2) Then there is a further ~ 100 km from the shelf edge to my recorder, through which the seismic wave would have been travelling as a p wave again

What speed did it travel through the continental crust to me from the shelf edge ?
normal P wave speed ~ 7km / sec or much slower ?

3) why is the T wave amplitude so large compared to the P wave

Unknown -- how far did the original P wave travel before impinging on the seafloor and producing an acoustic wave? Was it pretty much at the epicentre ? or some distance out from there ?

Now that I am aware of these T waves, I want to see if I record them from other under sea events around the South Pacific, or if it's something unique to quakes offshore of southern New Zealand ( maybe some set of conditions that allow it to occur)?

lots of questions and hopefully over time I will be able to answer some of them

cheers
Dave

davenn
#4
Dec23-13, 02:45 AM
Sci Advisor
PF Gold
davenn's Avatar
P: 2,599
Seismic T-waves

An FFT of the P waves show they are peaking between 0.7 and 2 Hz, centred on 1 Hz

An FFT of the T waves show they are peaking between 1 and 3 Hz, centred on 2 Hz

Interestingly not a lot of difference between them, I was expecting a bigger variation

Dave
billiards
#5
Dec23-13, 08:51 AM
P: 747
Quote Quote by davenn View Post

3) why is the T wave amplitude so large compared to the P wave
I would hazard a guess that it has to do with geometrical spreading. The normal P wave propagates with a spherical wavefront and so loses energy in proportion to 1/R2. Whereas the T wave is trapped in a layer and so propagates out cylindrically losing energy in proportion to 1/R.

An FFT of the P waves show they are peaking between 0.7 and 2 Hz, centred on 1 Hz

An FFT of the T waves show they are peaking between 1 and 3 Hz, centred on 2 Hz
Presumably low frequency energy does not get trapped in the SOFAR layer. The dominant frequency of the T-wave probably tells you something about the thickness of the SOFAR channel. I assume you have a broadband sensor with a flat response?
Redbelly98
#6
Jan5-14, 10:49 AM
Mentor
Redbelly98's Avatar
P: 12,069
Hey, this T waves stuff is pretty interesting. I just did a preliminary search, and have some basic questions.

From the first paragraph at this link, I get that there is a minimum in the sound wave speed at ocean depths of around 1 km, causing a waveguide effect:

Quote Quote by (see above link) Emile A. Okal, Northwestern University, USA
T phases are defined as seismic recordings of signals having traveled an extended path as acoustic waves in the water body of the oceans. This is made possible by the "Sound Fixing and Ranging" (SOFAR) channel, a layer of minimum sound velocity acting as a wave guide at average depths of 1,000 m"
My questions are:

1. Does anybody know by how much the speed of sound in the SOFAR layer varies? Are we talking at the 1% level? ppm level? Other?

2. Is there some reasonably defined, typical thickness for the SOFAR layer?

Or, if I could see a typical profile of sound velocity vs. ocean depth, that may pretty much answer both questions.
AlephZero
#7
Jan5-14, 01:54 PM
Engineering
Sci Advisor
HW Helper
Thanks
P: 7,115
Quote Quote by Redbelly98 View Post
1. Does anybody know by how much the speed of sound in the SOFAR layer varies? Are we talking at the 1% level? ppm level? Other?

2. Is there some reasonably defined, typical thickness for the SOFAR layer?

Or, if I could see a typical profile of sound velocity vs. ocean depth, that may pretty much answer both questions.
Is this good enough? http://en.wikipedia.org/wiki/SOFAR_channel
Redbelly98
#8
Jan6-14, 06:27 PM
Mentor
Redbelly98's Avatar
P: 12,069
Quote Quote by AlephZero View Post
Yes! Thanks. So basically the sound speed varies by a few percent.

Next time I watch Hunt For Red October, I'll have to listen for whether this reference in the book made it into the movie:
Quote Quote by wiki page linked to above
The novel The Hunt for Red October describes the use of the SOFAR channel in submarine detection.


Register to reply

Related Discussions
Seismic Activity Project Please help me! General Physics 4
Seismic Prospecting General Physics 6
Seismic waves General Physics 7
Difference between GPR from seismic General Physics 1
Seismic Stars... Astronomy & Astrophysics 0