1 March 2010
The Earth Really Can Ring Like A Bell
Posted by Dan Satterfield
I mentioned on air Saturday night that the quake in Chile may not have been felt in North America, but the ground definitely moved. The Earth Rang like a bell I said.
A friend of mine who trades emails with me frequently heard that, and sent me this story. Huntsville is a very high tech city as you will learn from reading this story:
********************************
Dan,
Heard your comments about the quake in Chile and about the
strength of those along the Andes.
Back in June of 1994 I was working at SCI and we were building
Cesium Atomic Clocks on a sub-contract with ITT. During the
month of June I was running a 30 day stability test on a new clock,
and there was an anomaly at the 18.87 minute point on the Alan
Variance chart that we could not explain.
Later I read an article in Science Weekly about an earthquake
in Bolivia that was 8.2 and was 300 miles deep that had caused
the earth to “ring like a bell”. And said the resonance of the
planet was about 0.0008 Hz. And that it caused the surface of
the earth to move up and down.
Finally, we connected the anomaly with the earthquake. Our clock
was designed to be frequency stable to one part in 10 to the 14th,
so it was sensitive enough to have its frequency altered by the
slight change in altitude as the “bell rang”.
Just thought this little tale might interest you.
Al Glenn
Dan Satterfield has worked as an on air meteorologist for 32 years in Oklahoma, Florida and Alabama. Forecasting weather is Dan's job, but all of Earth Science is his passion. This journal is where Dan writes about things he has too little time for on air. Dan blogs about peer-reviewed Earth science for Junior High level audiences and up.
I’m a retired scientist, age 73, and while my expertise is in chemistry and esr/nmr spectroscopy, I became interested in seismology after a 4.7 quake near Conway Arkansas shook the Ozark region (including Cotter, where I live). It made a noise, too, for about 30 seconds. That happened not quite a year ago. So I built a Lehman-type seismometer, and it’s now in my basement. Last Thursday I finally got it interfaced to the laptop that I’m typing on right now. The instrument uses a light beam and knife edge detector to pick up slight movements of the inertial mass. I did something unorthodox when I designed the circuitry: I thought that getting a voltage to this computer from the basement was a bad idea, that it would introduce spurious signals (and probably would), so I mapped displacement voltage as frequency using a voltage controlled oscillator (which is inside the seismometer’s thermal enclosure). The seismometer is thus an FM design, and I’m using a Fourier transform program to create the time versus amplitude display and do periodic screen captures. That’s the other thing. The Fourier transform program is capable of going to very narrow bandwidths. It was set for 0.24 Hz resolution when it detected yesterday’s 5.8 quake in Costa Rica (I’ve only had the instrument in operation since last Wednesday, Feb 8, 2012). It is capable of going to much finer resolution, and during the night it recorded at 0.061 Hz resolution, which reveals a periodic baseline roll that really caught my attention (I’d noticed the slight roll in my earlier recordings at 0.24 Hz resolution). A rough reading of the roll interval is 19.8 minutes, which comes out to be 0.00084 Hz! That’s pretty close to the planet’s putative grand resonance, stated to be at 0.0008 Hz. Looks like there could be an even longer roll, perhaps tides? You know, I wasn’t thinking about sensitivity to long period seismic waves when I built this thing. The Lehman resonates at 1 Hz, because my interest was in the quake swarms that are happening fairly nearby, e.g., near Guy Arkansas and along the New Madrid fault system east of us. But as it turned out, I picked up long seismic waves from an event at 3:23 PM CST yesterday (Feb 13, 2012). That event also registered on the professional broadband seismometers at FCAR and MGMO (which bracket Cotter). I’m cut off from a research library. I need to know more about the 0.0008 Hz resonance, e.g., who developed the theory? What is the vertical amplitude in SI units? Is there any other evidence than the atomic clock anomaly? Has anyone else ever built an FM Fourier transform seismometer? Apparently they are inherently broadband. Surely I don’t have the only one in the world! But then again, when I characterized the vitamin-E free radical by esr spectroscopy (at Washington University, St. Louis MO in 1968) I assumed that it had been done before. I was wrong.
There is no mechanism for sending images. You need to see the recordings at 0.061 Hz resolution. JRW
Following up:
I sent this to two of my colleagues (one a physicist and the other an electronics engineer). Detecting distant quakes with a home made seismograph is getting to be routine! (It’s the kind of toys scientists play with when they retire and start turning senile! Second childhood stuff, etc.) I can tell you now that my seismometer (COTR) would have easily detected the Japanese quake that caused the horrendous tsunami. But the baseline roll has my attention right now because it reveals a wave with a frequency not far from 0.0008 Hz. I’m being ruthless with my own speculations here, but that kind of thing is necessary (or it isn’t science). This would make a whole lot more sense to the reader if there was a means to attach the images. To my colleagues:
“I’m a doubting Thomas, and I’m still doing controls to find out if there is a simple explanation for the slow baseline roll that I’m seeing (i.e., not based on movements of the inertial mass but something spurious). The detector circuit does not have any intentional time constants anywhere nearly that long. I’ve had the recording in a very slow time base/ultra high resolution setting for over a day, and it continues to pick up the roll, which is not actually sinusoidal but something more like a sawtooth. A sample is shown in the first attachment, and the horizontal width of that screen is 5029.1 s (83 minutes, 49.1 seconds). The vertical resolution is 0.061 Hz, and as you can see from the scale at the right, the displayed frequency range is really narrow. The time stamps that Spectran II records in its top margin refer to the right-hand edge of its display. The FFT processing is actually part of the seismic amplification.
As it turned out, there was a 6.0 quake off the Oregon coast last night, and both COTR and FCAR picked up the seismic waves from that event starting at close to 9:37 PM CST. See this link for the USGS data on that quake:
http://earthquake.usgs.gov/earthquakes/recenteqsww/Quakes/quakes_all.php
The second attachment is COTR’s recording of the quake, and it is way too compressed along the time axis to be a useful seismic plot (again, I’m primarily interested in the baseline roll right now,). That first event has a lot of short period waves, and as you can see, COTR is much more sensitive to those frequencies than FCAR. Note the ~1 Hz modulation lines above and below the baseline in that initial shock and its decay (1 Hz is the resonant frequency of my mass balance). This might be the P-wave (but I’m not sure). FCAR does a better job of recording the long seismic waves, but COTR didn’t miss them! Attachment three is the FCAR record of the Oregon event.
I’d like to hear your opinions. The roll could be some kind of stray body capacitance or leakage that somehow affects the VCO. It’s probably too much to jump on the planetary grand resonance thing right now.
Rick”