Earth's Constant Hum Explained

MattSparkes writes "It has been known for some time that there is a constant hum that emanates from the Earth, which can be heard near 10 millihertz on a seismometer. The problem was that nobody knew what caused it. It has now been shown that it is caused by waves on the bottom of the sea, and more specifically 'by the combination of two waves of the same frequency travelling in opposite directions.'"

Re:10 millihertz

[FormOfActionBanana]

Actually, I think it works out to about 36 waves per hour.

10 milliHertz = 10 * 1/1000 waves per second
=> 0.01 waves per second
* 60 => 0.6 waves per minute
* 60 => 36 waves per hour

hertzs

[mapkinase]

10 milliHz is a beat every 100 seconds. Must be really tricky to detect. I wonder how far below that frequency the sensitivities of seismometers go.

http://en.wikipedia.org/wiki/Seismometer#Modern_re cording [wikipedia.org] mentiones only down to 1Hz. Need to see original article in Nature from work.

Re:Interesting, but wrong

[Prune]

This would have been funny except you seem to lack reading comprehension. The article said ten MILLIhertz! That's a single beat every hundred seconds. Bees' buzzing is about three orders of magnitude higher in frequency.

Re:10 millihertz

[Anonymous Coward]

Tomato, potato. 36 waves per hour is one wave per 100 seconds, or one cycle per hundred seconds or 100 seconds per cycle...

Re:hertzs

[mapkinase]

Ok, I skimmed briefly through Webb's article in Nature. Fig.3 shows the comparison between the model and experimental data. Coming from the computational biology field (3D structure prediction, gene modeling) I am somehow not convinced. To judge the matching of the model to the data one has to use a third more trivial approximation whatever it is.

For example, if you want to predict the quality of your prediction of protein 3D structure, you can assume very simple approach, like all the aminoacid residues on the surface of the protein are predicted less correctly than those inside. Or you can say: everything predicted in loops (those stretches of aminoacid chains that are not in regular local structure - alphahelix or betastrand) is bad, everything else is solid and then compare your quality prediction model threeways: to (a) actual quality and (b) to quality predicted by previously described simplistic models.

I do not see that here.

Re:Hmmmmmmmm

[ChrisA90278]

"Now tell us what causes the waves."

The answer is "wind" this has been known at least a century, no need to put it in the article. Now you ask what makes wind. This to has been known for a long time, the basic answer at the bottom of all of this is uneven heating of the atmosphere by solar radiation. Why "uneven"? The Earth is not uniform all over it's surface? Why is that? Something about plate tectonics? Why is that? The core is liquid and the "lighter" crust floats on the liquid while the liquid circulates. You can go on forever.....

But seriously, wind blowing over water causes ripples, the hight and period of the riples depends on the speed of the wind and the "fetch". Fetch being the distance the wind has to act on the water.

triangulation doesnt work for hum

[peter303]

A key breakthough was figuring out how to locate continuous signals. For normal earthquakes you have a sharp beginning. Using four or more seismographs you can invert for x,y,z and t0 (called triangulation).

For continuous signals you can find source by cross-corelating long pieces of signal from multiple locations. I first saw this in ambient noise submarine location, but the seismologists have now adopted it for analyzing some kinds of difficult signals like hum.