Heart of Earth's inner core revealed | Hacker News Reader

Heart of Earth's inner core revealed

bbc.co.uk

As for what the east-west axis means... According to the paper, they use a cylindrical anisotropy model and the best-fit axis is located at (9° N, 89° W) or its antipode at (9° S, 91° E). In other words, the crystals point along an axis near the equatorial plane through Central America and Southeast Asia.

The paper is pretty vague about why this would occur: "may represent a tectonic evolution of the inner core and thus our finding may indeed offer ‘clues about the inner-core history: its age, thermal process, and possibly an early convective event’."

Paper is available at http://www.nature.com/ngeo/journal/vaop/ncurrent/full/ngeo23...

But the core doesn't rotate at the same rate as the crust: http://news.harvard.edu/gazette/1996/08.15/PuttingaNewSpin.h... (or has this been proven wrong?)

That's very interesting. If the inner core rotates 3 degrees a year and the earthquake data covers 20 years, that would be a 60 degree shift from start to end of the data set. I looked over the paper again, and they don't mention this at all.

Edit: I emailed the author of the paper (Xiaodong Song). He says that more recent work suggests 0.3-0.5 deg/year for core rotation. (This would make the rotation much less important.)

http://www.geology.illinois.edu/people/xsong/Sites/papers/zh... http://www.geology.illinois.edu/people/xsong/Sites/papers/li...

What does it mean for the crystals of the 'inner inner core' to be aligned in an 'east-west direction'? Its a sphere. Does that mean they are aligned in a ring, like the grooves in an lp (or the dots on a CD/DVD)? Why doesn't the graphic reflect that?

Iron crystallizes at high pressure, while the iron is molten it aligns itself with any magnetic field, then as it crystallizes it can no longer change direction as the field changes.

This paper suggests that as the inner part of the core was solidifying, the earth had east/west magnetic poles rather than north/south magnetic poles. And then something "big" happened to switch the poles north/south. That "big" thing happened in a fairly short period of time (geologically) if the boundry between east/west alignment and north/south alignment in the crystals is thin. They also mention the core is solidifying at a rate of .5mm per year. Which is how they hope to "date" the event that occurred. Although it is unclear to me if they consider the process to be dependent on the surface volume or not.

Does that help?

JoeAltmaier is looking for a proper description of the alignment. Like, were the fields aligned like a donut, or were they aligned from longitude 0" to longitude 180", or some other vector along the xy plane (assuming that z is north/south)?

//edit: kens' top-level comment found the answer!

Not a bit. There's no such thing as an east pole or west pole, like there is a north pole.

I think the implication is that magnetic North was on the equator, rather than at the geologic pole.

exactly this. If you had a compass and were on the planet while those crystals were forming, the 'north' part of the compass would point you to a point along the equator rather than in the upper latitudes as it does now.

From what I understand, they mean that the "east" and "west" poles were, at that time, north and south. The something big flipped the earth so the solidified iron was east west, and then the rest of the core started solidifying to it's current north/south direction.

They aren't calling it a east/west pole, but that the north/south of that magnetic sphere are unnaturally aligned as so they are east/west relative to what we understand as north/south.

They mean that there's a preferred lineation to the individual grains. Consider cutting a sphere out of something similar to this: http://minerva.union.edu/hollochk/c_physicalgeology/images/L...

You can imagine making a sphere by packing together a bunch of small needles. Not all of them need to be oriented the same way, but they can align under the right conditions.

Because the needles are small in comparison to the size of the sphere, that alignment can change depending on where you are in the sphere.

In this case, they're hypothesizing that the "needles" are basically oriented differently at the outside of the sphere than near the center. More specifically, that they're oriented north-south at the outside of Earth's inner core and "east-west" near the center of Earth's inner core.

This is based on the observation that there's a strong anisotropy to the seismic velocities in the innermost parts of the Earth's core. The easiest way to explain any sort of seismic anisotropy this is through alignment of mineral grains (In many minerals sound travels faster along one axis of the crystal structure than along the other). However, this is one of the first times anyone's been able to clearly resolve differences in aniostropy with depth into the inner core.

The question is, if you cut the Earth through the equatorial plane and looked down at the exposed core, would the "east-west" grains follow circles around the plane, or would they form lines pointing across the plane?

Lines pointing across the plane. Otherwise there wouldn't be a fast axis.

Still not registering. There's no 'east pole' like there is a north pole. I get it, that the grains are in the plane of latitude instead of longitude. That still doesn't say which way they are aligned in that plane - randomly? Circularly? Radially?

Sorry, I should have been more clear. They have to be aligned linearly parallel to a line going from Central America to Southeast Asia.

If they were aligned any other way, you wouldn't have a fast axis in any particular direction.

The crystals are oriented due to the presence of a magnetic field. The different orientation indicates that the magnetic field has changed, but it's difficult to imagine a possible magnetic field inside an iron mass that wouldn't be essentially polar, unlike the ring alignment you describe. We know that the magnetic poles move around relative to the crust, and we know they occasionally flip. It's not inconceivable that at one time there was a large movement of the poles relative to the crust.

9N 89W is off the coast of Costa Rica, 9S 91E is off the coast of Singapore, which both locations seem to be kind of close to areas prone to earthquakes and tsunamis.

Also, it says it could have changed half a billion years ago. The current estimate of the formation of the moon is around 4.5 billion years ago, maybe these are related and they got their dating wrong?

The only thing I can find that happened half a billion years ago was the end of the the Neoproterozoic Era (the end of the Proterozoic Eon), which saw the first multi-celled life and the end of the Marinoan galciation, which could have subsided through global release of methane from equatorial permafrost.

Reply on news.ycombinator.com