Magnetic Poles & Earth's Rotation Alignment

[Vorde]

Why do the magnetic poles line up (relatively speaking) with the poles of Earth's rotation?

 

[Simon Bridge]

They don't - they are the other way around. xD
Do you know how the geomagnetic field is generated?

 

[Vorde]

Just roughly, I believe as the gross movement of liquid metals in the inner core.

 

[Simon Bridge]

geodynamo
OK - simplistically: the rotation of the core causes the field - imagine a system where the core rotates on an axis that is not the same as that for the overall rotation of the Earth... how would this happen?

Basically, the flows creating the dynamo effect are coupled to the Earth's rotation.

 

[Vorde]

Ok, I can see that. Thank you.

 

[geo101]

OK - simplistically: the rotation of the core causes the field - imagine a system where the core rotates on an axis that is not the same as that for the overall rotation of the Earth... how would this happen?

Basically, the flows creating the dynamo effect are coupled to the Earth's rotation.

One important consideration is that at any given time (i.e., an instantaneous snapshot) the rotation and magnetic poles do not align. Take the present day field, for example, the poles do not align. If, however, you average over a period of time then the average magnetic poles will align with the rotation poles. How much time to average over depends on the availability and distribution (in space and time) of the data, but when all things are good several thousand years is enough time.

 

[Simon Bridge]

It would be pretty unusual for the angle between the rotational axis and the magnetic dipole to be close to 90deg though. It's usually closer to 0 or 180 isn't it?

 

[Andre]

One important consideration is that at any given time (i.e., an instantaneous snapshot) the rotation and magnetic poles do not align. Take the present day field, for example, the poles do not align. If, however, you average over a period of time then the average magnetic poles will align with the rotation poles. How much time to average over depends on the availability and distribution (in space and time) of the data, but when all things are good several thousand years is enough time.

I agree that this doesn't sounds not unreasonable, but what is the actual status of this statement? Conviction, Axiom? Hypothesis? Theory? Any true observations that would support it?

 

[Simon Bridge]

It is empirically confirmed.

Look up anything geomagnetic.
Look at that link I gave you.

 

[geo101]

It would be pretty unusual for the angle between the rotational axis and the magnetic dipole to be close to 90deg though

It does pass through 90 degrees during polarity reversals and some excursions.

I agree that this doesn't sounds not unreasonable, but what is the actual status of this statement? Conviction, Axiom? Hypothesis? Theory? Any true observations that would support it?

This supported by a range of geomagnetic and paleomagnetic data. If you want to learn more look at http://magician.ucsd.edu/Essentials/WebBookse93.html#x109-18000014.5 from the http://magician.ucsd.edu/Essentials/index.html (a very good paleomagnetism book for those who are interested). The references in the chapter are good pointers for some more recent studies.

 

[D H]

One important consideration is that at any given time (i.e., an instantaneous snapshot) the rotation and magnetic poles do not align. Take the present day field, for example, the poles do not align.

Ten degrees is pretty dang close, though. If the orientations of the rotation and magnetic axes were uncorrelated and random (which they aren't), the expected value of the misalignment would be 60 degrees.

 

[Andre]

It does pass through 90 degrees during polarity reversals and some excursions.



This supported by a range of geomagnetic and paleomagnetic data. If you want to learn more look at http://magician.ucsd.edu/Essentials/WebBookse93.html#x109-18000014.5 from the http://magician.ucsd.edu/Essentials/index.html (a very good paleomagnetism book for those who are interested). The references in the chapter are good pointers for some more recent studies.

So I read in your first link:

One of the primary assumptions in many paleomagnetic studies is that the magnetic field, when averaged over sufficient time, averages to that of a GAD (Geocentric Axial Dipole) field.

Bold mine. Doesn't this suggest that the status is "assumption"? Of course it's certainly a very reasonable assumption and there is maybe no reason whatsoever to debate it. The point is, if the 'assumption' is used as a fact, then circular reasoning kicks in, proving the assumption based on an assumption.

 

[Simon Bridge]

When we do calculations, we often make approximations and "assumptions" for the sake of making the math simple or exposing some part of the physics we want to work on. These assumptions will be supported to various degrees by observation and experiment.

The assumption that the geomagnetic dipole averages to the GAD is well supported by the evidence. The language used reflects the fact that we do not actually have a record of the Earths magnetic field for all times in history which is what we need to turn it into an empirical fact.

It is the nature of science that anything we take to be true is only and assumption. However, some assumptions are so well founded that the difference between them and a fact is so small that pointing out the difference involves splitting hairs.

The use of empirical support is what stops these arguments from becoming circular. That and the fact that we don't really go around trying to prove things in science. A scientific proof is just the failure to disprove something by some particularly clever method.

But surely you know this already?

 

[geo101]

Bold mine. Doesn't this suggest that the status is "assumption"?

As Simon points out the assumption reflects the availability of the data that we have. The geomagnetic field has existed for at least 3.45 billion years, but high quality, high resolution data do not exist for that entire time period.

We only have enough to data to reliable test the GAD hypothesis over the last few million years (at least the last 5 myr, maybe as far back as 10-15 Ma), but we can confirm that it is valid. Beyond this there are further complications such as continental drift and a general lack of data make it difficult to test. There are tests that can be done and there is evidence to support GAD over the Phanerozoic, but the data are insufficient to provide the same level of support that we see over recent times

So in short it is strongly supported by observational evidence over recent time (e.g., http://paleomag.ucdavis.edu/research/acton/Publications/1996_Acton_Skewness_EPSL.pdf; I can find some more if folk are interested, but they may require journal subscriptions), but cannot be definitively proven over longer time periods.

If you are interested in reading more I can throw some references your way.

 

[Simon Bridge]

The status list is intriguing:

conviction
axiom
hypothesis
theory

Kinda semi-formal - very philosophy 101.

If we have to pick one, I'd say it is a hypothesis in that it is a specific synthetic proposition rather than an overall outlook. It is subject to being tested.

The trouble is, this makes it sound shakier than it is.
It is a very well tested hypothesis. So well confirmed it is often taken as axiomatic in research that attempts to advance our knowledge in this field.

Everything in science is either a theory, a hypothesis, or data.
Axioms are tools - subject to modification: we can take a proposition as axiomatic, work out the consequences and so design our experiment. The consequences don't pan out suggests the axiom needs to be revised. But I have not heard any actual scientist talk that way.
Convictions, as in strongly held opinions, are irrelevant.

We tend to talk about theories and hypotheses being variously; "established", "well supported", "under active investigation" and so forth. Hence, this is a well supported hypothesis.

To a certain extent you just have to accept that things are not all that well ordered - there being no clear hierarchy of theories to say one is better than another. There's just this: "has it been around a long time - has it been really cunningly tested?" thing.

 

[Andre]

Thanks for your elaboration. My point behind is, what to do when things get very iffy? An example is the Neoproterozoic tillite (evidence of glaciation). When that is found at sites around sealevel and it is hypothesed, after dating and paleo magnetic analysis that it would have been in tropical area's consequently it must have been rather chilly in that time.

But how valid is that? Which hypothesis might be suspect? How firm are all these suppositions? Alternately, Could the dating be wrong or the paleomagnetic assumptions and was the rock in high lattitude when it was formed? Or was it in tropical regions but at high altitude like the Andes and the Himalayas?

 

[Simon Bridge]

There is no way to tell from that statement. You have to read around the subject.

At any time there will be a range of competing ideas in any scientific field - in particular in areas of active research. In order to weight the various merits you have to become familiar with the literature in the field.

So to answer the question directly: when things get iffy, you hold everything as suspect until some definitive research is done ... or, since you cannot always wait, become familiar with the field (or hire someone who is).

 

[geo101]

Ten degrees is pretty dang close, though...

Ten degrees is pretty close, but that difference can hold a lot of information about the pattern of convection within the other core (when considered in the context of time series, as a single snapshot it is not so easy).

 

[geo101]

Snowball Earth...

Not a topic where I am well read, but there is evidence for and against and not just paleomag.

I do know, however, that some recent (currently unpublished) paleomag data suggests that some of the paleomag data that is viewed as being some of the most reliable evidence of low-latitude glaciation does not record Neoproterozoic directions, they are later remagnetizations.

And remagnetizations are one of the difficulties in getting good data with which to test the GAD hypothesis over very long timescales.

 

[Andre]

Maybe it is http://www.sciencedirect.com/science/article/pii/S1342937X11001250

Font et al 2011; Paleomagnetism and rock magnetism of the Neoproterozoic Itajaí Basin of the Rio de la Plata craton (Brazil): Cambrian to Cretaceous widespread remagnetizations of South America, Gondwana Research Volume 20, Issue 4, November 2011, Pages 782-797

Are we about to see some huge paradigm shifts?

 

[geo101]

That is an example of remagnetization, but not specifically related to Snowball Earth (at first skim read, it deals with Cambrian/Ediacaran meta-sediments and volcanics associated with orogenic belts).

Remagnetization is quite a common thing when you deal with rocks of all ages, but the older the rocks the more likely they are to experience remagnetization and the more likely it is that the remagnetization will completely overprint the original paleomagnetic recording. And when you deal with Neoproterozoic rocks (~542 - 1000 million years old) they have probably experienced quite a lot during their life time.

This is the unpublished work I was referring to.

Are we about to see some huge paradigm shifts?

It is certainly possible in terms of how much "reliable" data there is to support Snowball Earth, but as I said the specific details of all the supporting/contrary evidence is not something where I am fully read.

 

[Andre]

Thanks, that is very impressive.