6 fields a-flipping

A post by Chris RowanOn the 6th day of Christmas my true love sent to me: 6 fields a-flipping

About every millon years or so, on average, the Earth’s magnetic field reverses itself: the north magnetic pole becomes the south magnetic pole, and vice versa. The most visible evidence for this can be found in the world’s ocean basins, where the striped magnetic anomalies that provided compelling evidence for sea-floor spreading and mid-ocean ridges also provide a continuous history of the last 200 million years of so of geomagnetic field behaviour.

Magnetic anomalies in the North Atlantic. Source: SDSU (link to kml file)

As I have explained before, ‘on average’ has no real predictive power: the field can spend only a few hundred thousand years in one polarity state before switching to the other, then right afterwards spend a few million years in that state before flipping again. At least as far as we can tell, there is no periodicity to field reversals. This very unpredictability does have a rather useful consequence, however. Because, at least as far as we know, there is no real periodicity to the field, over any particular period the temporal pattern of magnetic field reversals – how often the field reverses, and when – is unique to that period, and unlike the pattern in any other period. So if, for example, I am trying to date a sedimentary sequence that spans six magnetic polarity reversals, with mostly normal (shaded black) periods, or chrons, interspersed with shorter reversed (white) chrons:


from the temporal pattern of reversals – at least a medium normal, very short reversed, longish normal, medium reversed, medium normal, short reversed, at least a long normal – I can fairly confidently match it to the series of reversals that occurs over the KT boundary. As long as your sequence spans enough polarity chrons, that pattern is not going to be like any over period covered by the geomagnetic polarity timescale (GPTS) principally constructed from measurement and dating of the seafloor magnetic anomalies.


This is effectively what magnetostratigraphy is all about: locating polarity reversals within a sequence and matching the pattern of polarity reversals to the existing. Reversals are extremely good chronological markers: they are usually well-dated, occur virtually instantaneously from a geological perspective (a few thousand years is not very significant if you’re looking at a sequence spanning a few million) and, most importantly, mark a global event that is simultaneously recorded by rocks forming in all sorts of ways and environments. Magnetostratigraphy is therefore especially good for long distance correlations between sequences that record environments so different from each other that the same index fossils may not be present in them.
The sequence of seafloor magnetic anomalies is often described as a magnetic barcode, and in a very real sense it can act as one: with the right palaeomagnetic tools, you can read the age of a geological section directly from the recorded sequence of reversals.

…5 focal mechanisms,

4 index fossils,

3 Helmholtz coils,

2 concordant zircons,

and an APWP.

Categories: geology, palaeomagic

Comments (5)

  1. andy says:

    So what’s the deal with the “superchrons”? Do these represent a long-duration tail of the chron length distribution or are they genuinely anomalous?

  2. Susannah says:

    I am loving this series!

  3. Chris Rowan says:

    Re: superchrons. In some ways they are clearly anomalous, because the two known examples (in the Cretaceous and Permo-Carboniferous) are so very much longer than any other known polarity intervals that they form extreme outliers on a frequency distribution. The question is whether they have a particular cause or whether the geodynamo just occasionally switches into a non-reversing state; at present, nobody really knows.

  4. Graham King says:

    Chris, I’m interested, but have no experience in this field.
    So mine may be naive questions: but, doesn’t the thickness of each chron depend crucially on both the length of time period, AND the rate of deposition during that time period, during which it was formed?
    Is there reason to believe these rates are or were uniform over time and even within a single chron?
    It seems to me that deposition rates from any geologic or even cosmic process cannot be assumed to remain constant.
    Is this a problem? Are there checks – checks against both mismatches and missed matches – the false positives and negatives, that might be caused by non-uniform deposition rates altering the relative widths of polarity-bandings?

  5. Chris Rowan says:

    You are correct that deposition rates are an issue. However, sedimentation rates in a particular depositional environment, whilst by no means uniform, do not generally vary over such a range that the relative lengths of different chrons are going to be skewed beyond recognition . If your section spans a change in depositional environment, of course, you have to be more careful.
    In reality, magnetostratigraphy is integrated with other dating methods – radiometric, biostratigraphy – and these, and consideration of the geological context, usually mean that you know roughly which bit of the timescale you’re looking to match a pattern to. That narrows down the chances of false positives.