AGU Dispatches: Superchrons and subduction

I started the day with my paleomagician’s hat on, sitting in on a session looking at the long term behaviour of the Earth’s dynamo. Changes in the strength and reversal frequency of the Earth’s magnetic field give a unique insight into what’s going on deep within the Earth – where the field is generated – far back in Earth history. Opening speaker John Tarduno probably went the farthest back: he has developed some extremely impressive kit to analyse magnetic grains enclosed in single quartz and plagioclase crystals. This has enabled him to measure the intensity of the Earth’s magnetic field as far back as three and half billion years ago, which indicate a field somewhere between 50 and 100% of the strength of the current field. However, there was a stronger solar wind emanating from the sun back then, so the magnetopause – the ‘edge’ of the geomagnetic field’s protective influence – was probably much closer to the edge of the Earth’s atmosphere. Tarduno argued that this would have meant significant loss of water from the early Earth as the atmosphere got sloughed off during solar storm events, meaning that Earth was either much more initially water rich than we thought, or cometary impacts supplied more water later. I was also intrigued by recent modelling of the geodynamo that predicts that even earlier in Earth history, there was no dipole field at all: Tarduno hopes to test this by analysing the ancient zircons in the Jack Hills conglomerate.

Several other talks in this session looked at magnetic superchrons – periods of Earth’s history where the Earth’s magnetic field is stuck in the same polarity for tens of millions of years without reversing. Studies of the most recent superchron, between 120 and 83 million years ago, indicate that, compared to periods when the field is reversing more rapidly, other properties of the field change too: it is stronger during superchrons, and the patterns of secular variation (drift of the magnetic pole around the Earth’s rotation axis, or true north) are different. Work is being done to check if other superchrons share the same characteristics, and so far it seems they do. The other big mystery is the apparently rapid transition between periods of frequent reversals and superchrons. Andy Biggin reported on the latest modelling which indicates that reversal frequency is highly dependent on equatorial heat flow at the core mantle boundary, with higher heat flow leading to faster reversals. He then went on to link the abrupt start of a superchron with an episode of True Polar Wander (movement of the whole solid Earth relative to the axis of rotation, driven by density imbalances): a provocative idea, even if I’m generally suspicious of invoking large amounts of TPW to explain dramatic changes when it has proven rather difficult to definitively prove it happens at all.

After that interesting session, I changed into my tectonic cap and sat in on a couple of sessions looking at the deformation associated with subduction. Some talks emphasised how lateral segmentation of the megathrusts on subduction zones is a very fluid thing. One particularly good example of this is the Sunda megathrust that generated the extremely destructive 2004 Boxing Day earthquake and tsunami: using historical records and uplift and subsidence recorded by corals, Belle Philibosian showed that each of the four rupture sequences in last 1000yrs is unique, with different segments of the megathrust rupturing singly or together in each one, and even the patterns of subsidence between the earthquake sequences being very different. In another interesting talk, Kevin Furlong argued that, in contrast to the standard conceptual model of how strain builds up across the plate interface between earthquakes, which assumes that it is mostly taken up by convergence and uplift in the overriding plate, is an oversimplification. This pattern is only one of a range of possibilities, that only occurs when the overriding plate is weaker than the subducting plate; if the situation was reversed, then most interseismic strain may actually accumulate in the slab instead. He provided some good examples that suggested that real-world subduction zones mostly fall on a continuum between these two extremes.

There were also a lot of new insights into how the structure of the plate entering the subduction zone effects things like the occurrence and frequency of great earthquakes, deformation in the overriding plate, and even the style of volcanism in the back arc. This is of particular interest to me due to my studies of the subduction margin off the East Coast of New Zealand, where the changing thickness of the subducting Pacific is an important control on how the coast there is deforming. Perhaps the most interesting insight is that sections of the megathrust where the plate entering the trench is rough, due to lots of volcanic seamounts or faulting, do not generate great (larger than magnitude 8) earthquakes, whilst smoother areas with fewer seamounts and/or faults next door do. I found this rather counterintuitive at first, since you might expect a rougher interface to be stronger, but it seems that having lots of strong ‘asperities’ close to each other prevents any one rupture from growing too large. There was also lots of interesting data from Alaska and the Aleutian trench, and I had a lot of fun learning about the complex tectonic evolution of this area.

All in all, a stimulating first day. Tomorrow, I will be braving the poster hall vortex, before an afternoon of pontificating about social media (I am talking about Twitter in session PA23B – Facebook, Twitter, Blogs: Science Communication Gone Social—The Social Media 101) and blogging in a panel discussion. Plus, there are plans afoot for a gathering of the great and good of the geoblog(and tweet)-osphere. The current plan is to meet at around 8.30 in Johnny Foley’s, which is here. The more the merrier!