So much for a honeymoon period; no sooner do I arrive at Scienceblogs than commentators start making demands. Fortunately, the story I’ve been ordered to write about was already in my ‘of interest’ pile, so in this instance it’s not too much trouble to oblige. I just hope I’m not setting a dangerous precedent.
The recent reports on the research of Jesse Lawrence and Michael Wysession highlight a novel application of seismic tomography, which uses the seismic energy generated by earthquakes to peer at variations in the Earth’s internal structure. Tomography requires a set of earthquakes where the time of rupture is precisely known, and a network of seismograph stations. Usually, researchers then look at the differences in the arrival times of seismic waves from a particular earthquake at neighbouring stations to identify places where the waves are traveling unusually fast or slowly: for example, if a stations starts picking up the waves much earlier than its neighbour, then the earthquakes are traveling through the rocks beneath that station unusually fast, and if they arrive later, they are traveling unusually slowly. Such variations are generally thought to correspond to temperature differences – the hotter the rocks are, the slower the waves will travel through them, and vice versa.
Lawrence and Wysession decided to look at something else as well; the differing degree of attenuation, or energy loss, of the seismic waves as they passed through the earth. This gives you information not about changes in temperature, but changes in strength; because they deform more easily, weaker rocks will tend to absorb more seismic energy as earthquake waves pass through them, increasing the observed attenuation.
Here’s what they found: a large zone of abnormally high seismic attenuation in the uppermost lower mantle (700-1400 km) beneath East Asia. The first images has removed the top 1000km of their 3D model to show the lateral distribution of this ‘Beijing anomaly’; the second shows a cross-section through it.
So what’s causing this anomaly? The fact that it doesn’t seem to correspond to any change in wave-speed suggests that it is not caused by temperature variations, but changes in composition, and the authors suggest that the most likely cause is the addition of water to the mantle rocks. This would be supplied from above, in the form of hydrated oceanic crust of the Pacific plate, which has been being subducted beneath East Asia for many tens of millions of years. We already know that some water is released from subducting plates into the upper mantle, which is what causes arc volcanism behind subduction zones (as discussed in this old post, hydrated mantle rocks have a lower melting temperature), but it seems that some is locked away in a stable enough form that it will only be released at the much higher temperatures and pressures present in the lower mantle, where it reacts to form a broad region of weak, hydrated mantle rocks. Bear in mind this conclusion is an inference, not yet supported by much data on how hydrated minerals actually affect seismic propagation in the lower mantle. But if it is correct, it is a very interesting result indeed; the fact that water from the surface can be recycled all the way back into the bowels of the planet calls into serious question the fairly common assumption that the lower mantle is pretty isolated, physically and chemically, from the tectonic action going on above it.
Interestingly, you don’t see quite the same thing over the other side of the Pacific, which is also a region of long-term subduction: there is a much smaller region of high attenuation beneath North America, and no anomaly at all to speak of beneath South America. Dave over at Geology News wonders why; my guess would be that it’s due to differences in the age of the crust being subducted. The crust being subducted beneath Asia is almost 200 million years old, so has had plenty of time for large amounts of hydrated minerals to accumulate through reactions with seawater. In contrast, the crust being subducted beneath the Americas is much closer to the spreading ridges which created it, and is therefore much younger, so will be relatively ‘dry’.
Much less interestingly, it seems that Joe Meert was spot on with his pre-emptive strike against creationists obsessed with flood geology, who he figured would start claiming that we’ve found the source of water for the Biblical Flood. You only need to look at the comments to the PhysOrg write-up to see how depressingly right he was.