The new geophysics results from the spreading ridge on the East Pacific Rise, just published by Toomey et al., have already been mentioned on Deep Sea News: as guest poster Kevin Zelnio discusses, this new mapping of where the mantle is upwelling relative to the ridge axis offers some clues regarding why the distribution of volcanism and hydrothermal venting varies along the plate boundary. But it turns out that these results may also have important implications for our basic understanding of plate tectonics.
Toomey et al. used high-resolution seismic tomography (see here or maybe here) to map the velocity structure of the mantle beneath the East Pacific Rise. Determining the bulk seismic velocity of particular regions allows the temperature of the mantle (fast = cold, slow = hot), and hence the concentration of melt, to be determined. The study also looked at seismic anisotropy. A common property of many minerals is that sound waves propagate at different speeds along the different crystal axes, and if something has caused the crystals in a rock to all line up in a particular orientation, this will result in seismic waves travelling measurably faster in certain directions through the rock compared to others. In the mantle, seismic anisotropy can be related to the direction of mantle flow.
The standard picture of mantle upwelling beneath a mid-ocean ridge is that it is a passive process: the plates move apart, creating a gap for mantle material to upwell into and melt as it decompresses. If this is the case, then you would predict that the hottest, most magma rich region would be right beneath the ridge axis. You would also expect any seismic anisotropy to parallel the direction of spreading, due to the flow of the mantle rocks as they pile up against the bottom of the diverging plates. As it turns out, at least in this part of the East Pacific Rise, you’d predict wrongly. As the upper image below (source) shows, the greatest concentration of melt (shown in red), and hence the location of maximum mantle upwelling, can be located up to 10 km away from the ridge axis itself (dotted black line). Not only that, but the lower image demonstrates the direction of mantle flow indicated by the seismic anisotropy (green arrows) is rotated 10° anticlockwise with respect to the spreading direction (black arrows – NUVEL 1A is a model of the last couple of million years’ plate motion). In fact, the melt zone also appears to have a similarly skewed trend, which is what causes the offset.
The most obvious explanation for this mismatch is that there has been a fairly recent and abrupt change in the configuration of the ridge system, causing the spreading direction to rotate clockwise, and the upwelling pattern in the mantle has still yet to fully respond to this change. However, if you actually look at how the spreading direction has changed in the last few million years, you find the opposite: the spreading azimuth is rotating anticlockwise. In other words, the plate boundary configuration appears to be lagging a change in the pattern of mantle upwelling, with the mantle flow actively driving a change in plate motions. That’s about as far from a passive process as you can get.
This is interesting, but it’s also a little weird; the drag which the upwelling mantle can exert on the base of the plates is probably fairly negligible as a driving force compared to, say, the pull of all those cold, dense subducting slabs at the other edges. It would be interesting to look for similar offsets and skews underneath other sections of the East Pacific Rise; if you saw the same trends everywhere, this would confirm the influence of some large scale mantle process, whereas less systematic variations (both clockwise and anticlockwise skews of varying magnitudes) would indicate more local processes at work. Either way, it seems that ridge-mantle interactions are not quite as simple as we thought.
Comments (4)