Most days I spend my mornings writing my blog, and then spend the afternoons and evenings developing my terrestrial glaciation simulation model.
Currently in this model when snow builds up on the ground the surface rocks beneath it heat up and start melting the bottom layer of snow. So after initially getting quite deep, even though it’s continually snowing, the snow begins to thin. And in the 100,000 year run shown at right, in which surface albedo varies with snow depth, air temperature falls and then rises again. So in this model surface rock temperatures are rising and falling some 15 – 20 degrees between interglacials.
When these surface rocks heat up, they expand. and when they cool down they contract – an idea I’ve explored before. And if these rocks are part of large monolithic tectonic plates, thousands of kilometres wide, these small local expansions and contractions will appear as large expansions and contractions along their perimeter.
And since the entire surface of the Earth is covered in large, irregularly shaped plates (above), their expansion and contraction will bring them into collision with each other. When the plates are expanding there will be convergent plate boundaries, and when they are contracting there will be divergent plate boundaries. And this offers a new explanation of plate movement.
But Earth scientists don’t think that this is the explanation of convergent, divergent, and transform plate boundaries. They think the plates are being moved around by convection currents in the Earth’s mantle. But how sure are they about that? Is it settled science? Yesterday I came across a video in which mantle convection was described, and which began with the declaration that this theory of plate movement was “not definitively proved.”
So I was wondering yesterday whether plate heating and cooling (during glacial cycles) might explain the observed phenomena just as well as mantle convection. And at the outset it looked like in might do quite a good job. For when expanding convergent plates collided there would be mountain ridge uplift (e.g. the Andes in South America), and where contracting divergent plates met the plates would move apart (e.g. in the mid-Atlantic ocean).
However, if plates are continually warming and expanding, and cooling and contracting, it’s rather hard to see how plates could continually move in one direction (e.g. the African plate away from the South American plate). Wouldn’t the plates just be jostling each other, with no net motion in any particular direction? In this mantle convection flows would appear to offer a much better explanation of plate motion, as the plates are carried along by the currents flowing beneath them, like icebergs on the ocean.
But then I thought that there might in fact be a fairly simple explanation for how plate movement could happen with plate warming. And it was that as two adjoining plates contracted, this would create an opening in the surface into which molten rock from beneath could rise. And when this molten rock had risen into this gap, it would then cool and harden. And so when the plates stopped cooling and contracting, and started warming and expanding again, the plug of solid granite rock that had filled the gap would prevent the plates coming back together. Instead a mountain ridge would get uplifted where the plates met. So while it was easy for the plates to split apart, it wasn’t easy for them to rejoin each other. And so the net motion of the plates would be to slowly move apart.`So this account explains both surface spreading of the ocean floor, and ridge (and island) uplift.
And the same reasoning might explain the formation of island chains (like the Hawaiian Islands). For when islands have been uplifted by plate expansion, they form a solid plug which will prevent not only plate expansion, but also plate contraction at that point. So when the plates begin to cool and contract, the surface rocks will not fracture at the island, but at some point nearby where the crust is thinner and weaker.
The mantle convection explanation of this sort of island chain creation uses a controversial hot stationary mantle plume which bores through the crust to create new islands as the plate moves over it.
Anyway that’s my plate expansion explanation of ocean floor spreading. A further corrollary of it would be that the entire monolithic 5,000-km-wide South American plate would be moving westwards, and during episodes of plate expansion it would ride up over the adjoining Nazca plate under the Pacific ocean, pushing up the Andes mountain range in the process, with the Nazca plate subducting under them.
The idea obviously needs further development. But it suggests quite different plate movements would take place. And the size of these movements would be determined by the extent of adjoining plates, and the temperature of them.
P.S. Using sandstone’s coefficient of thermal expansion, 12 x 10-6, and a temperature increase of 20° C, the 5,000 km wide South American plate would increase in width by 1.2 km, and contract by the same amount with a subsequent temperature decrease of 20° C.