For the past two years I’ve been developing an idea about how ice ages work. It’s a very simple idea. And the very simple idea is that when large areas of the surface of the Earth get covered in deep snow, the snow acts as thermal insulation, and causes the rocks beneath it to warm up. And when the rocks have warmed up enough, they melt the overlying snow. And without their insulating snow cover, the rocks then cool down. And when they’re cool enough, the snow starts settling on them again, And that’s the glaciation cycle. It seems to be an idea that’s absent from contemporary climate science. So maybe there’s something wrong with the idea. But, to date, nobody has explained to me what’s wrong with it. And, actually, I don’t think there’s anything wrong with it at all.
But it also occurred to me that, if ice ages work the way I’m suggesting, then that means that large areas of the surface of the Earth heat up during ice ages when they’re covered in deep snow. And also that, during interglacial periods when most of the snow has melted, the same large areas of the surface of the Earth cool down. And most materials expand when they heat up, and contract when they cool, these large areas of the surface of the Earth must also be expanding and contracting during glacial cycles. And when they expand these warm surface rocks must push against each other. And when they contract they must pull apart from each other, with cracks opening up between them.
So I’ve been wondering whether plate tectonics might be explained not by convection currents in the Earth’s mantle (the usual explanation), but instead by the expansion and contraction of warming and cooling plates. Could these plates that are all jostling with each other act to move each other around? Could it explain continental drift?
In principle this seems quite plausible, because the continents are all float on top of molten rock tens or hundreds of kilometres beneath them, just like ships on an ocean. So maybe it’s actually quite easy for continents to push each other around, just like it’s quite easy for tugboats to push around much bigger ships. But since I’ve just got expanding and contracting plates, wouldn’t they all push against each other when they expand, and go nowhere at all?
But the Earth’s plates wouldn’t all expand and contract in step with each other all over the world. Plates at the equator would probably not heat up during ice ages, because they probably never get covered in snow. Bur plates at the poles would heat up during ice ages, and cool down during interglacials. So there’s be differential warming and cooling, expansion and contraction, all over the surface of the Earth.
And I put forward an explanation for how cyclic expansion and contraction could result in the motion of plates in one direction. And the explanation was that as one plate warmed and expanded, it pushed an adjacent cool plate away from it, and when it cooled and contracted, and the boundary between plates opened up, molten rock from below would rise to fill the gap, and then cool to form a solid plug, so that when the plate warmed up again, and started pushing, it would push both the adjacent plate and the the solid plug in between away from it. Plate movement wouldn’t be continuous. but instead cyclical.
The idea seemed to work quite well where two plates abutted each other, and would result in the formation of long ridges of volcanic mountains along the plate boundary (e.g the Andes in south America).
But how could it explain single volcanic islands like Hawaii, with a string of extinct volcanic islands adjacent to it? Isn’t the idea of a mantle plume a much better explanation for why there always seems to be a volcanic island in one place?
But I think there may be a good explanation for how single volcanic islands form. They’re not found where two plates meet: they’re found where three or more plates meet. For when these plates contract, and gaps open between them, the area of the small triangular junction between the three plates expands twice as fast as the gaps between any two plates. And this means that vulcanism is much more likely to happen at the triple junction point than elsewhere. So if there was a triple plate junction at the location of Hawaii, that’s where you’d expect to see an island appearing. (The explanation from this is geometrical: When the distance between two plates expands from 1 km to 2 km, the area between them doubles. But when the side length of the triangular junction increases from 1 km to 2 km, its area quadruples.)
And, strangely enough, Hawaii actually is a (roughly) triangular island, with side lengths of about 11 km. None of the adjacent older islands in the chain are particularly triangular. But this might simply be be because the older islands have been eroded by waves from their original triangular shape.
And if Hawaii is still erupting, it’s because it’s sitting over a triple plate boundary which is still widening during our current Holocene interglacial, during which plates are cooling and contracting.
So that’s a new explanation of how single volcanic islands like Hawaii are formed. The same explanation might offered for Iceland, and for numerous islands elsewhere in the Pacific and Atlantic.
And one argument against the mantle plume hypothesis would seem to be that, if the mantle plume is continuous, shouldn’t it produce a single long island stretching away from it, rather than a series of islands?