Take another look at the ice age record, “stages of glaciation” on bottom of the chart. Although it is similar to the solar forcing, it is not the same, and that means there must have been other influences. Not surprising, but it’s worth asking what they might have been. Look, for example, at the low spot — the ice age maximum — at about 640 kyr.
It’s one of the two lowest spots on the graph. much lower than most of the lows, though there’s a similar low at 440kyr. It’s not just the Milankovitch cycles; something else is going on. Also it’s interesting that the temperature climbs so rapidly out of it. Well, all the temperature climbs are fairly rapid; ice melting seems to go fast when it goes. In any case, the Milankovitch cycles are not the only large-scale influence on our climates.
Yellowstone is another. Any volcano can be important, but Yellowstone is (or was; we hope it’s more settled today) a supervolcano. The difference, though still under debate in its details, is fairly clear: volcanoes build mountains; supervolcanoes erase them. Yellowstone’s largest eruption was 3.1 million years ago, and is not on this graph; it spewed 600 cubic miles of stuff, enough to cover California 20 feet deep, had it been there to cover. But the eruption of 640,000 years ago spewed 240 cubic miles of material, a thousand times more than Mount St. Helen’s, and still enough to cover California 7-8 feet deep. Of course it was spread over most of the western United States and indeed some of the ash went round the world. It has to have diminished the sunlight for quite a while, must have caused a cooling, and thus we can be sure that we are looking at one of the records record of this event.
It’s easy to find maps of the Yellowstone caldera. Here’s an image of part of one edge. The total caldera is 50 miles long and 30 miles wide. It’s so large, and so completely overgrown, that it takes a special eye to see it as a caldera, but there is no doubt of it.