The Science of Discworld Revised Edition

organism in its own right. The idea has been romanticised into the Earth being a kind of Earth-mother, but what do you expect when you attach the name of a goddess to your new scientific concept? Stripped of the romantic frills, the point is that our planet acts as a single system, and it has evolved mechanisms that keep it functioning effectively. This development is a consequence of innumerable subsystems – organisms, ecologies – evolving mechanisms that keep them functioning effectively. If every member of a team gets better at playing their role within the team, then the team as a whole improves too.
    Complexity is a double-edged sword. More complex forms of life find that the ordinary problems of living on a planet are more and more under control … except for those confounded problems from outside, like meteorites, which can be disastrous.
    The Moon, Mercury, Mars, and various satellites are covered in circular craters, some large, some small. Nearly all of those craters, we now know, result from the impact of a big lump of rock or ice or a bit of both. A few are volcanic. Not so long ago most were thought to be caused by volcanoes, but that turned out to be wrong.
    Several planets, among them the Earth, do not show obvious signs of impacts. Is that because nothing hit them? No. An atmosphere helps: smaller bodies burn up before they hit the ground. It’s the closest to a protector-turtle that we get. But bigger rocks can still get through the defences. The main reason why some planets show no clear signs of impacts is because those planets have weather, like the Earth, which erodes the craters until they disappear, or episodes of massive vulcanism, like Venus, which resurfaces the planet, or are gas to begin with, like Jupiter and Saturn, and don’t show permanent marks.
    In Quebec there is a lake called Manicouagan. You can’t miss it on a map: look near 51°N, 68°W. It’s circular, and it’s big: 44 miles (71 km) in diameter. It is the weathered remains of a gigantic crater that formed 210 million years ago when a rock two or three miles (3–5km) across collided with the Earth. There is a central peak made from rock that melted in the heat that the impact generated and then solidified; more molten rock flowed across the floor of the crater and still can be found today. The lake fills a ring-shaped valley that glaciers carved out of soft rock that originally formed the crater walls, and was eroded away and collapsed.
    Also in Canada is the Sudbury impact structure, the largest on the planet. It is 190 miles (300 km) across and 1.85 billion years old, and the rock that made it was about 20 miles (30km) in diameter, and the energy released in the impact was equivalent to one quadrillion tons of TNT, or about ten million really big hydrogen bombs. In Vredefort, South Africa there is another impact structure of a similar size, formed 2.02 billion years ago. These may not remain the record-holders: an even bigger impact structure, about twice the size, is suspected to exist in the Amirante Basin of the Indian Ocean. Altogether, more than 150 impact structures – remnants of craters – have been found on the Earth’s land-masses, and many areas have not yet been thoroughly surveyed. More than half the Earth’s surface is ocean, and incoming rocks should hit pretty much at random, so the total number is probably closer to 500.
    These are all fairly ancient craters, but there is no good reason to believe that such impacts could not happen again. Big impacts are rarer than small ones, because big lumps of rock are rarer than small ones. Impacts the size of Sudbury or Vredefort should happen about once every billion years. (It should not be a surprise that when such impacts finally did arrive, about two billion years ago, two of them came along together.) Since nothing that size has happened for two billion years, it might seems that we are overdue for another one, but that kind of reasoning is a statistical fallacy. Rare, isolated events usually obey the so-called ‘Poisson distribution’ of probabilities, and one feature of this distribution is that it ‘has no memory’. At any time, whether two major impacts have just happened, or none have happened for ages, the average time to the next one is always the same – in this case, a billion years.
    It
could be
a few decades, mind you. But it couldn’t be tomorrow, or even next year, because we would have spotted such a body coming by now.
    The most