This terrifying video depicts the collision of a small planetoid with the Earth.
You can read a somewhat awkward translation of the Japanese commentary here. The following is my adaptation of the commentary.
The video is an extract from the 2005 NHK television series Miracle Planet II. Episode 1, The Violent Past, portrays the formation and early history of the Earth. During the “late heavy bombardment” of the Hadean era (about 4 billion years ago), the planets of the inner solar system were bombarded by many thousands of meteorites. During this period it’s thought that the Earth received several impacts creating craters 5,000 km across. (Earlier impacts very likely included some that were much bigger than this, especially if the giant impact hypothesis of the origin of the Moon is correct.)
This is a simulation of one of those giant Hadean impacts, but transposed in time to the present day to give a visceral impression of the scale and consequences of the collision.
The bolide is about 250 km in diameter,1 similar in size to the asteroid Juno. In fact, since the video doesn’t give it a name, I’m going to call it “Juno”. Juno is a typical rocky planetoid with a density of 3,400 kg m−3 and a mass of 3×1019 kg.
The opening frames of the video show that “Juno” is molten below the surface, with lava flows and glowing impact craters. This heat must be the result of recent collisions: even by the Hadean period bodies of this size had long since lost the heat of their formation. Its spherical shape may indicate that it has recently been completely molten: a solid planetoid of this size isn’t big enough to form a sphere under its own gravity.2 A 20 km s−1 collision with a 20 km body would yield more than enough energy to completely melt Juno; such a collision could have been the result of whatever violent event precipitated the late heavy bombardment.
Approaching the Earth3 at 20 km s−1, Juno carries 6×1027 J of kinetic energy. It’s estimated that the meteorite responsible for the Chicxulub crater and the Cretaceous–Ternary extinction event was 10 km in diameter and released 5×1023 J, so this impact is about ten thousand times as powerful.
The shock wave of the impact travels through the Earth’s crust at 6 km s−1, peeling off the crust4 and flinging it into space.
At five minutes from impact, the shock wave has travelled 2,000 km and has weakened to the point where it is unable to further destroy the crust. The resulting crater is 4,000 km across, with rim walls 7 km high. The reduced shock wave continues with the power of a magnitude 12 earthquake, destroying most man-made structures in the hour it takes to encircle the globe. (This earthquake is not depicted in the video.)
The earthquake is followed by an air blast, travelling at 300 m s−1 (the speed of sound in air), taking about 18 hours to reach the antipodes of the impact site. The air blast flattens trees and destroys any buildings left standing after the earthquake. (Again, the air blast is not depicted.)
Meanwhile the heat yielded by the collision has vaporized 1020 kg of rock, heating the resulting gas to 4,000 K, two-thirds of the temperature of the surface of the sun.5 This mass of incandescent vapour bulges up in a huge dome above the crater (about 02:40 in the video) and falls back to Earth, covering the globe.6
The seas boil away rapidly; in a month the deepest trenches are empty and bedrock of the ocean floors is melting.7 The surface is lifeless and dead. In time it will cool and the volatiles will condense, raining down again onto the surface, scouring the barren lands and refilling the oceans. However, it is likely that some life has survived: there are lithoautotrophic bacteria that live deep in the Earth’s crust, metabolising minerals and not dependent on sunlight, atmosphere, or surface water. In a temperate zone between the sterilized surface and the depths heated from below, these microbes may survive to emerge and re-populate the world.
↩ The size of the planetoid is a bit tricky to determine. The commentary says only, “The diameter of the meteorite is slightly bigger than the breadth of Honshu”. Honshu is 230 km across at the widest point.
At 00:54 in the video there’s a frame with Juno and the curve of the Earth simultaneously visible. I make Juno to be about 100 pixels high and the Earth about 4,000 pixels high. If their centres were the same distance from the camera Juno would be about 320 km in diameter, but although Juno’s shadow is close to the horizon it’s not quite there, so we should take off a bit to correct for perspective.
Some other commentators8 had a bit of trouble estimating Juno’s size from the video; “about the size of the Moon” (3,500 km diameter) is a popular option but is clearly much too big; astronomer Phil Plait guessed 100 miles (160 km); and bloggers at the New Scientist guessed 100 km.
↩ See for example the oddly-shaped Vesta, nine times the mass of Juno.
↩ No known objects of the size of Juno have Earth-crossing orbits; the largest known Earth-crossers are about 10 km in diameter, like Sisyphus, and it’s unlikely that there are any unknown bodies the size of Juno in the inner solar system. So nothing like this impact can happen in the foreseeable future, nor has anything like it happened for 3.8 billion years.
↩ You can see this effect in miniature in Meteor Crater, Arizona, where the overlying strata have been peeled off and turned upside down.
↩ Figures from the commentary. The vaporization of rock takes about 1027 J.
↩ The Westminster Clock Tower (04:47) and the Temple of Poseidon (05:00) would probably not have survived: these structures would have been demolished by the earthquake and the air blast. But their survival to see the rain of rock vapour is not impossible and their mournful presence brings home the human scale of the catastrophe in a way that an undifferentiated pile of ruins would not.
↩ The boiling of the Earth’s oceans takes about 1026 J, about a tenth of the heat energy of the rock vapour.
After the impact the Earth’s surface would be hidden by a blanket of incandescent gas, which would cool to an unbroken cover of hot clouds, like Venus. This wasn’t shown: maybe the hellish surface of molten rock was considered the most dramatic image with which to end the sequence.
↩ One of the curious aspects of this animation is how keen people seem to be to rubbish it. You can see some quite bogus criticisms here ("If it struck out at sea the water would cushion the impact … there will be survivors depending on how much oxygen remains and other environmental damage"), here ("the meteorite doesn’t have any gas trail, the shadow is way too big on the earth, and something that big would rip a third of the earth away"), here ("This scenario simply couldn’t happen, it defies Newtonian Physics at several levels"), and here ("Totally implausable, close to laughable, completely disingenuous science … only a moron would believe this is something to be credible").
There may be lots of simplifications, but the general story, the sizes, speeds, and energies involved seem right, at least to within an order of magnitude or so. The events match the results of the Impact Effects program at the Lunar and Planetary Laboratory.9 In particular, the video gets some things right that I would have guessed wrong: for example, I would have imagined that ballistic ejecta would thoroughly bombard most of the Earth’s surface. But according to Collins, Melosh & Marcus, Meteoritics & Planetary Science 40(6):817–840 (2005), the intensity of the bombardment is greatest at the crater rim and falls off from there with the cube of the distance (equation 43 of that paper), so that it’s negligible at the antipodes.
So why the general strong disbelief?
↩ If you’re trying this out yourself and wondering why the damage seems to be limited to that caused by the fireball, earthquake, and air blast, see the authors’ paper for a list of consequences not considered by the program, especially page 833: “Currently, we do not make any estimates regarding the potentially global environmental consequences of large impact events.”