Story in brief: An unknown agent blows up the moon, initially into 7 major pieces and a bunch of little ones. The pieces are gravitational bound, but collisions between the pieces continues the break up process. Stephanson postulates that the number of pieces reaching the earth's atmosphere will increase to the point that in 2 years Earth will experience a 'white sky' from re-entering rocks, which will burn off the surface of the planet, and that this will last for several thousand years. I don't understand how this is possible. I know that in elastic collisions a small object can come away with a larger velocity. The classic experiment is to drop a tennis ball 1" apart from a basketball at the same time. The basketball coming up meets the tennis ball coming down, and the tennis ball leaps away at about twice the speed relative to the floor. But collisions are not going to be elastic. Much of the energy will be consumed breaking big rocks into little rocks, and relative speeds are going to be low. (The description has the 7 major chunks being indistinct in a cloud of debris a week later, and being a patch several times the diameter of the full moon.) To burn off the surface of the earth, you need dump an appreciable fraction of the amount of energy the earth gets from solar radiation. The moon is 1/80 of the mass of the earth. Positing 5000 years of bombardment we are talking about 1/400,000 of the mass of the moon per year, if the entire moon came down. This would be a planet wrecker: At 22 billion cubic km 1/400,000 is still a lot of rock. But from a gravitaionally bound cloud of rocks, what would be the ejection rate of mass? Given the initial descriptions, I figure that most of the mass would recollect into a new moon.