| They got a lot closer than that in recent times, but not quite as massive: http://en.wikipedia.org/wiki/Tunguska_event In the long run cometary impact is pretty much a certainty and close passes are a lot more likely than impact. So even if this particular observation is probably attributable to something else there is a good chance that something like it did happen at another point in time in the last 2,000 years or so. http://en.wikipedia.org/wiki/Chicxulub_crater was 65M years ago, http://en.wikipedia.org/wiki/Sudbury_Basin 1.9B years ago, http://en.wikipedia.org/wiki/Vredefort_crater 2B years and so on. The http://www.passc.net/EarthImpactDatabase/index.html Earth Impact Database holds currently 183 recorded impacts over a period of about 2.4 billion years. 2400M / 183 gives a probability of 7.6x10-8, but that's for impacts, and near misses don't leave any evidence other than people that happened to be looking in that direction. The question then is how many near misses before you get a 'hit', probably a fairly large number depending on how you classify a near miss. Anything closer than the earth moon orbit would be fine with me as a 'near miss' (according to George Carlin an accident is a near miss, but never mind that). The exposed target of the earth-moon bounded sphere is substantially larger (~2000 times or so?) and would increase the chance of a near miss from the impact figure to about 0.000167 for any given year, or about once every 6,000 years. And the base figure here is craters that we know about, which means they're likely on the land, that misses 2/3rd of the planet. I'm sceptical about this but I would not rule it out entirely. |