🚀 go-pugleaf

|> 
|> Upon reaching the atmosphere, the comet would begin heating up as do
|> meteoroids and space craft. Like the Tunguska object of 1908, the
|> outer layers of the comet would begin heating up and a large trail
|> of dust would begin forming behind the nucleus and hanging in the
|> sky. If at night, this dust trail would exceed the Milky Way in
|> brightness, while in daylight the trail would appear grayish against
|> the blue sky.
|> 
|> Somewhere around 20-40 miles above the surface of Earth, the comet
|> nucleus would explode. At this point, the Tunguska object, which has
|> been theorized as perhaps 100 feet across, may have been completely
|> vaporized and a shock wave circled the Earth twice. On the ground,
|> a forest was hit by the downward moving shockwave and was leveled for
|> a distance of 50 miles in every direction. In other words, a 100-mile
|> circle of trees was laid flat. The evidence leading to the belief
|> that the Tunguska object was vaporized was the lack of any apparent
|> craters. The shock wave from P/Swift-Tuttle should be tremendously
|> larger, though I have no means of computing exactly how large.

Surely this is wrong. The square-cube law prevents you just scaling up the
atmospheric interaction like this.  The comet hits atmosphere at say 150 km
altitude at 50kms^-1. We have no way of predicting the angle of impact, but it's
unlikely to be very flat. Thus the comet will almost certainly be in the
atmosphere for less than ten seconds. In this time th outer layers will
undoubtedly heat up, and volatiles near the surface will boil and explode, but
the bulk of the 10km diameter nucleus would almost certainly survive intact, not
having had time to notice the atmosphere much before it hits the ground.

On diverting it:

In their book, posisbly called `comet' or maybe not, Greg Benford and David Brin
discuss two methods of steering a comet:

1) gentle thrust at aphelion, where a delta-v of a few dozen ms^-1 can make a huge
difference. This would require us to get out to rendezvous at aphelion in 2059 or
so. 

2) controlling outgassing at perihelion where huge forces are at work. Does
anyone know whether P/S-T meets Earth before or after perihelion?

Someone has already suggested a more dramatic variant of 2, using nuclear
explosions to create additional outgassing to order. This strikes me as the best
idea so far. Of course the earlier the better.