Hacker News Reader

Comment by lr1970 | Hacker News Reader

lr1970

> The whole approach is very sensitive to the detonation height, asteroid composition and color, rotational characteristics, etc. So Burkey’s group has really made a simulation framework for modeling the right thing to do for a given asteroid.

And don't forget that most asteroids are spinning over its own center-of-mass. The thrust generated by evaporating rocks on one of its sides will be changing direction as the asteroid spins around. Think of a rocket with thrust but no stabilization. Cool idea but very hard to make practical.

View on HN

Given the dimensions involved, the billions of miles, any deflection would create a miss. If we are certain that an asteroid is actually going to hit earth many years/decades out, then even a microscopic nudge in any direction would shift the closest approach by an earth-width. So precise aim might not be as important as getting the bomb out there asap. The earlier it is done, the less push is necessary to generate a miss. Of course, this is premised on us having near-perfect knowledge of the initial trajectory.

Detecting an asteroid years or decades in advance in the first place would be an extremely difficult task, I believe.

As I understand it we are tracking certain asteroids that we predict have a chance of impacting on the following encounter with earth depending on what the outcome is of passing through a per-basis volume called a "keyhole". So it isn't that we detect them when they are a decade out on a direct trajectory.

Those are the ones we know about. If you think about the problem, it really is a very difficult one to solve. We're attempting to find and track tiny objects that could be anywhere in an absolutely massive volume of space using almost entirely ground based equipment that is hobbled by competing demands, atmospheric distortion and a very inconveniently nearby star that halves the observation time and makes it really hard to find objects on that side of our orbit. For all the fancy animated maps of the many asteroids we have located there are likely still a very large number of other asteroids we don't know about yet and may not find out about until after the fact.

Oh certainly, from that angle yes it is a massive undertaking indeed!

Heaven forbid their calculations aren't confident given that large distance, and the theoretical nudge instead generates a hit.

A chilling tale, but step back and look at the odds: Not only does it rely on a significant original measurement/prediction error, but it also assumes the exact wrong nudge will occur that happens to compound that error.

In contrast, the vast majority of possible nudges will be neutral or beneficial, since the Earth is a very small target compared to the rest of space.

Why would we be nudging ones that are going to miss?

If we mistakenly think it will hit because "their calculations aren't confident given that large distance."

Presumably the same distance that gives the calculations uncertainty allows time to adopt a wait and see approach, refine the data and calculations and then make a decision based off of stronger inputs and still avert disaster if necessary?

Probably better to hit it twice, the first time right after detection and a second time if the probability of impact is still too high.

Every tumble has a "slow" pole, where the outgasing would form a vector, nearly almost away from earth.

Also at least we cant have a ice-ball earth due to impact-winter.https://en.wikipedia.org/wiki/Snowball_Earth

So if the universe threw a perfect spiral at us, we'd be fucked

No because the vector can also change in magnitude. Even if the spiral was perfectly aligned to impact the perfect center of the Earth we could change the magnitude of the object by nuking either pole to either slow it down or speed it up. Enough energy on any pole could make it miss regardless of the orientation.

If you can make the asteroid cross Earth's orbit 430 seconds sooner or later that's an Earth diameter.

Two approaches to getting that number...

https://www.wolframalpha.com/input?i=diameter+of+earth+%2F+%...

https://www.wolframalpha.com/input?i=%28430+seconds+*+earth+...

Granted, you'd probably like more than a diameter... but we're talking about it getting there a few minutes faster and it would be a miss.

Perfectly chaotic rotation might be worse. If it was a spiral then you could hit it at a pole. I think most things rotate about a semi-fixed axis (not totally fixed, often propagating over time)

> Think of a rocket with thrust but no stabilization. Cool idea but very hard to make practical.

Given that the asteroid was heading directly at Earth prior to disruption, wouldn't any change in trajectory be welcome, even if it's chaotic?

Unless it's spinning at 100rpm the rotation would have very little impact, the whole x-ray-heatup-boiloff cycle is under a second.

The paper only simulated it up to 10 microseconds. But they also only used a simplified 2D toy model. They acknowledge that a realistic simulation will need to factor in a ton of other things, including rotation of the asteroid.

The impacted side might outgas for years.

In that case, detonate the bomb above its axis of rotation. Unless the rotation axis is precessing as well...

Reply on news.ycombinator.com