2009年2月14日 星期六

Satellite collisions = Epidemic?

A satellite collision high up in the Siberian sky has generated a plethora of news in the last few days. I'm not going to belatedly repeat the story and information here. For background, please see the following:

First of all, I'm no expert in any fields. I'm putting forward the following idea only because I haven't seen anyone mentioning it and I think we might benefit from a change of angle, IF (that's a BIG IF) I got my logic correct. I'll let all the intelligent folks out there to decide whether this is an approach worthing pursuing.

The problem of two things colliding in orbit doesn't stop when all pieces are set flying. The collision produces a cloud of fragments, increasing the chance of later collisions, which produce even more fragments, further increasing the risk of collisions. This scenario of cascading collisions (see Kessler Syndrome) was contemplated by Donald Kessler, a NASA consultant, back in 1978.

Such collisions getting out of control strikes me as alarmingly similar to an epidemic outbreak. Thus, I'm wondering about the possibility of modelling satellite collisions as the spread of diseases. The disease model, if adopted, does not predict satellite crossing points, but can shed light on conditions that might lead to runaway cascading collisions and how best to avoid them.

It is true that there is much object-monitoring and collision-mitigating going on. Some may think that the latest collision is more an example of negligence than inevitability, and the disease model an overkill. However, let me quote a few segments from this article:

  • ... Iridium [the company to which one of the colliding satellites belonged] had been receiving a weekly average of 400 conjunction reports from the U.S. Strategic Command's Joint Space Operations Center that tracks debris in space ...
  • ... so the ability actually to do anything with all the information is pretty limited ... describing a kind of data overload.
  • Bryan Whitman, a Pentagon spokesman, said it was not possible for the U.S. military to track and predict the movements of all 18,000 objects in space all the time.
  • "Because there is so much, you have to prioritize what you're looking at," he said. "There are limits on your ability to track and compute every piece ..."
  • "We did not predict this collision," he said.

Apparently, with the number of objects up there, we can't control everything. In addition to 'micro-managing' all the orbital objects, it can be beneficial to take a more 'macro' approach and try to avoid catastrophe if something bad does happen.

In epidemiology, populations are divided into the susceptible, infectious, and recovered, and there are equations describing their dynamics. In orbit, satellites are 'susceptible' to being hit by orbital debris; when the hit is powerful enough to break up a satellite, the new cloud of resulting debris becomes 'infectious' (imagine a sick person carrying a cloud of virulent air around him); when debris gets picked up by some collection mechanism (if we do introduce such at some point in the future), it is removed from the environment, 'immuned' from further collisions. Lastly, for objects we CAN monitor, control, and thus in no risk of being hit in the first place, they are the vaccinated.

Admittedly, there are differences. The susceptible and infectious somewhat overlap, because a satellite has the potential to hit and be hit. And it might not be clear how to represent the cloud of debris: as many individual infectious agents moving in a formation, or as one agent that is somewhat more infectious than usual. I suppose the differences can be taken care of. (Then again, I'm no expert.)

Treating satellite collisions as human infections, we have:

  • the susceptible: anything that may get hit and potentially cause more hits after being hit;
  • the infectious: anything that can hit;
  • the recovered (or removed) and the vaccinated: anything that we can control and will not be hit.
  • even the dead: anything that re-enters the atmosphere.

To all you intelligent folks out there, does it work? Can we apply insights from epidemiology to cascading collisions?

Granted, it is premature to worry about a 'collision outbreak' at this moment. But 40 years ago, neither would anybody worry about two satellites colliding.

沒有留言:

張貼留言