click on the image for a larger version
Cataloged debris associated with the 10 February 2009 collision of Iridium 33 and Cosmos 2251 as of 23 October 2009. The orbits associated with the major pieces of each satellite are indicated.
As we wait for the next shoe to fall in the Iranian nuclear crisis, I feel a real need for a short break to consider something completely different at least for a little while. The evolution of the debris from last February’s collision between a dead Cosmos satellite (Cosmos 2251) and Iridium 33 always represents an interesting and important digression. First off, here is the score card as of 23 October 2009:
Debris Associated with Iridium 33’s Orbit:
Cataloged: 484
Decayed: 17 (3.5% of those cataloged)
Debris Associated with Cosmos 2251’s Orbit:
Cataloged: 1102
Decayed: 35 (3.2% of those cataloged)
Qualitatively, the plot of the debris positions shown in the image above shows that the debris associated with the cosmos satellite’s orbit has precessed to a greater extent around the globe than that associated with the Iridium’s. This does not represent a wider angular distribution to the pieces following the cosmos’ orbit but rather a significantly greater change in the ensemble’s average speed than that associated with the Iridium. (It appears that observational biases eliminated those pieces with large angular differences from the catalog, at least it keeps them from being associated with either of these two satellites.) This difference in orbital speed distributions is shown below, which I made several months ago when there were far fewer pieces cataloged:
click on the image for a larger version
The orbital speeds of the collision debris. The arrows indicate the original satellites’ orbital velocities.
Why did the pieces associated with the Cosmos’s orbit end up with such a significantly greater shift in orbital speed than for the Iridium’s? It would be tempting to say that this asymmetry is somehow related to the geometry of the collision but it is hard to imagine how that asymmetry could arise. Perhaps the Cosmos passed through a portion of the Iridium that had relatively light components and that “all” those were bounced into the Cosmos’s orbit or at least into trajectories that made large angles with both orbits and were not associated with the collision. Those pieces that ended up in the Iridium orbit might have been broken apart by the shockwaves that traveled through the satellite after the collision. That might explain the factor of roughly twice as many pieces in the Cosmos orbit as in the Iridium orbit. Such “delayed” breakup is probably not properly modeled in NASA’s computer programs that predict the numbers of debris created (at least not in my limited understanding of those programs, though I could be wrong about this). In any case, this collision should reveal interesting phenomena is hypervelocity collisions of extended objects. Phenomena that will be important for, among other things, missile defense where the increasing emphasis is the destruction of specific components of incoming warheads. (Richard Lloyd has an updated version of this book but for some reason I cannot find a link for it.)
Update (one minute later): It looks as if the other shoe has indeed fallen with NPR reporting that Iran has reneged on the TRR refueling deal.
Up-update: First media reports are always so problematic. Other reports now say that Iran will respond to the deal next week. Thanks for pointing this out Josh! ( Here is a more “official” statement.
Yale Simkin | October 23, 2009
Is this the update?
Geoff Forden | October 23, 2009
I think Richard has had a book come out in the last couple of years. Unfortunately, my copy is at my office and Im working at home today so I dont know the details.
Yale Simkin | October 23, 2009
To see the extremely freaky “Iridium Flares” from non-obliterated Iridium satellites check the links here
Geoff, has there been any collisions noted between fragments in the debris streams?
Geoff Forden | October 23, 2009
Yale brings up an important point and one I considered including in this post: how much has the increased density of debris so close to Iridium orbits increased the chances of other satellites being struck. As of today, no new Iridium satellites have been hit though I have heard anecdotal reports that there has been an increased number of orbital maneuvers. As for debris-on-debris collisions after the initial event, I dont know that we have any way of knowing that.
Azr@el | October 23, 2009
Could it be possible that most of the of the alleged increased orbital burns to avoid the Ir-Mos or Cos-mium debris field, are in fact safe side hedges as opposed to imminent dire cases. Thus if a tracked debris will pass within a certain circle of radius r_s whose normal line thru its origin is its orbital vector. r_s would be derived from the time averaged projection of the surface of the satellite with respect to its normal in the direction of orbit. Computational and observational errors will inflate r_s by a safety cushion. But the orbital element of the debris must also contain these errors which are further compounded by the fact that it’s a low mass object of unknown composition and angular momentum, subject to potential boil off, fragmentation or magnetic interaction all of which could cause it to be a curve ball, thus to model this they no doubt assign a large bounding r_d to the debris field’s nominal orbital element. So instead of trying to keep two rather discrete lines from intersecting you have a situation where you’re attempting to avoid to one small, given by r_s, and one large, given by r_d, flyswatter from intersecting. Thus all of these alleged orbital maneuvers could be attributable to a cautious, yet wasteful, policy of giving a wide berth to debris elements as opposed to avoiding dead on collisions.
On the subject of the Iran update; how can a party renege on an agreement it hasn’t acceded to? Someones’s been reading a bit too much 1984; double plus ungood.
John Schilling | October 24, 2009
The fact that Cosmos 2251 was substantially larger than Iridium 33, is probably sufficient to explain why there is substantially more (trackable) debris in the Cosmos 2251 orbit. Indeed, there’s actually a shortage of debris in the Cosmos 2251 orbit, if we go just by the mass ratio. It would appear that each trackable piece of Cosmos 2251 is about 45% heavier than each piece of Iridium 33 debris, assuming similar fragmentation patterns.
It’s important not to try and apply too much common sense to this; hypervelocity impacts don’t work at all like the more common sort. Both vehicles will have been substantially fragmented and partially vaporized by shock waves in the immediate aftermath of the collision. The fragments will have been accelerated by the shock waves, by entrainment in the expanding vapor clouds, and by occasional fragment-fragment collisions.
And yes, a fair bit of Iridium fragments will be entrained in the mostly-Cosmos debris, and vice versa. Surprisingly little will wind up anywhere else.
If Cosmos 2251 was a more heavily constructed satellite whose fragments tend to be heavier than those of Iridium 33, and a smaller fraction of the vehicle literally vaporized, it is not at all surprising that the Cosmos debris was not subject to quite so great an acceleration in the immediate post-collision environment than the Iridium fragments.
Geoff Forden | October 24, 2009
I waited quite awhile after making those plots of orbital speed, the single most important aspect of the collision, thinking that it might still reflect an experimental bias. Now, after about nine months, I believe it is a real effect.
I guess we all have our ways of understanding hypervelocity collisions. I’ll stick with mine.