Jeffrey LewisDon't Shoot Down USA 193

After a getting multiple e-mails about my quotes in Global Security Newswire and the New York Times, let me make my advice to the Bush Administration, who decided to shoot down USA 193, explicit: DON’T FREAKING DO IT.

This is bad policymaking. Take this explanation of the rationale for the decision to intercept the falling satellite:

At the end of this, just from my perspective, what to me was compelling as we reviewed the data is that if we fire at the satellite, the worst is that we miss, and then we have a known situation, which is where we are today.

This logic — “hey, why not?” — is always suspect. It reverses the burden of proof, placing the emphasis on those who oppose the intercept.

Yet, this is an “extraordinary” measure (General Cartwright’s phrase) against a “small” risk (his phrase again). Justifying it requires demonstrating not just that one risk is greater than another, but that we have high confidence that estimates of the risks are accurate and complete.

Holding aside my general worry that this Administration is not to be trusted with sharp objects, there are specific reasons to be skeptical of both the accuracy and the completeness of this Administration’s calculations. I strongly suspect that they are systematically discounting two types of hard-to-quantify risks — the possibility of error within the estimates and the political costs to conducting an anti-satellite intercept.

Worst Case Scenarios

The Administration, according to General Cartwright, compared the “worst downside” of intercepting. But read carefully, and the “wost” case appears to be that they miss — not that they are wrong about the debris estimates or how much debris would reenter with a successful intercept.

The real worst case is that they are wrong about those risks. Yet the possibility of that error is difficult to capture in models.

On the other hand, the Administration did consider the worst case for hydrazine: that the tank remains in tact, that it comes down in a populated area, that people don’t evacuate:

The worst scenario is that you have a person who either is not mobile or does not, for whatever reason, sense that they’re in danger, and therefore doesn’t take any action. But those variables are very difficult to put minutes or time to.

The comparison here is apples-to-oranges: worst case on the human health risks, median-case on the debris risk. The deck is stacked.

The Hydrazine Story

Yet, there are reasons to wonder whether we know, or are simply concerned, that the hydrazine tank will remain intact. Administrator Griffin indicated that the “analysis that we’ve done is as certain as any analysis of this type can be” that the tank “will survive intact” and the “hydrazine will vent.”

Well, one might ask how certain can one be in this business? “I mean, one can never be certain,” Griffin said elsewhere.

It is worth noting that, on January 30, 2008, General Gene Reneurt, Commander of NORTHCOM, discounted the risk from the hydrazine.

The satellite includes some small engines that contain a toxic chemical called hydrazine which is rocket fuel. But Renuart said they are not large booster engines with substantial amounts of fuel.

Guess he didn’t get the talking points.

The idea that the hydrazine might survive apparently comes largely from the experience of the Columbia. That experience featured heavily in the presentation by Jeffrey, Cartwright and Griffin, as well as a story by Craig Covault indicating that data from the Columbia “are now being used operationally for the first time by Pentagon, NASA and NRO analysts to calculate better how much debris will survive…”

Whatever the probability that the hydrazine will survive and vent, it is less than 1. That probability must then be discounted by the probability that the tank will land near enough a person (let alone an immobile one) to cause human harm.

What About the Debris Risk?

We’ve had a great discussion in the comments about the risk from debris. Simple models suggests a very large amount of debris, much of which comes down within a few orbits.

We’ve seen, however, that the NASA model pretty seriously underestimated the amount of debris created by the Chinese ASAT intercept. Most of the debris will be very short lived, but the risk to space assets may be higher, at least in the very short term, than NASA imagines.

Those risks are not negligible — Administrator Griffin stated that “risks to shuttle and station … are at least a factor of 10 smaller than risks we take just being in space anyway…”

That, to my mind, is strange way of looking at the problem — we take many risks to be in space. Yet astronauts chose those risks freely for their own reasons. The folks sitting on the space station — and to a lesser extent those who have paid for it — are do not have a choice about the risk they will now incur.

We haven’t seen hard numbers yet, but my suspicion — and it is just a suspicion — is that the expected risk to the ISS crew will be surprisingly competitive with the expected risk to persons on the ground even without considering the possibility that our debris estimates might be low.

Of course, that’s because they both round to zero.

The Politics of this Suck

I don’t know how to express the political risk. Not knowing the risk, however, is different from it being “zero” — which is how the Bush Administration, at best, seems to count it. At worst, some members seem to assign a positive value to conducting an ASAT test.

The Chinese will use this to excuse their January 2007 test and, perhaps, future ones. The Russians seem interested in playing along, too. I’d like to be able to argue that they’re wrong; That this is different.

I have argued, in the past, that we have a strong interest in constraining the development of debris-creating anti-satellite weapons. Sadly, our intercept will make that outcome harder to achieve, not easier.

Given the extremely small risk to people on the ground, as well as the three people in orbit, these risks — though difficult to quantify — almost certainly should dominate the discussion.

But what loser is going to go to bat for confidence building measures in outer space when there is a giant tank of hydrazine bearing down on a Cub Scout Jamboree and one really awesome, heroic chance to blow it out of the sky? Hell, I bet the thing explodes into fireworks with red, white and blue stars and streamers like over the Mall on the Fourth of July.

Let’s face it, supporting the shot is the “safe” thing to do. After all, the debris risk will probably work out ok, while we’ll never know if the satellite would have hit a populated area. The cost, in terms of space security, is so difficult to identify, that one can simply explain it away with facile counterfactuals. “Oh, the Russian’s were just looking for an excuse, they would have done it anyway.”

Can Cowboys Do Math?

But the safe decision isn’t always the wise one. My sense on this was captured by my friend Jonathan McDowell, who called the decision regrettable and, in a moment of frustration, added:

Clearly someone in the administration who has the instincts of a cowboy has decided this is the perfect excuse to rattle our sabres and show the Chinese that we have the same capabilities.

Of course, Jonathan could be all wrong. The way for the Bush Administration to dispel his skepticism, and mine, would be to publish the risk estimates — both of the debris risk to the ISS and the hydrazine risk to the population on the ground.

Let independent observers check the homework.


  1. Lao Tao Ren (History)

    I am afraid that the political / strategic risks of this ABM / ASAT test are greatly underestimated.

    The least of my concern is the Chinese will just take this American ASAT / ABM test as carte blanch to do a lot more ASAT tests of their own. Developing a Chinese ASAT capability is bad for strategic stability and generating more debris, but there are even worst outcomes.

    Right now, the United States can sleep very soundly knowing that China has a desultory 20 or so land based missiles that have the theoretical capability of reaching CONUS. Paired with death trap Jin class Type 094 subs that don’t patrol, the Chinese threat is ‘just right’, not enough to be real serious, but important enough to justify US defense spending on many toys to counter the ‘threat’.

    By demonstrating a rudimentary terminal intercept ABM capability, it clearly signals to the Chinese that their nuclear deterrent is, at best, limited in credibility today, and in the future, likely to be not credible at all.

    That is a real good way to encourage the Chinese to start thinking about a Soviet style nuclear arms buildup to restore their nuclear deterrence (or minimal means of reprisal) capability.

    Do we really want China to get serious about defense spending on nuclear weapons aimed at the USA?

    I think not.

    Do the Chinese really want to have to deal with this problem?

    Every RMB spent on the nuclear deterrent takes away from China’s pressing needs to address real issues in raising living standards, deterioration in the environment, resource depletion, and other dire serious issues that the Chinese government face on a daily basis like electricity and petroleum shortages.

    Let’s not give the Chinese more reasons to get into an arms race with the US.

    Mutual restraint is in order here.

  2. CKR (History)

    Ah well, a couple of small tanks of hydrazine are just not the same as a frozen thousand-gallon chunk. I’m not the greatest heat transfer engineer in the world, but the experience I do have with heat transfer suggests to me that the big frozen chunk might well survive reentry.

    I’ve still been doubtful that that risk warranted the shoot-down, though.

    I will maintain that there may well be something on that satellite that our government doesn’t want others to know about. I don’t know whether it’s excessive paranoia about classified stuff or just stuff that shouldn’t be there.

    Physics Today asks whether there’s a plutonium power source on board.

  3. mark F (History)

    Except by ignoring its entire record, how could anyone believe that the administration has any concern about risk to any person “who is not mobile, and therefore doesn’t take any action”

    Whatever the result of the “shot”, it is reasonable to expect that it will be used to justify more money for the anti ballistic missile defence contractors.

  4. Lao Tao Ren (History)

    Post Script: Sorry, this thought got chewed up and omitted when that last message was written:

    The Chinese have been watching intently the Kosovo situation for quite some time in order to gauge how the international community will react to a similar event happening with Taiwan.

    The formal recognition of Kosovo by USA, Britain and France within days of a UDI was a rather unpleasant surprise to China.

    For those who are not area specialists, Taiwan is scheduled to hold elections on March 22, 2008, and at the same time, a provocative “Referendum on applying for United Nations membership under the name of “Taiwan”“ is being held at the same time.

    It goes without saying that China took note of the presence of UN troops in Kosovo which ensured that Serbia cannot take military action at will.

    Politely put, doing the shoot down now is certainly… bad timing.

    At the very least, it drives the Chinese to closer relations with the Russians.

  5. Muskrat

    A few days to go, apparently. I wonder if anybody in the administration is starting to think like a place kicker, facing the game-saving kick, who gets time out called by the opposition.

    As a test of NMD-like capabilities, this is as easy as it gets, right? Big fat target, days to prepare, the plans, materials and schematics of the target on file, no maneuvering, no decoys. How bad are they going to look if they miss on the first shot?

    Now, the actual military people involved are professionals, and I yield to no one in my admiration of them and their ability to perform under pressure — and no test is ever as much pressure as real combat. Still, I’m thinking there are some mid-level NSC or Pentagon staffers (and contractors) who are going to be on an all-Pepto-Bismol diet for a few days.

  6. yousaf

    There is a relevant paper at:

    Also, NASA had done estimates for a similar uncontrolled re-entry scenario when (purposely) de-orbiting CGRO:

    where it says:

    NASA to Deorbit Compton Gamma Ray Observatory

    NASA announced Friday that the Compton Gamma Ray Observatory
    (CGRO) will be deorbited June 3, coming down over the Pacific Ocean to
    prevent the possibility of an uncontrolled reentry in the future.

    NASA made the decision one day earlier to deorbit CGRO, which
    continues to return scientific data, after considering various options
    to deal with a gyroscope on the spacecraft that failed in December.

    While CGRO can operate normally with two of its three gyros,
    it would lose attitude control if either or both of the remaining
    gyros failed — an event with a 10% probability of happening in the
    next three years. This would eventually lead to an uncontrolled
    of the 15,000-kg (33,000-lb.) spacecraft in three to 11 years,
    depending on the amount of solar activity and its effects on
    atmospheric drag.

    The large size of the spacecraft meant that much of it would
    survive reentry and crash to earth. Officials estimate that 30-40
    pieces weighing more than 18 kg (40 lbs.) and up to one ton, as well
    as many more smaller objects, would survive reentry and crash at up to
    320 kmph (200 mph). Experts estimated a 1-in-1000 chance that such
    debris could kill someone if the spacecraft came down uncontrolled.

    this number is also reported by:

    Although CGRO was bigger and there was no talk of the hydrazine threat, I think this is useful ball-park number. It could be re-normalized to the orb. incl. of the spysat using the paper by Ailor & Patera.

    I think the real problem in this is assessing the hydrazine risk: I don’t know what the 27m contamination zone is supposed to be?? Is there supposed to be a chemical hazard if in there, or risk of getting hurt by a piece of debris??

    Bottom line: I believe the real risk prob. of death by debris due to an uncontrolled re-entry of the spysat would be significantly less than 1-in-1000 since the spysat (2500kg) is substantially less massive than CGRO (15,000kg).

  7. Mike (History)

    Everyone is spending a lot of time talking about risk. Risk is usually defined as the probability of the event happening times its consequence. What if they are defining risk as 1 minus the probability of doing nothing and its consequences? Then you have a small risk. The real problem in this case is in the uncertainty of the judgment of the risk (how big are the error bars). Are they Iraqi WMD big?

    The shuttle and space station are in a really bizarre orbit right now. The shuttle is there because it is at the station. The station is there so it can accept deliveries from the Kazak launch facility and US launch facility.

    Hydrazine burning should create a lot of N2, water and NOx, which should create a brilliant red streamer as the thing enters the atmosphere.

  8. Pavel (History)

    Columbia reentry data is not particularly relevant here – all its components were protected by the shuttle heat shield through (most of) the toughest part of the reentry. Columbia made it safely to about 64 km and, as I understand, the temperature at the heat shield was decreasing, indicating that the heaviest heat load part of the reentry had been already passed. USA 193 would have to withstand much heavier loads.

    Overall, the story about “frozen hydrazine” surviving reentry looks like BS. As the AW&ST article indicates, the pieces that can survive reentry are normally lightweight. There is a reason for that – large mass means a large ballistic coefficient, which means low deceleration, higher speed, and higher heat load. Basically, if there hydrazine in a tank, it will probably be going with high speed and will burn down. It seems that the only way for the tank to survive the reentry is to be empty and fall apart.

  9. spacemanafrica
  10. Mark Gubrud

    Coupla points:

    1) I can reliably estimate the risk of fatality if the hydrazine tank lands intact at one in a thousand or less. That is, if it does land intact, the probability is still 99.9% that no one dies.

    2) The risk to space objects from debris created by this test appears to be negligible, if not zero.

    3) The odds of an intercept that destroys the tank cannot be estimated better than 50-50, i.e. no information about a binary outcome.

    4) The cost of the shot will be the order of $100M and the payoff will be to change a one in a thousand chance of someone dying to one in two thousand, or one in ten thousand if you assume the mission has a 90% chance of success.

    5) The SM-3 missiles and the Aegis system have not been physically modified. The modified software can be downloaded to other fleet missiles and associated systems at any time. Therefore, if successful, this test will establish the Aegis TMD system as an operational, ASAT-capable weapon system.

    6) The LEAP interceptor reportedly has an axial motor or “third stage” which could propel it up to several km/s. This will not be used in the shootdown. It would give the SM-3 interceptor more crossrange reach or allow it to reach satellites at working LEO altitudes.

    So, this is an actual test of a real, live ASAT weapon system.

  11. Chap (History)

    I must admit that your choice of wording, “this Administration etc”, tends to make the rest of your argument sound less compelling to me. There’s a big bureaucracy you’re talking about, most of which won’t change when the CINC does.

    Some useful points might also be explaining who’s arguing what and where they’re coming from, or explaining who’s got what data and why you worry about it.

  12. Anonymous

    At the briefing Griffin said that the uncertainty of the time of re-entry will remain at ~10%. With the now announced Thursday planned date for the “shot”, and the anticipated ~March 7 re-entry — 14 days later, we get a ~1.4 day uncertainty as of Thursday. This corresponds to ~22 orbits worth of possible impact.

    The Columbia Investigation Board report shows a map of the debris field. While some pieces have found their way as far as 15-20km from the path projection, the vast majority have fallen within a 5km band. It is very reasonable to assume that the 1000lb, 50cm spherical fuel tank will fall very close to the projected path.

    The uncertainty of where the debris will fall is much lower than the briefing suggested. A rough estimate is that at the equator, only 5km*22=110km out of 40000 km (0.3%) are possibly in danger (albeit that danger is far greater). The number gets higher as the latitude get closer to the inclination.

    Whether or not the “shot” is attmepted, a map of the projected path needs to be disclosed ASAP, either by the government or by the wonks.

  13. J.Lo (History)

    It’s not surprising that a number of people are at a loss to gauge the fallout, both physical and political, of this decision. The law of unintended consequences, the dynamic of the unexpected when a more simple system attempts to regulate a more complex one, seems to follow space policy around like the kid in “Peanuts” with the cloud over his head. Without repeating what is already an excellent analysis, it’s worth noting that

    – a hit probably does more than create debris, including emboldening others (perhaps beyond China) to play Missile Command with their own toys;

    – a miss probably does more than embarrass the US, perhaps by exposing flaws in US systems and again emboldening other actors; and

    – doing nothing probably does more than simply risking an intact descent, including signaling irresponsibility in space or a reluctance to test missile systems (which carries a history), which might also embolden other actors.

    The thought exercise for unintended consequences is to take these and other possibilities as a jumping-off point to imagine scenarios that have not yet been imagined and push against the conventional wisdom on all sides.

    A final concern, and one from which I hope all will learn, is that in each of the above scenarios, there is a near certainty that any action (including no action) emboldens other actors. Having put the SM-3 on the table, the administration is almost compelled to use it. Once it’s in the New York Times, it gets increasingly unlikely that the Pentagon will say, “oh, that was just something we were talking about. We’re not actually going to do that.” I’m sure they knew what they were doing when they made it public, but it’s worth noting that doing so takes an option off the table. As another thought experiment, what might happen if the Pentagon decided not to intercept USA 193 and China executed another ASAT test next month?

    Unintended consequences are a bitch.

  14. Jeffrey Lewis (History)


    That’s a good point, as I understand that there was significant resistance to the decision within NRO and State, based on the debris risk and the politics, respectively.

    I wanted this post to be about the logic, rather than the bureaucratics, of the decision.

  15. Bruce Roth (History)

    This is probably a naive question, but is there any reason why astronauts on the space shuttle could not repair the satellite? If that is not feasible, could they remove the hydrazine, disable or dismantle the satellite, or just bring it down to Earth safely?

  16. Lao Tao Ren (History)

    Making the debris model and risk calculations public may not compromise very much sensitive knowhow.

    It would set a precedent to all other countries that if they intend on intentionally causing the creation of orbital debris, there is a standard to be met in technical analysis showing risk / cost / benefits.

    Revealing the data in a timely fashion before the shoot down would similarly set a strong precedent for transparency to international concerned parties and strengthen a consensus toward environmentally friendly use of space.

    The chance of this happening anytime before Thursday…. is probably lower than the risk of an individual being beaned by a tank full of hydrazine.

  17. Andrew Higgins (History)

    “Whether or not the ‘shot’ is attempted, a map of the projected path needs to be disclosed ASAP, either by the government or by the wonks.”

    The “projected path” following the intercept is increasingly looking like it will be over an obscure third world country. 😉

    Interesting analysis and discussion found here.

  18. Gridlock (History)

    But the logic is hard, if not impossible, to reconstruct given all the unknowns. From the satellite’s functionality to its power source to the tech on board we cannot possibly hope to assess the risk/reward calculation.

    That said, all we have to operate on is perception, supposition and interpretation. These being the same as other actors have access to, I have to say this is a dangerous precedent on the global stage.

  19. Andreas Persbo


    Here is a legal perspective:

  20. Juan-John (History)

    Be kinda funny if China decided to, “in the spirit of international space cooperation and safety,” shoot the doggone thing down itself before Thursday…

  21. yousaf

    Here is some info on hydrazine:

    In particular:

    Autoignition Temperature varies with contact surface:
    23C (74F) in contact with iron rust;
    132C (270F) in contact with black iron;
    156C (313F) in contact with stainless steel;
    270C (518F) in contact with glass.


    Sealed containers may rupture when heated. Vapors can flow along surfaces to distant ignition source and flash back. Above the flash point, explosive vapor-air mixtures may be formed. Sensitive to static discharge.


    The auto-ignition temperatures seem low to me; I’d naively expect such temperatures to be met at the metal-hydrazine surface of the tank during re-entry.

    As for the likely debris path post-intercept:

    http://www.agi. com/corporate/ mediaCenter/ news/spySatellit e.cfm

  22. Mike H

    Lets keep in mind this might not work. It is an enormous short-term engineering challenge for a missile not designed for the task, even if , in theory, it would be easier to shoot down a satellite than a missile.

    Flip around the question — presume a missed shot or two . What are the consequences? Are the consequences better than if the US did not take a shot at all, or worse? I could imagine reasonable people arguing either way.

  23. James (History)

    “The least of my concern is the Chinese will just take this American ASAT / ABM test as carte blanch to do a lot more ASAT tests of their own.”

    They have already demonstrated that they will do ASAT tests if they think it is in their interest to do so, whether we like it or not, no matter what we do or do not do.

    “By demonstrating a rudimentary terminal intercept ABM capability, it clearly signals to the Chinese that their nuclear deterrent is, at best, limited in credibility today, and in the future, likely to be not credible at all.”

    The Chinese already know that their nuclear deterrent would have limited credibility in a confrontation with the US, whether or not we actually test this intercept capability. It’s really kind of insulting to the Chinese to assume they aren’t aware of our missile defense programs, and that only the “signal” of testing an intercept capability will wake them from their slumber.

    “That is a real good way to encourage the Chinese to start thinking about a Soviet style nuclear arms buildup to restore their nuclear deterrence (or minimal means of reprisal) capability.”

    This test alone will not “encourage” Chinese thinking along these lines. What will do so is the entire US missile defense program plus China would need to make a political decision to confront the US. The state of their deterrent is just fine if they don’t plan to butt heads with the USA at some point.

    “Do we really want China to get serious about defense spending on nuclear weapons aimed at the USA?”

    If they do so it will certainly not be as a result of this intercept, but as a result of a larger political decision to confront the US. If China has not decided to confront the US, then they will regard this satellite shootdown with indifference.

    “Every RMB spent on the nuclear deterrent takes away from China’s pressing needs to address real issues in raising living standards, deterioration in the environment, resource depletion, and other dire serious issues that the Chinese government face on a daily basis like electricity and petroleum shortages.”

    That sounds like an excellent reason for them to do the rational thing, which is not to make that political decision to pursue confrontation with the US.

    “Let’s not give the Chinese more reasons to get into an arms race with the US.”

    More reasons? Right now they have no reason whatsoever to get in an arms race with the US that I can see.

  24. Lao Tao Ren (History)

    James, being old and cranky like me have the advantage of seeing things from a WWI perspective.

    I see striking parallels to 1905, where all the Great Powers were quietly preparing, arming, and building a system of interlocking alliances / contingency plans which created a tinderbox.

    War got started not because anyone really wanted to, but once the wheels got going (from some obscure event in the Balkans), it didn’t stop until after a second world war, a cold war, and more.

    If you were to really push it, I would say —- I am concerned about what might happen if the election in Taiwan delivered results that are provocative (e.g. the Referendum to apply to join the UN under the name “Taiwan” passed), and in the last days of office, Chen decide to make use of that and go past the ‘red line’ China have set, and in the last months of office, the Bush administration decide to play with fire.

    Maybe all of this is just the product of an overly active imagination —- but I fear that without some hard nosed, frank and candid exchange of views between China and America’s ‘hardliners’, and the development of an understanding of both side’s concerns and ways to address those issues peacefully, the risk of conflict between China and the US is increasing in the coming decades.

    On that note, I beg to differ with you that the Chinese will do things that are provocative (like the ASAT test) regardless of international concerns. They do respond to these concerns, though not necessarily in a way and degree that is expected of a great power with their present statue and capabilities.

    Back to the issue of appearances and politics: if they have no reason to get into an arms race with the US, why build a fleet of SSBNs? A land based deterrent does the job better, cheaper, and more securely. That is the kind of question that China need to answer and throw some light on. They are doing plenty of things that can be construed as provocative to the US.

    To reiterate: both the US and China are doing provocative things that can be taken by the other side to be threatening to each other.