Jeffrey LewisShoot-Look-Shoot

I have a column up at Foreign Policy (“Billion Dollar Baby”) on the announcement by newly installed SECDEF Chuck Hagel that 14 additional ground-based interceptors ought to do it with regard to North Korea.  I noted in the piece that the focus on North Korea led to a comparative neglect of the decision to cancel Phase 4 of the European Phased Adaptive Approach intended to provide an early shot at an Iranian ICBM. (Not everyone ignored it, of course, but most of the major daily papers did.)

After I submitted my column, the Washington Post wrote a strange (even by its standards) op-ed lamenting the loss of Phase 4, explaining:

In doing so, the administration has eliminated the possibility of a defensive system that would give the United States two shots at an Iranian ICBM — what in Pentagon jargon is called a shoot-look-shoot capacity. [sic]

Actually, no.  This is pretty much all wrong.

Let’s just observe that, as a factual matter, the earth is round.  As a consequence, the interceptors based in Alaska (near the north pole) do have a chance (in hell) of intercepting an Iranian missile.  The current doctrine is to salvo-fire five interceptors at each target.  That is five shots, not two.  The reference to shoot-look-shoot, which seems like plain English to me, refers to the ability to shoot at the incoming missile, then look to see whether the first shot killed the missile before expending a shot again.  Shoot.  Look.  Shoot (if necessary).  The improvement in efficiency is obvious, but what that takes is time — sweet, precious time.

(I should also mention the two slightly different definitions of shoot-look-shoot.  In the second version, you fire additional interceptors if the first shot gets close enough and sees more than one credible target.  We don’t configure the kill vehicles that way, but we could.)

First, Phase IV of the EPAA might not have provided a shoot-look-shoot capability.  The report of the National Academies is a little coy about this, because I think the situation is pretty borderline and delves into the classified realm.  The basic problem is the Alaska is far enough away on a curving earth that the window to shoot at incoming Iranian ICBM is pretty narrow — probably right around the apogee.

I would encourage the collective you to make some formal models, but I suspect the answer will depend very much on the assumptions in the model — assumptions that are hard to make given the fact that neither the SM-3 IIB nor the notional Iranian ICBM yet exist.  The uncertainty is probably enough, however, to discourage heavy investment in Phase 4 for the purpose of gaining a shoot-look-shoot capability, especially given the alternatives.  (Choices are always about the alternatives.)

I make the apogee something like 6,000-7,000 km from the Fort Greely site. (I note the test range maxes out at 8,000 km, which I suspect probably bounds the worst-case Iranian engagement for it.)  That’s a long, long way for an interceptor to travel, which means that the interceptor is looking at well in excess of 10 minutes of flight time, maybe even in excess of 15 minutes. If the engagement window from Alaska is around 1,800 secs into the flight, then the interceptor might need to be launched as early as 800 seconds after the Iranian launch.  One might be able to fit in a full engagement cycle in the first 800 seconds, but I think it’s pretty close.

The tight kinematics mean that it would not be responsible to rely on Phase 4 of the EPAA for a shoot-look-shoot capability if there are better options.

Then there are the other criticisms leveled by the National Academies:

(1) Phase 4 is not necessary for European defense (and therefore need not be “in” Europe), (2) that the rationale of an “early” intercept as a solution to the midcourse discrimination problem is an illusion, (3)  that Iran might overfly the site in Poland, (4) that the National Academies concluded the interceptor likely couldn’t achieve 5 km/s in the VLS launcher volume, and (5) that an East Coast site with a new interceptor, new radars and new concept of operations would be much, much better.

The editors at the Washington Post simply skip over all this, as well as the fact that the Administration is starting the EIS process for an East Coast site (under Congressional duress).  As I have noted before, an East Coast site is a terrible idea unless it is a packaged with the other changes recommended by the National Academies.  If the editors of the Washington Post took ten minutes to think through this, they might have written an op-ed usefully advocating for the broader set of recommendations in the National Academies report.

Instead, I get the feeling that they wrote the editorial as a sop to certain Republican politicians.  “See, we can be critical of the Administration!”   One of my complaints about editorials in general, but also bipartisanship, is that we tend to compromise on outcomes in ways that have no inherent strategic logic.  So, for example, if the President says he will end nuclear testing and some opponent calls for seven tests next year, the Post might write in favor 3 1/2 nuclear tests — a nonsense solution that achieves nothing other than standing in the middle.

Rather than compromising on outcomes, what we ought to do is try to make policies that address the concerns expressed by the other side of the argument. I realize that can hard to do when much of official Republican foreign policy boils down to “I hate the President and his works,” but there are certainly conservatives who speak in complete sentences and chew with mouths closed.  (Even if they eat babies. Kidding!)

What the Post might have done is observe that conservatives are not wrong to assert that we have a real (if still emerging) challenge from North Korean and Iranian missiles.  In that context, one might certainly explore the best architecture to meet that challenge, without breaking the bank. I still think back to a conversation I had when I was a young RA at CSIS, with someone who would not mind one bit being described as a “neoconservative Star Wars fanatic.”

He told me he was opposed to space-based defenses.  I was stunned.

“It’s simple.  I want missile defense to work.”

Liked him ever since.

Comments

  1. yousaf (History)

    “The basic problem is the Alaska is far enough away on a curving earth that the window to shoot at incoming Iranian ICBM is pretty narrow — probably right around the apogee.”

    Ted and I looked at this in fig 8(a) and (b) in our report for FAS — Sept 2011, and reproduced in yesterday’s post at the Strategic Security Blog at FAS.

    GBIs can engage notional Iranian/Russian ICBM at post-apogee midcourse.

    • j_kies (History)

      Properly speaking, GBI’s are not kinematically prohibited from intercepting items on those trajectories. Engagement is driven by sensing and tracking capabilities and the forward propagation of the error states to the GBI acquisition timeframe. With poor tracking the GBI’s may never have the threats within kinematic access of interceptor divert in the end-state.

  2. j_kies (History)

    Jeffrey

    I tossed you a snotty softball response in the spirit of your mocking the public announcements in the current climate of test results. Its pretty fair to state a technical opinion that the prior ‘leadership’ of the MDA was an epic fail. (also ran the GMD program during CE-2 engineering)

    To get the gut of the NAS/NRC report value, you should have a copy of the 2004 APS study at hand and run some kinematic cases with reasonable sensing timelines.

    It’s also safe to assume none of the critiques / accusations from the Bulletin of Atomic Scientists OpEd on the NAS/NRC report appear valid (and the authors should publically retract the OpEd and apologize).

    BMD is a difficult problem but doable with sufficient scientific and engineering diligence, the offense is stuck with the same physics as the defense. The real question – does the nation chose to make that investment rather than accept the consequences of mistakes or failed deterrence.

    • Jeffrey (History)

      It doesn’t seem snotty to me. I agree with all the things you’ve said … and have my little copy of Mosher’s masterpiece right here.

      Not kinematically prohibited, but everything depends on the radars. Yes, that’s right. My point was that Phase 4 wasn’t necessarily going to give us an SLS capability.

      “Leadership” in name only. I agree here that we’ve gotten less than our money’s worth given how these programs have been run.

      We might disagree about how doable it would be to keep up with countermeasures, but I don’t have any objection to making investments in technology to allow us to see one way or the other.

      My objection is to spending money on GBIs, not investing in missile defense generally.

    • Jeffrey (History)

      I put a comment on the Foreign Policy site, but I’d rather have that conversation here where I can block the trolls.

      j_kies. I would add that there is some evidence that the KN-08 is in the field, although I find the Administration’s refusal to say so infuriating.

      I have stated before that I believe the warhead is under 1,000 kilograms, see: this and this.

      So, I do believe that North Korea is moving much faster toward an operational nuclear capability than most people think and that the result, in the short-run, will be a rougher neighborhood.

      I support the NAS/NRC recommendations, with the stipulation that I would not commit to an East Coast site until we had more data about the feasibility of GMD-E approach or Iran demonstrated more on the ICBM front. (I am being sort of evasive about “more data” and “Iran demonstrated” because I’ve been talking with colleagues about what those thresholds might look like. Suggestions are welcome.) Absent one or the other, the danger is that we end up with another site of GBIs that provides no defense.

      As you will see from the links above, I am very worried about preemption by the South Koreans. I don’t think 14 GBIs solves our problem there. And, I think there are basically no realistic scenarios to use nuclear weapons on the Korean Peninsula, though that might not be the declaratory policy I would choose at the moment. (I might observe that they wouldn’t be necessary, then ask how to say “Fuhrerbunker” in Korean.)

      I think we are left with a credible conventional deterrent to a large-scale war (which North Korea will lose, conventionally) and a not terribly credible deterrent to a North Korean use of nuclear weapons in desperation.

      We also have a conventional provocation problem, which sucks. We don’t have a magic wand, however, that can make it go away. Again, 14 GBIs doesn’t get us there. Sometimes the best policy is face up to what you can and cannot do.

      I’d probably spend some of that billion on theater defenses and the rest on the first steps toward the NRC/NAS midcourse architecture. I’d also be honest with the South Koreans about what we can, and cannot, do. Some will say the South Koreans will respond by building a bomb, but we offer a lot of conventional capability and we can point out our friends in Israel have one yet remain worried that a nuclear Iran will engage in all manner of provocations.

      Allied containment has risks. That’s life. I try take, as my general attitude toward defense problems, that of Michael Howard as he surveyed the Soviets:

      “How are we to deal with this problem? How are deterrence and reassurance to be once more reconciled? This is the task that will confront statesmen and strategists for the rest of this century.

      [snip]

      Our first task must therefore be to get Soviet power and intentions into perspective. The exaggerated melodrama implied in the term ‘The Soviet Threat’ seems and has always seemed to me unnecessary and counterproductive.

      [snip]

      One does not have to attribute to the Soviet Union either predatory intentions or ambitions for global conquest to persuade all but a stubborn minority that the states of Western Europe have a problem of military security that must be solved if normal intercourse with the Soviet Union is to be sustained on a basis of equality.

      [snip]

      The second task therefore is to show that Europe can be defended, and that the costs of doing so would not outweigh the benefits.

      [snip]

      So where does this leave us? First, the requirement for effective deterrence remains, if only because the Soviet Union cannot be expected to observe a higher standard of conduct towards weaker neighbours than other states, whatever their political complexion, have shown in the past. Second, deterrence can no longer depend on the threat of a nuclear war, the costs of which would be grotesquely out of proportion to any conceivable benefits to be derived from engaging in it. Third, proposals to make nuclear war ‘fightable’, let alone ‘winnable’ by attempting to limit its targets and control its course, however much sense this may make in the military grammar of deterrence, are not persuasive in the political language of reassurance. And finally the problem cannot be solved by any massive transferral of resources to conventional capabilities. The immediate social costs of doing so, whether one likes it or not, are unacceptably high.”

      With the exception that, in fact, I do believe the optimal allocation of defense dollars would provide a very substantial capability against North Korean above what we might have been able to achieve against the Soviets at a reasonable level of defense expenditure.

    • j_kies (History)

      Jeffrey – that was a really nice exposition to clarify things over on FP – Your heartfelt acceptance of the NAS/NRC view is singing the Coyle tune of ‘MDA needs to have competent personnel for hard science / engineering to make stuff work’. Your response feels different than the orginal piece and you could productively revise the original with that material.

      As to the topic of your post here

      In addition to kinematic timeline, SLS has the amusing need for 2 things that don’t appear to be in play 1) Sensors to view the engagement (placement?) and 2) the Kill Assessment function in the Command and Control.
      Kill Assessment was in play back in the 1994 GAO Homing Overlay ‘fraud’ investigation and report (hypervelocity impact phenomenology and observables) and MDA used to have a R&D program …(Check the Congressional R2’s)
      http://www.smdc.army.mil/FactSheets/KillAssessment.pdf

    • Jeffrey (History)

      It’s funny you say that. I was worried about that outcome so I shared the draft with at least one person who … well … lets say that I had ample reason to believe my (qualified) endorsement of the NAS/NRC approach was clear.

      Part of the reason for this blog post was to reinforce my point that we should take the KN08 seriously, while also stating that adding 14 GBIs doesn’t count as taking it seriously.

      In fact, I think it is worse than doing nothing. As I wrote to a colleague who is far less worried about the accidental nuclear war scenario:

      First, it costs a billion dollars. That’s a lot of money right now and it will come from somewhere. My view continues to be that we ought to spend our defense dollars as best we can, with a heavy emphasis on training and readiness, then explain to our allies how we’ve done the best we can. I am prepared to have a conversation with them about ensuring that our operational plans are consistent with their needs (no falling back to Busan, etc.), but I cannot accept the idea that deterrence or reassurance is enhanced by spending money suboptimally in search of making another world leader “feel better.” Madame Park needs to put on her big boy pants, so to speak.

      Second, spending $1 billion on GBI and starting the third site now will choke off any hope of the technological transformation outlined in the NAS/NRC report. I am still more skeptical than they are that we can stay ahead in the countermeasures game, but I am persuaded that it is worth a reorienting our a significant portion of our current investment in missile defense to see if we can move toward the sort of SLS approach they’ve outlined given that nothing else will work.

      Third, the short-term benefit of the buy will pass as soon as the North Koreans stage another provocation to which have no answer. My concern remains that the North Koreans are in fact moving toward a real operational nuclear deterrent that will enable them to repeat events like Yeongpyeong-do and the Cheonan. The GBIs won’t enhance our willingness to initiate a major conflict with North Korea over these sorts of events, so as soon as they shell another island the benefit of the buy is gone.

      I am now persuaded on the basis of the North Korean test, as well as the implicit confirmation that North Korea is putting the KN08 out into the field, that we need to think seriously about a thin but resilient homeland defense. I am skeptical that the GMD-E system will, in the end, provide a cost-effective defense but I would be prepared to spend the money from within the current budget to see and, if it does provide a defense against the DPRK, would deal with the Chinese consequences (which are by now probably programmed in anyway).

      But I could not advise a US policymaker that the current the GBI system in Alaska (and VAFB) — at any level, either the current 30 or the planned 44 — enables him (or her) to initiate a conflict with North Korea that has deployed nuclear-armed KN08 ICBMs. I’d like to believe that Obama, Hagel and the rest understand that this is not a capability that can be used, but they are publicly professing otherwise. So, my goal is to mock them for it.

      Just like I mock Obama for suddenly deciding to wear a flag pin and the million other little things politicians do in the name of pandering. If, in private, they are thoughtful and sensitive guys with reasonable opinions, rather than the raging egomaniacs I see on TV, all the better!

    • Jeffrey (History)

      I did say in a preceding column that dealt with the NAS/NRC report at length:

      “One has to give the National Academies credit for proposing a complete redesign of the current architecture — new interceptors, radars, and a concept of operations — to deal with the problem. For years, many of my colleagues have been arguing that the Missile Defense Agency has systematically ignored the challenge posed by countermeasures. While I am not sure I share the National Academies confidence in the ability of the United States to stay ahead in the countermeasures game — particularly in the current era of budgetary austerity — this is the first missile defense study to take the problem seriously.

      [snip]

      Given the challenges associated with countermeasures and the far off prospect of an Iranian super-ICBM, I am inclined to think we could hold off on the East Coast site for now, while making many of the programmatic changes suggested by the National Academies panel, as well as entertaining one other modest notion.”

      The “modest notion” was returning missile defense programs to the services, which I take as an implied recommendation of the mandate-constrained NAS/NRC report.

    • j_kies (History)

      ” I would add that there is some evidence that the KN-08 is in the field, although I find the Administration’s refusal to say so infuriating.” … really? where is the ‘evidence’ and does it require consumption of Hallucinogens to see it?

      I am a really boring guy, I restrict myself to physics constrained rational approaches with US and Soviet developments as a guide to where ’60s and ’70s technology go if you know what worked.

      I really like the stuff that Markus Schiller and Robert Schmucker did as it has the attributes of analysis that I like. Perhaps Mr Gates as a historian got a bit carried away and the present folks are backing off from those ‘assessments’.

    • Jeffrey (History)

      As I state in the piece, Clapper said the DPRK was moving to field the KN08. I presume the leak to Sanger and Shanker, which they garbled, was supposed to provide the backstory.

      I don’t know if the claim is true or not, but if we’re going to spend $1 billion, I’d like an IC judgement to discuss.

    • George William Herbert (History)

      Jeffrey wrote:
      Second, spending $1 billion on GBI and starting the third site now will choke off any hope of the technological transformation outlined in the NAS/NRC report. I am still more skeptical than they are that we can stay ahead in the countermeasures game, but I am persuaded that it is worth a reorienting our a significant portion of our current investment in missile defense to see if we can move toward the sort of SLS approach they’ve outlined given that nothing else will work.

      I think I would object to this.

      Strategically, BMD is a numbers game. We have a GDP of about $15 trillion, NK has a GDP of about $32 billion. At any realistic cost ratio of interceptor to missile (3:1, 5:1, 10:1) we can afford to buy enough GBI interceptors to shoot down every missile they could build.

      Playing BMD as a defense budget zero-sum game is a mistake. It’s a national strategic component. We should commit to spending what’s required to defend against the maximum credible NK strategic attack. They should never be in a situation where they think that they have enough missiles to push through our defenses.

      The reality of a mass attack and possible leakers if we don’t have enough time/kinematic envelope/sensors/sites to effectively shoot-look-shoot against a NK threat means that even that effective defense level is not “100% guaranteed”, but we should make it clear to them that they are not a partner we consent to allowing to enter a strategic MAD deterrence relationship.

      We have consented to strategic MAD with the Russians and China, by the size of the BMD system we already built. But we do not have to, and should not, with NK. Spend them into the ground. We have the billions. That’s what the world dominant economy is FOR, in a strategic sense.

      This is not an argument to ONLY spend on GBI; better more cost affordable options are appropriate and necessary. But we should commit to overmatching any nearterm deployments they go for, and denying them credible MAD or first strike capability against the US.

    • Jeffrey (History)

      What you say sounds reasonable, but let me press you a bit.

      We talk a lot about accepting mutual deterrence. The reality has always been that the DPRK has had a conventional deterrent that induces caution on our part. I teach the 1969 EC-121 shoot down in my “Security and Arms Control in Northeast Asia” class precisely because it is such a clear exposition of how the DPRK got away with the Pueblo, multiple assassination attempts on Park, the EC-121 shootdown, killings in the DMZ including the ax murder, the assasination attempt on Chun in Rangoon that killed his cabinet, the plane they blew up, etc. The DPRK always gets away with it because of their conventional ability to make war. Nuclear weapons make that problem worse, but let’s be honest that we have a mutual deterrent relationship. I don’t like it, but that’s what it is.

      Now, I agree with you that “strategic stability” is not the goal here. My objection to missile defense in the Russia context is that it is madness to think we can escape from vulnerability and that trying to do o will make things much, much worse. The DPRK is a closer call. If we can defend against the DPRK, I say do it. And, to make good on that premise I would give the NAS/NRC architecture and CONOPS a try.

      But we might not succeed. Sure maybe a 1:1 spending ratio favors us, but there is surely some ratio at which we can’t keep up. We’re richer, but we also have to plan for two road games while the DPRK gets to prepare one home game. At a billion bucks to defend against three missiles, I think we’re hosed.

      I can see the DPRK deploying 100 ICBMs. (Let’s hope they don’t figure out how to make their own TELs.) I struggle to see us deploying 500 GBIs. Moreover, the 97 percent intercept probability looks a lot less appealing against 100 missiles than it does 10. We haven’t started on countermeasures yet. At the end of the day, I am not convinced that we can spend enough to keep ahead. A thin defense might still be worth a modest investment, but lets not kid ourselves.

      My point is that dealing with the DPRK’s nuclear forces is going to be tough enough. We simply can’t be pissing away a billion dollars on 14 GBIs as a political statement.

    • j-kies (History)

      WOW –
      George bless you

      Exceptionally well stated and I will shamelessly quote you in every forum where I can.

      As to the countermeasures issues, I laughed my way through the 2000 ‘report’; lets just say that the 1999 NIE was hopelessly misinformed as to the ease of defeating BMD sensors. If we are afforded the opportunity to define the sensor construct to enforce observability the target ID problem can trivially be ‘brute forced’. So that economic argument again plays to US advantage.

    • yousaf (History)

      George — NK may wish to consider a less attributable method of delivery were they truly interested in a first strike: e.g. a sailboat or fishing boat. Smaller risk of attribution, no need to miniaturize. Perhaps get Waikiki instead of LA?

      Is it sensible to create incentives for NK to do so? I’m more comfortable with them placing nukes on missiles — instead of being dissuaded from doing so. We can tell where missiles came from which might make them think twice.

      BMD, *if* vastly technically improved, could conceivably dissuade primitive missiles w/o decoys as a delivery method — but does nothing to dissuade possession of nukes.

      Not sure that won’t make the situation substantially worse.

      Anyway, *if* you still wanted BMD against a small nation like NK — despite the downsides mentioned above — I’d go with Garwin’s suggestion of boost-phase from surface ships.

      Personally, I’m happy not to incentivize them to put their nukes on boats. (And I like my tax dollars in my pocket.)

    • Mark Gubrud (History)

      I am simply astonished at this exchange.

      J_kies, for “a really boring guy” who claims to base everything on physics, you display here only a theatrical gift for sounding like you might know what you are talking about, plus exactly zero physics or physics-based argument. In fact, it’s rather hard to discern what point is being made (if any) in much of your verbiage above.

      With regard to Ted Postol and George Lewis’s outraged response to the NRC boot-phase “study” which they show to have been technically flawed and exempted from any responsible peer review process, you say that “the authors should publicly retract the OpEd and apologize.” But they haven’t, and their complaints about the process are factual, referencing correspondence which they make available, while their technical arguments are explicit, physics-based and publicly available. Where are yours?

      A key point is that straightforward E&M calculations on a cone-shaped warhead using commercially available radar-absorbent materials show that the warheads would be invisible to the NRC panel’s proposed radars at ranges where the panel expected the radars to be capable of tracking the warhead, and that discrimination would in no case be possible, whereas the panel expected the X-band radars to play a key role in discrimination. And by the way, this is before you even think about the use of balloons, decoys, antisimulation, chaff, etc.

      With regard to the 2000 Countermeasures report, co-authored by Richard Garwin (who also joined in protesting the 2012 NRC travesty), you report having “laughed” but do not tell us anything, not even a hint, of what you might have found so amusing.

      You tell us that “If we… enforce observability the target ID problem can trivially be ‘brute forced’.” OK, how? By using nuclear weapons to strip off the balloons? “Brute force” fantasies were the stock out of which the SDI was cut, and there are budgetary and strategic as well as technical reasons why these pipe dreams don’t work. But, let me guess, you can’t tell us how you’d do it because that’s classified.

      Well, Jeffrey, maybe you can tell us, since you “agree with everything” j_kies writes, and are not similarly constrained. Have you independently done the E&M calculations to determine that Postol & Lewis are wrong about the radar cross sections? I’ll admit I haven’t either, because I am pretty sure I have better things to do with my time than second-guess them on such a simple physics calculation which is squarely within their domain of expertise and about which they will be pleased to provide you complete information (which the NRC panel will not do).

      As a general point, it may be true that “the offense is stuck with the same physics as the defense.” But this depends on what you mean by “same physics”. The engineering problems are NOT the same. Neither are the strategic problems. Apart from the ease of countermeasures in midcourse (radar and IR/optical sensors cannot see through aluminized mylar balloons), the outcome is overwhelmingly biased in favor of the offense by the key underlying physical fact that nuclear energy levels are a million times greater than chemical (a factor a thousand times larger than the ratio of US to North Korean GDP), which means one nuclear weapon can ruin your whole war. And that ain’t going to change. Ever.

      Which brings me to George’s bombastic stance that we should not “consent to… a strategic MAD deterrence relationship” with North Korea. But this is not something where our consent is needed; anyone who can build nuclear weapons let alone nukes plus ICBMs (Jeffrey does not tell us what evidence he sees that North Korea has the latter, and even j_kies does not share this vision) gets to force themselves on us. Did we really “consent” to this relationship with China and Russia? Such an understanding of “consent” can get you into a lot of trouble.

    • Jeffrey (History)

      I was responding to the comment on the blog. I listed the areas of agreement — “Not kinematically prohibited, but everything depends on the radars. Yes, that’s right.” “Leadership in name only. I agree here …” — before stating that that “We might disagree about how doable it would be to keep up with countermeasures …”

      If I have a problem with something Ted or George writes, I’ll take it up with them directly and not in the comments.

    • Mark Gubrud (History)

      I forgot to add:

      The reality is that a deterrence relationship already exists between the US/South Korea and North Korea. It has pretty much since the end of the Korean War, which is why that war ended (pretty much). But the North Korean regime has not felt secure, which is why they pursued nuclear capabilities and, after GW Bush’s display of reckless aggression in Iraq, decided to go for it.

      We don’t have to tolerate North Korea’s outrages but we don’t have to turn them into another Korean (or wider) holocaust, either. Perhaps giving the DPRK a taste of its own medicine, a la recent South Korean saber rattling, is not entirely uncalled for, but I would still urge that we act like the sober adults and let the Norks be the ones playing crazy, and I would point out that it really doesn’t seem to be getting them anywhere, and at best can only delay the regime’s inevitable decay and collapse.

      Better yet, when and if little Kim shows us some leg, we should be ready to seize the opportunity to start bringing this psychotic romance to a satisfying closure. I mean, make love not war.

      Jeffrey, I really admired and enjoyed your piece in FP, which is why I can’t understand why you undermine it here with these dark hints of deeper strategic calculations in which North Korean ICBMs and effective missile defenses possess some exotic kind of ontology which escapes my grasp.

    • Jeffrey (History)

      I don’t see “dark hints” in what I’ve written.

      I think we have a reasonable problem of defense against North Korea that stems primarily from a conventional military threat and, now, a nuclear capability.

      I don’t think this remedied by our own nuclear weapons or 14 GBIs. There are probably some defense programs I would fund that you might not, but obviously this problem is best managed diplomatically — even if we buy the same horse over and over.

      I have no objection to managing the problem for the foreseeable future.

    • SQ (History)

      One important limit on the sheer-numbers approach is the U.S. willingness to spend. Or lack thereof. GDP ratios aside, the DPRK is less constrained when it comes to prioritizing defense over all else.

      Another important limit on interceptor numbers is the anticipated Russian and Chinese reactions. It’s hard to imagine they would believe, or would be prepared to assume, that high numbers weren’t about them, or relevant to them. They more or less have to assume that GMD works better than the test record alone suggests, because someday it might, right? An observation that we might extend to the KN-08, for that matter.

      On that point: DNI Clapper has said enough of late to make me sit up and take notice.

    • j_kies (History)

      Mark
      Ted and his co-authors had three major arguments in the OpEd: 1) the NRC study inappropriately states that boost phase is infeasible (they argue for faster interceptors against long burn [>8 minute] ICBMs), 2) they claimed, based on their assumed radar operations motif and assumed parameters, that the NRC radar discussion was non-physical, 3) and finally they cite the 1999 NIE as the authority of the ease of BMD countermeasures that defeat BMD sensors.

      1) While the 2004 APS study notionally considered interceptors faster than the KEI objectives (>6 km/s), the APS study certainly did not address the development, basing or logistics of such fast interceptors and the NRC experience base did not see such vehicles as being feasible. The successful 12 December 2012 Unha-3 launch demonstrates the capability to configure the first 2 stages as a ~ 4 minute burn time ICBM capable of ranging WDC. I see the BAS critique as clearly wrong on both aspects.

      2) The NRC had nationally recognized radar expertise on the panel where they applied very different operational parameters from the critique assumptions due to use of the Early Warning radars to cue their proposed X-bands to operate in a pure tracking and measurement role. After dissolution of the NRC panel, one key expert undertook to explain to Ted and his co-authors the sources of the differing expectations. I will not speak for the correspondents, but I see this as a difference in engineering approaches between the methods chosen by a published radar designer and analyst with 50+ years of work and the critics. No physics were violated so the extreme tone of the BAS OpEd was inappropriate.

      3) The US spent in excess of 30 years in research, development and flight test of ‘Penetration Aids’ involving hundreds of test events. That community of practice and specific expertise was well represented on the NRC panel so their opinions on the actual engineering difficulties of building and flying successful ‘simple BMD countermeasures’ are fully informed. I cannot similarly speak to the expertise of the analysts supporting Walpole but I cannot see that opinion if they had been involved in the historic ‘penaids’ development efforts. I see the 1999 NIE as incorrect in its assertions due to lack of engineering experience and supporting analysis of “BMD Countermeasures”.

      As to the ‘NRC travesty’, that panel was established at the behest of Congress, held data collection meetings in the open as well as in closed environments and Ted (and others) certainly had the opportunity to and did present their views to the panel. Due to the subject matter, the overall reporting had to be screened by the MDA for publication (the infamous security review). I believe that Ted and others were ‘outraged’ that their preferred views were not reflected in the report as portions were being revealed. The overall issue of private citizens regardless of their perceived stature attempting to leverage reviews of classified documents (prior to security review completion) is a really murky question. As to how others perceive the NRC report, Coyle cites it as a serious basis for recommendations and programs they recommended for cancellation are apparently being cancelled. Not bad for a travesty in my opinion.

      On the 2000 ‘Countermeasures’ monograph, the sources of my humor came from several aspects. The topic of countermeasures to a system intended to defend the US cannot be extensively discussed in the open literature without being self-defeating. Non-specialists thinking about how to defeat BMD sensing without the discipline of addressing the engineering to make / deploy the Countermeasures can easily fall into flights of fantasy. While certain ‘discrimination proof’ concepts may be allowed by the physics that you consider; the issues of putting such things on a missile, leaving that system alone until use and deploying them in a manner that confounds the defense are really hard engineering and science issues and I assure you any ‘2nd order’ effects and other issues you failed to consider will cause the item to not behave in the manner you intended. So no, the 2000 Countermeasures monograph was not credible in my view.

      Sorry if I appeared ‘coy’ or failed to make suitable arguments in your views, I worry about missiles and the means (or lack of) to deal with them. Since senior people do read blogs, I am essentially crowd-sourcing the current NK situation such that the educated discussions may suitably inform the decision making layers.

    • Mark Gubrud (History)

      J_kies:

      1) I’m not going to argue that 10 km/s interceptors are feasible, but I have not previously heard the claim that Unha-3 can carry a nuclear warhead to “WDC” even if it might be able to drop its empty nose cone somewhere in the vicinity. Throw weight, you know? Assuming a 1 ton warhead is possible, it will still need a much bigger, heavier, slower (given the same technology) rocket.

      2) What I understand from Ted is that the NRC based their assessment on a minimum RCS for the warhead provided without justification by “intelligence sources.” Their response:

      “Because cone-shaped warheads are ideal for atmospheric reentry, one must assume that any missile defense will face warheads that are either cone-shaped or combinations of cone-cylinder shapes. The radar cross sections of such warheads, which are not necessarily designed for stealth, but instead for efficient aerodynamic reentry, can easily be as small as 0.001 m^2. Even using assumptions of radar cross sections 10 or more times larger (we used 0.01 m^2), calculations immediately show that the radar ranges that could be achieved with the AN/TPY-2 and the stacked AN/TPY-2 will be much too short for the NAS recommended applications for both the EPAA and GMD systems.

      In other documents and presentations they show how they came up with their RCS estimates based on standard known formulas for these shapes plus the use of commercially available radar absorbing materials.

      3) Postol and Lewis’s complaint is not that “their preferred views were not reflected in the report” but that their explicit critiques and demonstrations of the report’s technical inaccuracies (based on simple physics) were not responded to, nor was there any independent peer-review process to determine who was correct if there was indeed an impasse or unresolvable difference of opinion.

      4) Your response re Countermeasures boils down to the usual “You’re wrong but I can’t tell you why because it’s secret,” but in this case it is even less credible because the report (again, co-authored by a bunch of other highly credible scientists including Garwin who I don’t think can be accused of lacking relevant engineering experience) states exactly what simple countermeasures it proposes. You tell us that they won’t work but you don’t tell us why.

      It’s not like I’m asking you to reveal the blueprints of secret weapons. They’ve shown you blueprints. Pray tell why they will fail. Why are you unable to do so, why has no BMD proponent been able to do so in the past 13 years?

      Jeffrey: Your first reply to j_kies above stated: “I agree with all the things you’ve said…” This would appear to have included his statement that “none of the critiques / accusations from the Bulletin of Atomic Scientists OpEd on the NAS/NRC report appear valid (and the authors should publically retract the OpEd and apologize).” Perhaps you have already taken this up with George & Ted, but I’m guessing you were just in a hurry, and appearing to agree with his rejection of everything they wrote was an oversight.

    • Jeffrey (History)

      Mark:

      I specifically noted that the commenter and I disagree about countermeasures. I was looking at the comment on the blog which only mentioned kinematics and radars. Christ almighty, it’s a fucking comments section. I’ve clarified what I meant. Accept it or don’t.

    • Mark Gubrud (History)

      Accepted.

    • George William Herbert (History)

      Mark Gubrud wrote in part:
      Which brings me to George’s bombastic stance that we should not “consent to… a strategic MAD deterrence relationship” with North Korea. But this is not something where our consent is needed; anyone who can build nuclear weapons let alone nukes plus ICBMs (Jeffrey does not tell us what evidence he sees that North Korea has the latter, and even j_kies does not share this vision) gets to force themselves on us. Did we really “consent” to this relationship with China and Russia? Such an understanding of “consent” can get you into a lot of trouble.

      The ICBMs of Russia and China predate our ability to effectively defend against such weapons. The question of consent was not relevant at the time.

      At THIS time, it is.

      Your proposition is that we must prostrate ourselves before any tinpot dictator who develops deliverable nuclear weapons and has appropriate range missiles.

      My response is what I said first – there’s no “must” in that. We have effective – inefficient, but effective – defensive technology today. We have an economic and technical advantage over the smaller proliferators. We can submit to them gaining effective ability to threaten us, or we can chose to spend the money to simply overmatch them and deny that. That choice exists, enabled by physics and technology and economics.

      My assertion and opinion is that in this case, we should say no, we do not submit or consent to NK having that ability to threaten us. We’re going to do the needful and deny them that. The reasons for that should be obvious. The costs are real, but worth it.

    • George William Herbert (History)

      Yousaf writes:
      George — NK may wish to consider a less attributable method of delivery were they truly interested in a first strike: e.g. a sailboat or fishing boat. Smaller risk of attribution, no need to miniaturize. Perhaps get Waikiki instead of LA?

      Is it sensible to create incentives for NK to do so? I’m more comfortable with them placing nukes on missiles — instead of being dissuaded from doing so. We can tell where missiles came from which might make them think twice.

      This is the terrorism vs deterrent question. No defensive system can prevent a well shielded container (or insert alternate delivery mechanism) full of atomic weapon of wreaking havoc on the US or other western nations.

      That delivery method is useful for actual nuclear warfare surprise attacks, but useless for deterrence. They cannot threaten to start delivering nukes on containers or fishing boats; our response will be to end their regime and halt any fishing boats or containers en route for rigorous checks. If they announce they already delivered some, we launch an intense search, and end their regime.

      They might conceivably be useful for short-term blackmail, but the end result of the game is regime change. Any of these actions are acts of war, and ones that it’s doubtful that we’d see serious opposition to ending their regime over.

      Missiles, by posing a stable passive threat from the regime’s home territory, can be used in geopolitical deterrence or saber-rattling without it being an offensive act.

      That is not to say that we should discount a crazed madman or terrorist organization trying to sneak nukes in, but it fails to serve their national interests to do so.

      BMD, *if* vastly technically improved, could conceivably dissuade primitive missiles w/o decoys as a delivery method — but does nothing to dissuade possession of nukes.

      Not sure that won’t make the situation substantially worse.

      I disagree for the reasons above.

      Anyway, *if* you still wanted BMD against a small nation like NK — despite the downsides mentioned above — I’d go with Garwin’s suggestion of boost-phase from surface ships.

      Surface ships are weather-dependent, vulnerable to conventional air or sub or surface ship interdiction, and about as expensive a basing mode as the silo facilities we’re talking about for more GBI. Not that I don’t believe in their utility, but we should be aware of their limitations.

      Personally, I’m happy not to incentivize them to put their nukes on boats. (And I like my tax dollars in my pocket.)

      I like my tax dollars in my pocket, too, but I also like the idea of not allowing the Kim family a veto over my continued existence. They have a distressing tendency to kill people to prove points and as a prestige game.

    • Mark Gubrud (History)

      George, my observation (please don’t consider it a proposition) is that at this time, no, we still do not have effective defense against nuclear weapons nor against ballistic missiles and certainly not (not to imply that the first two assertions are in any way lacking in certainty) against ballistic missiles carrying nuclear warheads and equipped with simple countermeasures. Nor does it appear that we will have such a defense in the future. Ever.

      But of course, some of us have been discussing the technical substance of that, so I just wanted to set the record straight about what I was and wasn’t saying.

      Of course we do not have to “prostrate ourselves before any tinpot dictator who develops deliverable nuclear weapons and has appropriate range missiles.” But neither is any such “tinpot dictator” going to have to prostrate before us in fear that we would otherwise “Saddamize” him, unless we are willing to ante up some cities for the pleasure. Short of that, with or without consent, we will have to cohabit with said tinpots until such time as their little fiefdoms go the way of the mighty Soviet juggernaut.

    • George William Herbert (History)

      (warning long)

      Jeffrey responded to me:
      What you say sounds reasonable, but let me press you a bit.

      We talk a lot about accepting mutual deterrence. The reality has always been that the DPRK has had a conventional deterrent that induces caution on our part. I teach the 1969 EC-121 shoot down in my “Security and Arms Control in Northeast Asia” class precisely because it is such a clear exposition of how the DPRK got away with the Pueblo, multiple assassination attempts on Park, the EC-121 shootdown, killings in the DMZ including the ax murder, the assasination attempt on Chun in Rangoon that killed his cabinet, the plane they blew up, etc. The DPRK always gets away with it because of their conventional ability to make war. Nuclear weapons make that problem worse, but let’s be honest that we have a mutual deterrent relationship. I don’t like it, but that’s what it is.

      Now, I agree with you that “strategic stability” is not the goal here. My objection to missile defense in the Russia context is that it is madness to think we can escape from vulnerability and that trying to do o will make things much, much worse. The DPRK is a closer call. If we can defend against the DPRK, I say do it. And, to make good on that premise I would give the NAS/NRC architecture and CONOPS a try.

      But we might not succeed. Sure maybe a 1:1 spending ratio favors us, but there is surely some ratio at which we can’t keep up. We’re richer, but we also have to plan for two road games while the DPRK gets to prepare one home game. At a billion bucks to defend against three missiles, I think we’re hosed.

      I can see the DPRK deploying 100 ICBMs. (Let’s hope they don’t figure out how to make their own TELs.) I struggle to see us deploying 500 GBIs. Moreover, the 97 percent intercept probability looks a lot less appealing against 100 missiles than it does 10. We haven’t started on countermeasures yet. At the end of the day, I am not convinced that we can spend enough to keep ahead. A thin defense might still be worth a modest investment, but lets not kid ourselves.

      My point is that dealing with the DPRK’s nuclear forces is going to be tough enough. We simply can’t be pissing away a billion dollars on 14 GBIs as a political statement.

      Keep in mind that we have about 500 Minuteman-III and 350ish Trident D5 missiles right now. The total scope of our national nuclear forces is in line with this worst case defense you posit.

      The economic ratio is 500:1. That covers a lot of inefficiency and sins. Our existing (pre-sequester) defense budget is 22 times their GDP, about $664 billion.

      Even if they try to go big on this, it eats up the resources they need for other things that threaten us.

      The reality of what they could do with conventional forces is serious. In the past that was a serious renewed conventional military threat to the south, currently at least the ability to fire thousands or tens of thousands of shells an hour into Seoul, in addition to much other provocation along the border, commando raids, etc.

      My read on the conventional military situation is that the NK manpower and mass is not useful compared to the SK armor and artillery quality, so a conventional invasion is not as credible a threat anymore. But that fight would tank the GDP of SK for a while and possibly severely damage Seoul, which is a lot of South Korea’s industry and people.

      The conventional deterrence there is asymmetrical. SK cares about its people and economy. NK has for various internal political reasons found it expedient to be provocative at times. SK and the US and others have a significant amount invested in SK’s stability and prosperity, so we are very reluctant to risk a more massive NK response. But NK also is running big risks here.

      SK probably means more to China now than NK does, from a practical political and economic point of view.

      SK’s stability is enhanced most by total peace, and it’s not too destabilized by a semiannual shelling / sinking / commando sub / whatever incident.

      However, too many of those, or sufficiently worse of an incident, inverts the SK risk perception. At some point South Korea’s stability is LEAST threatened going forwards by blowing a bunch of Pyonyang up with cruise missiles and pushing the North Koreans off the peninsulas that have range to bombard Seoul.

      The alternative assumption is that NK is and always will be “conventionally deterrable”. As we’ve seen the closer we look at big deterrence and specific incidents over time, it’s not nearly as good and stable as people assumed. A lot more of the underlying assumptions, including sufficiently aligned mindset and goals that the two sides can have a meeting of the minds on “reasonable” behavior, are not met for NK. The risk of any deterrence falling apart there is much higher. The risk of a nuclear war starting with them is higher than with China or the Russians, though they have fewer weapons.

      The idea that 97% reliable intercept is less important against 100 missiles than 10 is almost inverted. Yes, there’s always a chance of leakers, and that increases exponentially as the number of threat missiles increases. But the potential cost of NOT defending goes from (a million people) to (many tens of millions of people).

      The appropriate response is time-dependent. For one, what we can build now (is on the shelf) is more of the same silos and GBI missiles / kinetic kill vehicles. They also don’t have 100 ICBMs now, even if they could in 10 years. What we can do immediately is buy and deploy a countervailing defensive force for what they’re deploying now.

      What we should do next is make a strategic decision about handling NK going forwards. In my opinion, this boils down to one of contain – defeat – deter – defend. One of these has to be a primary strategy and has to be executed vigorously and effectively.

      Contain has failed. Defeat risks so much, primarily to South Korea and the North’s innocent population, but also the possibility of nuclear damage to SK or Japan (or China or Russia). Deter seems unstable and ineffective, though we should use it. Defend is just money.

      If we chose to emphasize defend, then we commit to it, we develop effective boost phase and put it on ships or on land in Japan or whatever. We get Russia on board and have them put boost phase missiles in Vladivostok. We develop the E-model GBI that the NRC suggested and better kill vehicles. We deploy enough of what we have to in order to keep the NK risk down. We tell NK we’re doing this and that we’re committed to matching them five or ten missiles per missile for however long they want to go building their missiles and TELs.

      There are three risks in this. One, we can’t build effective defenses. I think ours are non-optimal now. I think interceptor reliability of 95% plus – US ICBM standards – and a successful flight’s kill probability of 90% plus against the target are reasonable goals here. We’re not there yet. But we’ve demonstrated the technology is workable and build operational missiles.

      Two, we build enough defenses that we can stop China’s ICBM force, which then threatens the US / China defensive balance and relationship.

      Three, we build enough defenses that we could stop the Russian ICBM force, with the same issue there.

      Three would be a ways away. Two is not. A 10 ICBM NK force is not that far from numerical parity with China’s US deliverable deterrent force. I believe that China’s interest in seeing this situation not develop further should be clear. Expressing our resolve to China and attempting to get them to lean further on NK to avoid us having to do this would be a good thing. China reacting very badly to all this would be the most significant negative outcome I can forsee (and therefore is most important to focus attention, diplomacy, and patience on…).

    • Jeffrey (History)

      Having estimated the cost to replace each leg of the Triad, I believe those levels are largely unobtainable now for fiscal reasons. We might end up building 10 SSBNs with 160 tubes, but that’s max.

      The ICBMs are in a lot of trouble, especially if the USAF continues to insist it be mobile.

      Five hundred GBIs is not a realistic near-term outcome.

    • George William Herbert (History)

      Mark Gubrud writes:
      George, my observation (please don’t consider it a proposition) is that at this time, no, we still do not have effective defense against nuclear weapons nor against ballistic missiles and certainly not (not to imply that the first two assertions are in any way lacking in certainty) against ballistic missiles carrying nuclear warheads and equipped with simple countermeasures. Nor does it appear that we will have such a defense in the future. Ever.

      But of course, some of us have been discussing the technical substance of that, so I just wanted to set the record straight about what I was and wasn’t saying.

      I’ve been staying out of that sub-discussion, but I am somewhat technically aware of the BMD technology assessments and have been following it for some time.

      I recall a decade’s worth of deeply heartfelt assurances that hit-to-kill was obviously unreasonable and could never be made to work reliably. I also recall that once the basic technology was fully developed, the next 3 test missiles that didn’t suffer a technical failure rendering it a no-test all hit the target dead on.

      I recall a lot of conversations about decoys, credible decoys, various decoy types, RAM, balloons, chaff, flares, MRV designs, etc. I also recall a lot of “It’s not that easy to develop” and watching some test vehicles punch right through the decoys and hit the right target.

      This is not to say that the difficulty of HTK or the real threat of decoys and noisemakers are unreal. Failure to respect the issue is unwarranted.

      But these are technology questions, and there are a lot of assertions by skeptics that turn out to be solveable problems. I can’t count how many “Missiles will render manned aircraft obsolete”, “missiles will never be jammable”, “missiles will never be able to burn through the latest jamming and decoy suites” types of circles have gone around and around in the generally similar aircraft ECM/decoys/stealth field.

      I do not agree with your observation. I don’t think it’s invalid questions to be asking, but I don’t agree with the conclusion.

      Of course we do not have to “prostrate ourselves before any tinpot dictator who develops deliverable nuclear weapons and has appropriate range missiles.” But neither is any such “tinpot dictator” going to have to prostrate before us in fear that we would otherwise “Saddamize” him, unless we are willing to ante up some cities for the pleasure. Short of that, with or without consent, we will have to cohabit with said tinpots until such time as their little fiefdoms go the way of the mighty Soviet juggernaut.

      Which is one tradeoff to make when the question is about Iraq and apparent but ultimately fictional WMDs or North Korea and a Cheonan. But it’s a very different tradeoff when it’s a Cambodia and a million skulls stacked in soccer stadiums, Rwanda, another Hitler and tens of millions more dead.

      Or, more likely, a thousand artillery shells into Seoul, with the threat of another hundred thousand if we don’t capitulate on some major point, and an inability to then promptly end their regime without the threat that Seattle or LA or Chicago or DC or New York or all the above would be damaged severely with a million casualties each.

      Some issues demand intervention. Being effectively deterred under those circumstances is unacceptable in the neighborhood of China / Japan / South Korea.

      If you think that the NKs are effectively deterred from larger atrocities now, I urge you to re-read the last couple of decades of studies on deterrence failures, and the last 50 years of North Korean adventurism. The last ten things which demanded intervention weren’t there, but there hasn’t been a decade that the highest probability hotspots weren’t there and the Middle East / Israel region.

      The threat they would continue to hold against Seoul with all the conventional artillery will still deter the US and South Korea from rash aggression. There’s no need to let them hold multiple US cities at nuclear ransom to keep any reasonable balance of anything.

    • George William Herbert (History)

      Jeffrey wrote:
      Having estimated the cost to replace each leg of the Triad, I believe those levels are largely unobtainable now for fiscal reasons. We might end up building 10 SSBNs with 160 tubes, but that’s max.

      I think it’s more likely to be more stretched Virginias than the big-hull proposal, and perhaps even fewer tubes, but…

      The ICBMs are in a lot of trouble, especially if the USAF continues to insist it be mobile.

      Not ultimately their call. Presidents and SecDefs need to remember that.

      Five hundred GBIs is not a realistic near-term outcome.

      Neither is 100 NK KN-08 missiles. If they committed to it as their national focus goal it would take a huge part of their resources and the better part of a decade to get there. If we assume 1/month as a realistic production rate they could meet, we match that with 5/month GBIs (worst case, assuming no tech/systems improvements), $4 billion a year.

      If SecDef and the President can’t find $4 billion in a $665 billion DOD budget and $3.8 trillion total federal budget next year, we have a national problem.

      Again – we have the money. You’re suggesting and asserting we do not have the willpower. I believe you when you say we don’t for a full-force replacement of the existing ICBM/SLBM forces – not only is that more expensive than this, but it’s strategically unwise, if we want to build down towards under a thousand weapons. There’s no reason to muster the willpower to rebuild at current force levels. But for defenses against NK we can find the money and willpower.

      The real risk here is that NK’s behavior ends up making it harder to get Russia to build down their strategic forces, if we increase our defenses enough, and what China’s reaction would be. Again, I believe that getting China on board from the beginning would be wise.

    • Mark Gubrud (History)

      George,

      Some assertions made by BMD opponents may certainly be wrong. I cannot apologize for people who are on the right side of the issue but don’t know the physics or get their facts wrong.

      That HTK works is surprising from a naïve point of view until you realize that the delta-V impulses required for course correction in the end game are of a constant order of magnitude as you approach the target. So, it turns out to be “technically sweet.” But only if the target is non-evasive.

      Target discrimination in the presence of countermeasures is a much different matter.

      Tests in which you claim “test vehicles punch right through the decoys and hit the right target” have all been scripted frauds. The “decoys” looked completely different from the “warheads,” and the interceptor knew what each would look like in advance. In one case, analysis after the test showed that the “decoy”, a large balloon much brighter than a warhead, had actually served as a beacon for initial acquisition of the “threat cloud.”

      All this has been amply reported; I’m quoting from memory but you can easily look it up yourself.

      The number of tests in which successful discrimination of serious decoys from seriously-disguised warheads has been demonstrated is zero. In fact, the number of such tests which have been attempted is zero.

    • George William Herbert (History)

      Mark Gubrud writes in part:
      Tests in which you claim “test vehicles punch right through the decoys and hit the right target” have all been scripted frauds. The “decoys” looked completely different from the “warheads,” and the interceptor knew what each would look like in advance. In one case, analysis after the test showed that the “decoy”, a large balloon much brighter than a warhead, had actually served as a beacon for initial acquisition of the “threat cloud.”

      All this has been amply reported; I’m quoting from memory but you can easily look it up yourself.

      The number of tests in which successful discrimination of serious decoys from seriously-disguised warheads has been demonstrated is zero. In fact, the number of such tests which have been attempted is zero

      Here’s where I and the critics drift apart some.

      I could easily design some exceptionally credible decoys. I know enough about the bus systems, sensors and defeat measures, likely counter-countermeasures, etc. I even know enough about launch vehicle upper stages and separation and the like, for the bus integration issues.

      I have very low confidence of high reliability of the above without test flights.

      AT THE VERY LEAST – seriously good decoy systems are as hard as a multi-reentry-vehicle bus (MRV), and arguably they’re as complex as a MIRV bus (adding independent targeting and complex bus maneuvering along with maintaining targeting accuracy during the maneuvers).

      Even getting a single satellite to orbit successfully and deployed out from under a payload shroud is a maneuver that fails quite a bit often, particularly in new spacefaring countries. Shrouds do the damnest things. Collisions between stages have doomed quite a few launchers, and collisions with shrouds or with satellites after separation are not at all rare.

      When you’re talking about a serious massive decoy set you’re talking about an engineering effort and testing effort that resembles that of the total launch vehicle under it. It took quite some time for the US and USSR to get it right, with far more resources and test flights than NK has had a chance to try so far. Dozens.

      If you want to suggest that some day, a single warhead missile coming at the US will deploy both a bunch of environmental fuzzing (chaff+flares+dumb balloons) AND a bunch of really good cone decoys that are engineered (and mechanismed) to match the real warhead thermal as well as radar signatures across several bands, I would agree. The USSR can do it to us today, and probably China.

      If you want to suggest that NK is going to do that to us without a long and very visible test program, I suggest that you’re engaged in decoy threat inflation. Decoy sets that advanced will not just happen. They will require test programs, which we’ll see. If they tried to use them without testing they’re more likely to crash their own buses and warheads than successfully mislead our interceptors.

      Those tests have to be in vacuum and zero-G. Even a large ground-based vacuum drop tower would not be sufficient for large parts of the testing. They could conceivably put a vacuum chamber on a passenger or cargo jet and do vomit-comet type flights but the quality of microgravity isn’t good enough for most of the tests I have in mind. I strongly suspect they require exoatmospheric flight tests. Which could be on vertical sounding rockets never leaving the NK territorial footprint but would of necessity be going up high where China, SK, and Japan (plus any ships nearby) can get a really good look.

      In short, the tests we did perform match somewhere between what I judge to be the middle and extreme outside edge of what a NK threat could do with decoys without flight testing. And therefore are perfectly applicable and valid, and not “faked” as has been repeatedly claimed by skeptics.

      Do I fear that NK could sometime get the better super-decoys? Yes. Do I think that they could do so in the next ten years? … Maybe. Fifteen or more is more likely. They’re still working on the rocketry end of things. Without us noticing? No. In my judgement we’d know it’s coming.

      I do not think that we should dissuade ourselves from protecting ourselves for 2014-2025 by worrying about Brilliant Comrade’s Miracle Decoys. Staying wary for signs of them starting a development and test program, yes, but do not assume the program or a successful conclusion on timescales shorter than their missile program has taken to date.

    • FOARP (History)

      “The USSR can do it to us today”

      One of the surprising things to a layman is how many of the wonks on this site still speak as if the USSR was a going concern. Yeah, I get that he meant that they “could do it”, but all the same . . .

    • j_kies (History)

      Mark –
      “…the claim that Unha-3 can carry a nuclear warhead to “WDC” even if it might be able to drop its empty nose cone somewhere in the vicinity. Throw weight, you know?” No claims, no need to believe in intelligence assessments; pure kinematic analysis does the job on public information.

      Consider the mass placed on orbit on 12/12/2012, a satellite, some deployment debris and the empty 3rd stage rocket body (choose numbers that are internally consistent). The path that vehicle took to orbit includes a large ‘plane change’ maneuver so the effective Vbo of the stack was rather higher than a direct ascent / insertion. When you have constrained your estimates of the vehicle against the stage and fairing drop points and the final orbit, consider what the first 2 stages did to drive that 3rd stage to where it could make the orbit shift. This type of constrained estimation shows those 2 stages would generate ICBM launcher Vbo carrying sufficient mass for a warhead.

      George and Jeffrey –
      While some might argue against such an ICBM, I note the Soviets had the R7a on ‘alert’ with a preparation time in the ½ day range. If people are really driving towards views of a KN-08 as an ICBM, I look to the history of the R-16. http://en.wikipedia.org/wiki/Nedelin_catastrophe

    • John Schilling (History)

      To expand on George Herbert’s latest, it is important to note that every serious look at BMD allows for the possibility of credible decoys; things that look enough like warheads that even after you’ve stared at them with X-band radar and space-based LWIR sensors, you just plain have to commit interceptors to shoot them down because they might really be warheads. And, for all its faults, America’s BMD system is at least somewhat serious in this regard.

      But, as George points out, credible decoys are Really Hard. Testing, as he points out, is a serious challenge that most emerging nuclear powers are a long way from overcoming. Most of them are a long way from even building anything worth testing. And when you do find you have built and tested a credible decoy, it will turn out to have grown quite a bit from the naive initial concept of e.g. an overgrown toy mylar balloon, into a rather heavy and expensive piece of equipment. And no, I’m not going to explain all the reasons for that here.

      But the end result is that while the Russians might be able to put a few dozen credible decoys onto every SS-25, that only comes as a result of decades of work in the field, and the fairly substantial throw-weight of an SS-25. The North Koreans are not going to be putting dozens of credible decoys on a KN-08. Right now, the best decoys they can plausbly decoy, will be decoys that we can sort out from live warheads with some effort. When, maybe five or ten years from now, the North Koreans can deploy credible decoys, the marginal throw-weight and unshrouded warhead of the KN-08 will limit them to very small numbers.

      Which means a proper defense against the North Korean threat will involve things like testing GBI and SM3 kill vehicles against crappy “unrealistic” decoys, carefully observing North Korean test activities, and deploying interceptors in numbers sufficient to counter all known Nork long-range missiles even if each one carries a few credible decoys. Which is to say, about what the MDA is actually doing.

    • Mark Gubrud (History)

      George and John:

      What both of you are ignoring:

      1. US pen-aids in the 1960s had to deal with nuclear detonations in their vicinity, and were intended to work into the first seconds of reentry, since interception was mostly going to be terminal-phase. The discrimination radars also had an easier time (until blackout set in) because of relatively short range as compared with current scenarios.

      2. There have been no successful tests involving even reasonably serious, e.g. cone-shaped, decoys. Bill Broad had a nice review of this in the NYT some years ago, and there is nothing new since. In fact, the 2012 NRC panel reported that MDA seems to have given up working on discrimination some years ago. Much easier just to pretend the problem doesn’t exist, I guess, when you’ve got both political parties on board and not asking serious questions.

      3. Antisimulation. Instead of trying to make a perfect decoy, you make a bunch of random ones, and also randomize the warhead signature, possibly by hiding it inside a “decoy.”

      4. Neither IR/optical nor radar sensors (which are the only ones used in our current BMD architecture because they are the only ones that are physically practical to use from long range) can see through very thin and lightweight aluminized mylar. This material can be inflated as a balloon with a bit of gas, or held on some kind of unfolding framework. One can easily come up with many creative ways you could potentially use to create a shell game.

      5. Another nice material to work with is Eccosorb, which will reduce your RCS to a shadow, and is also opaque to light.

      6. Thermal signatures can be randomized with heat sources, cryogens, and paint patterns.

      8. Unlike in MDA tests, in an engagement with North Korea we can’t depend on the Norks to provide us with signature data to make the discrimination problem easy.

      Finally, I am, again, not accepting claims that you know reasons why these things wouldn’t work but you won’t tell us. The discussion has been ongoing for many years. No BMD proponent, not even an irresponsible one with a strong engineering and physics background but no clearance to lose or risk imprisonment, has been able to explain why these simple countermeasures won’t work or would be too hard for a state that can produce ICBMs.

    • John Schilling (History)

      Mark: You are “not accepting claims that you know reasons why these things wouldn’t work but you won’t tell us”. That’s fine, because we aren’t making any such claim. If you actually pay attention to what George and I are saying, instead of what you expect us to say, we are claiming that it is VERY HARD to make these things work. That it will consequently take much longer to accomplish, and lead to much smaller results, than you or anyone else in the “thin aluminized mylar balloons” crowd are likely to understand. And we are making that claim on the basis of decades of experience in making things that do work in this environment. It’s really damn hard. Rocket-science hard.

      But, yes, these things will work. On a limited scale, after enormous effort. Please do not assert that we have claimed otherwise, because that would make you a liar.

      And if you are going to demand a detailed explanation of the limitations and of the effort required, I at least will not oblige you. First, because it would require far more effort than a series of blog posts would warrant. Usually, I get paid for that level of effort. Sometimes I will do it pro bono, but not today. Second, because it would in large part amount to an enumeration of the pitfalls and obstacles anyone trying to turn “thin aluminized mylar balloons” into reality will encounter – at least some of which the North Koreans haven’t figured out yet, and maybe one or two of which will lead them far astray. All this against, what, the possibility of enlightening Mark Gubrud if his mind happens to be open that day?

      Your views on this subject are quite clear. As is our disagreement, and at least the broad basis for our disagreement. If that’s not enough for you, so be it. I fully expect that we will continue to disagree on this.

    • Mark Gubrud (History)

      John, my mind is open any day of the week. It’s been 13 years since publication of the Countermeasures report, which really only reiterated what had been discussed publicly for many years. If there is a single technical paper out there, which I have missed, rebutting the report and showing why the proposed countermeasures would fail against the combination of radar and optical/IR sensors, or would be too difficult for a threshold state that can build an ICBM, and that consists of something more than these claims that “we’re rocket scientists and we know how hard this is, so trust us and believe what we say,” please let me know and I will rush to retrieve whatever reference you provide.

    • George William Herbert (History)

      Mark Gubrud writes:
      John, my mind is open any day of the week. It’s been 13 years since publication of the Countermeasures report, which really only reiterated what had been discussed publicly for many years. If there is a single technical paper out there, which I have missed, rebutting the report and showing why the proposed countermeasures would fail against the combination of radar and optical/IR sensors, or would be too difficult for a threshold state that can build an ICBM, and that consists of something more than these claims that “we’re rocket scientists and we know how hard this is, so trust us and believe what we say,” please let me know and I will rush to retrieve whatever reference you provide.

      You’re still making the mistake John just pointed out.

      The physics of capable decoys is well known. They’ve been flown. Multiple nations around the world have them, the best talked about (but by no means newest or most capable) the UK Chevaline program designed to punch their SLBMs through the nuclear interceptors around Moscow, so they could credibly independently hold the former Soviet now Russian capitol at risk.

      Existence proofs abound.

      What we’re trying to communicate is the scope of the development effort and engineering challenges between “we know it can be done” and “we have a reliable deployed system” in this case.

      We know it can be done, which is why I don’t personally expect that we could credibly build a NMD system to defend against an all-out Russian strike (and don’t count on it in the nearterm against China).

      John and I are asserting – and please believe us, we both ARE quite literally rocket scientists – that it’s a really large difficult engineering development program to be able to actually fly and disperse those types of decoys. Which is arguably at as complex as the ICBM development or the nuclear weapon development programs are. Which will require multiple highly visible test flights to validate before it could be considered credible.

      The US tests with the decoys we did fly are more competent decoys by a large margin than what NK could credibly stuff in a KN-08, between volume and mass constraints and a constraint of not having flown decoy dispersal tests.

      That is not “they can’t do that”, that’s “it is a well founded statement that it’s exceptionally unlikely that they can do that now, or in the near future, though they could credibly start developing that capability at any time”.

    • George William Herbert (History)

      FOARP wrote:
      “The USSR can do it to us today” One of the surprising things to a layman is how many of the wonks on this site still speak as if the USSR was a going concern. Yeah, I get that he meant that they “could do it”, but all the same . . .

      I would like to say that was some sort of clever play on words / names, but it was a thinko as I typed.

      I am fully aware that the USSR disintegrated a couple of decades ago; I watched the Berlin Wall fall on TV live while I was in college, and the shelling in Moscow in the abortive coup attempt against Gorbachev a few years later, in which Yeltsin rose to prominence.

      Russia continued a number of the ex-Soviet ICBM progams and associated warhead, bus, and penetration aid programs.

  3. Greg Thielmann (History)

    Thanks for addressing technical flaws in the Washington Post editorial. I submitted an LTE that pointed out the editors had misquoted DNI Clapper on Iranian ICBM timelines. I didn’t have space to criticize the editorial’s Cold War logic, implying that anything that might please Russia must be a loss for the United States. I think an entire supplement is necessary to elaborate on the many failings of this editorial.

  4. Rene (History)

    A question: are there any North Korean plans for building SSBNs? If there are, how long would it take them? And if they get there, how effective the current defenses would be against it?

    Sorry I’m rather ignorant military-wise.

    • Cameron (History)

      Not that I’ve ever seen, but IANANorK. As far as a desire, I can’t see it as a priority. The cost of development would be prohibative, and the safety of the weapons, given US naval technology would likely decrease vs digging bigger/deeper holes for missles.

      Not to mention the “trusting the people on the boat” issue.

  5. Anon2 (History)

    I am sorry, but it appears to me that allowing the DPRK to achieve nuclear weapons capability is an unmitigated disaster.

    As Yousaf points out, there are always multiple “leaky” means of delivery, i.e. planes, ships, midget submarines, and so on, not to mention an untested missile defense.

    We have on the other side the Junche Republic building what, 5 weapons per year, maybe more. They have promised to take out the United States.

    I am afraid this ends badly and yes, I live in a target zone. This is the best my government can do for me?

    What we are left with is counting on the NK Junta to behave rationally like the Soviet Union and China did before. Where does the DPRK Junche propaganda end? Who outside of the NK Junta knows?

  6. SQ (History)

    If memory serves aright, a shoot-look-shoot concept involving pairs of interceptors might have been entertained once, long ago.

  7. Egad (History)

    The current doctrine is to salvo-fire five interceptors at each target. That is five shots, not two.

    Just a comment on that. Back in the 1990s when MDA was BMDO and GMD was NMD, I had a chance to become somewhat familiar with official thinking on such matters. (At the time the firing doctrine was four-on-one, so the confidence in hitting a target with a GBI doesn’t seem to have increased since then.)

    But, more importantly, “target” was “credible target”, meaning something that we couldn’t be sure wasn’t a warhead. So, in the case where shoot-look-shoot was feasible, four interceptors would be launched at each non-discriminable object deployed from an ICBM. If the result of the first salvo left some credible objects/targets, four more interceptors would be fired at each of those. In cases where the time needed for initial discrimination didn’t permit such shoot-look-shoot, the idea was to launch enough interceptors to do four-on-one engagements of each “anticipated credible target” and hope that they would be enough. Variants of each case existed, but they all involved pretty high ABM/ICBM ratios — even with perfect confidence in our ability to discriminate warheads from not-warheads, the ratio was still 4 to 1. And, at least at the time, confidence in ability to discriminate was considerably less than perfect.

    A bit of reflection indicates that, under such a firing doctrine and unless the ability of the defense to discriminate actual warheads from decoys is very good, the required numbers of ABM interceptors vs ICBMs goes up rapidly. E.g., if each ICBM deploys three credible targets then, under five-on-one, we’d need at least 15 GBIs per DPRK ICBM.

    • John Schilling (History)

      The whole point of shoot-look-shoot, if and when it is technically feasible, is that you don’t need to fire five GBIs up front against each credible target. Firing two GBIs at each target, looking at the results, and firing three more if the first two missed, gives the same total Pk as salvo-firing five interceptors, but requires roughly 2.75 interceptors per target instead of five. A little over eight GBIs for each of your postulated Nork dual-decoy ICBMs. Actually, you implicitly assume Nork ICBMs have 100% operational availability and reliability; as that is not the case, half a dozen GBIs per deployed KN-08 would probably be about right.

      That’s half a dozen GBIs plus the radars and other C3I infrastructure needed to support shoot-look-shoot. Which is a non-trivial but largely fixed cost; against a small threat, it may be more cost-effective to just add enough GBIs to support five-shot salvoes against everything. At some point, you get a better return by building the radar sites that effectively double the capability of whatever GBIs you already have.

      The whole thing works better still if you have something better than GBIs to do the shooting with, of course.

  8. yousaf (History)

    S-L-S may be sensible if there exist comparable numbers of opportunities for shooting as there are presumed decoys.

    Presumed decoys usually number in the dozens. So am unsure if SLS makes substantial difference when CMs/Decoys are involved. For a single warhead with no chuff/no natural CMs/no man-made CM/no decoys it may work.

    @George — I think we can agree to disagree: if a given adversarial nation has nuclear weapons, I do not wish to incentivize them to place them on poorly attributable boats, which may tempt them to actually use them.

    The sad reality is this: if an adversarial nation has nuclear weapons you will be deterred, no matter how good you think your defense may be. And the adversary likely knows this.

    @j_kies: the Pentagon’s own DSB and NAS both agree with the inherent difficulty of the decoy/CM problem.

    Ted and George Lewis’ difference with the NAS study has to do mostly with that study dismissing boost phase defense by setting up a strawman boost phase defense that would not do the job. Their strawman boost-phase example does not do the job because it is set-up (burnout speeds etc.) to not do the job. I agree with Ted and George.

    And this is what Dick Garwin thinks:

    http://www.ucsusa.org/assets/documents/nwgs/RLG-comments-sept-2012-NAS-BMD-rpt.pdf

    • John Schilling (History)

      Why would shoot-look-shoot require a number of opportunities comparable to the number of decoys? As the term is usually understood, S-L-S means firing a salvo of 1-2 interceptors at each credible target simultaneously, observing the results, and firing a second salvo of interceptors at each remaining credible target. Doesn’t matter whether the target set consists of a single unprotected warhead, or a hundred warheads hiding among ten thousand decoys – you need only a single opportunity for a reshoot in either case.

      You need more interceptors in the second case, yes. But any reasonable implementation of shoot-look-shoot will roughly halve the number of interceptors required to achieve the same level of protection.

      Also, I am skeptical of the presumption that decoys will number in the dozens. I don’t doubt that the presumption is common, at least among BMD skeptics. But it is not realistic except in the case of modern US, Russian, and maybe Chinese ICBMs. We are talking here about near-term North Korean and Iranian missiles; by my math the Norks can’t squeeze more than half a dozen credible decoys into a KN-08 under even the most optimistic assumptions, and the Iranians aren’t likely to do better.

    • j_kies (History)

      Yousaf – (and Mark)
      Regarding “Countermeasures” and the 1999 NIE, these are engineering problems not science problems. Working with the practical problems of making Penaids AKA BMD countermeasures has always been classified so the assessment of the 2000 report (including assessments of what does and what does not work as expected) isn’t up for discussion.

      I would assert that while Dr Garwin is expert in the engineering behind the MIKE device that does not make him expert in the issues of engineering BMD Countermeasures.
      David Montague Co-chairing the NAS/NRC panel provides vastly more useful and profound insight from a guy that has BUILT and TESTED SLBM’s and SLBM PENAIDS including operational ones in the following quote from the press conference:

      MONTAGUE: Well, first off, nobody has built and provided countermeasures to deal with — effectively with both optical systems and radar. But — and I say effectively.
      However, we believe the major powers, such as Russia and China, have a lot of work going on in that area. And I believe it’s fair to assume that that technology would quickly be available to even third world countries.
      So I think it’s a — the contest that I described earlier would have to go on and continue as you face the likelihood of a certain kind of countermeasure.
      And by the way, you know, cheap, light balloons are not effective.
      SLOCOMBE: Yes, I think — I think that’s an important point.
      I’m not — I’m not quite sure which intelligence community report you’re referring to. But yes, any country that can develop an ICBM will probably think about a countermeasure.
      But for reasons that are explained in the report, it — it — it’s not an easy thing to do.
      My favorite example is the British spend a — gobs of money trying to develop a system called Chevaline, which eventually they deployed for a brief period of time.
      And they had a lot of — they had a lot of resources and a lot of desire. And it didn’t work very well.
      And they came to the conclusion that it was not a productive way to maintain the viability of their deterrent.
      The report discussion indicates some of the problems that the countermeasure people have. But it’s in — and I’m not saying that the discrimination problem is easy to solve.
      I’m only saying that the fact that you can — you can say well I — I can imagine a — a technically feasible countermeasure, doesn’t mean you’ve — you’ve got it and it’ll work right.

      CHEVALINE as part of the POLARIS A3 family held 42 flight tests to sort out all the major elements of the propulsion and payload deployment. CHEVALINE was explicitly a first world countermeasures development with fully informed science and engineering efforts on par or better than anything presently available to MDA and their partner organizations. After the UK declassified much of their historic nuclear deterrent; they authored an informed engineering history of the CHEVALINE program – much can be found in the exceptional Wiki article http://en.wikipedia.org/wiki/Chevaline

  9. Jeffrey Lewis (History)

    Chevaline is a fascinating case, although some challenges in penetrating a nuclear-armed ABM are different from penetrating a hit-to-kill system. My sense — and the British are never terribly direct about this — is that cost-growth was driven by rad hardening the warheads.

    Having said that, I think this is a close enough call that we aren’t going to be able to reach analytical conclusions on a a priori basis. We probably have to spend the money to see how the defense-countermeasures race shakes out.

    • j_kies (History)

      Jeffrey –

      Thanks for unintentionally hosting a great crowd-source, I can only hope the philosophic opponents to our current policies take the time to review the current events threat ‘facts on the ground’ and compose more useful and actionable objections than arguing about ‘discrimination’. I am completely onboard with Coyle’s article in ACA; when does the US plan to get properly serious and bring the personnel expertise necessary to do the engineering in a robust and reliable manner?

      BMD is a difficult problem but doable with sufficient scientific and engineering diligence, the offense is stuck with the same physics as the defense. The real question – does the nation chose to make that investment rather than accept the consequences of mistakes or failed deterrence.

    • Jeffrey (History)

      What do you mean by “unintentionally”?

    • j_kies (History)

      Laughter, touche’ sir, I merely meant the SLS topic spawned one of the better discussions I have seen regardless of forum and I deliberately exploited the opportunity to see what the smart public thinks

      I should recall, your an academic with a blog, please do continue with your intentional means of stirring discussions.

    • Mark Gubrud (History)

      It’s a fallacy to assume that the burden is on the offense to come up with countermeasures that are certain to work, and that this would impose the requirement for an extensive testing program (which, incidentally, would not necessarily provide the defense with enough data to ensure discrimination). Rather, it is the defense which has to have high assurance of effectiveness against unknown or only partially known countermeasures (which can be randomized in unknowable ways). Otherwise, the Norks will have us deterred.

      Rather than commit to spending limitless sums on a measure-countermeasure race that is not known to have any ultimate limits, we could maybe subcontract the Norks to provide the targets we have to discriminate in our tests. Make them the red team. Maybe even propose this as a new Olympics event. Think I’m kidding? Seems a lot less wild to me than some of the other ideas I see thrown around.

      I think what this “crowd-source” shows is that at the end of the discussion those claiming discrimination will work and countermeasures will fail are reduced to thumping claims of superior knowledge which they can’t or won’t share. It’s been the same con all along.

    • George William Herbert (History)

      Mark Gubrud writes:
      I think what this “crowd-source” shows is that at the end of the discussion those claiming discrimination will work and countermeasures will fail are reduced to thumping claims of superior knowledge which they can’t or won’t share. It’s been the same con all along.

      This in no way resembles the actual arguments we’re making. Please listen to us.

      There is no hermetic knowledge. It’s all laid out for you at the high level.

      The idea of advanced decoys is easy. Existence proofs of exist. Credibility of the threat is established and acknowledged.

      The difficulty – time, technical talent, development risk and test and validation challenges – is the entire point. It is a difficult enough task to go from theoretical knowledge that it is possible to having reduced it to reliable repeatable practice that we can safely say NK are not there yet, and that it is extremely likely we would see their decoys and dispenser bus testing many many years before they would have completed development and have a credible flight-ready system.

      They could have started development already, or could this week, but until they start flying tests we have very strongly credible reasons to believe that they are on the order of 10 years from a viable system.

    • j_kies (History)

      Mark-
      Brilliant idea on getting the Norks to build our targets; one of my SETA’s did one of his master’s papers at GWU on exactly that thought.

      I spent significant time under the scrutiny of the GAO and the JASONS so the ‘con’ description doesn’t play unless you want to toss rocks at those groups as well as the NAS/NRC. We choose not to have discussions in public in order to force prospective opponents to replicate some fraction of our development activities. 30+ years of Penaids development makes for an astounding amount of data/analysis products as comprehensive as you might wish in the BMD countermeasures domain (which the NAS/NRC points out the MDA is far less informed on than would be expected). If you happen to have clearances, the history of what was done and corresponding problems / failures to deploy in flight are accessible. I believe MDA is looking to hire competent scientists and engineers so you could consider a sabbatical to contribute to the security of the country.

      Where BMD has failed is the reliability and relevant engineering of the interceptors and support systems. What can and should be done (if anything) to rectify those basis engineering issues is a significant question of national importance.

    • kme (History)

      It would seem to make a lot of sense to commission a local Red Team, funded and staffed to the presumed level of the North Korean decoy program, and see how well they can perform against the MDA’s finest.

  10. Mark Gubrud (History)

    I called it a “con” not to be rude but to point out that you keep deploying claims of authority, experience, expertise, and other rhetoric intended to induce people to believe that you have this inside knowledge which you can’t or won’t share, instead of responding to clear, substantive technical arguments which incidentally have been advanced by people other than myself whose credentials are at least as good as yours. That is the MO of a confidence artist. It seems to be where this discussion always ends up.

    • George William Herbert (History)

      Mark – We’ve addressed the clear substantiative technical arguments. They’re right. Good decoys work. We’re not and have not been denying that. There is no coverup or hidden knowledge.

      Our point is simply that NK is for the sake of argument in the realm of $5-15 billion dollars, 5-15 years, and 10-100 test flights away from being able to credibly deploy and use more advanced decoy systems.

      This is not a physics problem, it’s an industrial R&D program depth problem.

    • Mark Gubrud (History)

      They don’t have a nuke warhead ICBM yet either.

      Countermeasures that can defeat present US BMD, at least to the level that we won’t know if it will work (if it works even without the CMs) are easier to do than the warhead and the rocket.

      This has all been laid out, and no specific reason has been given why what has been laid out won’t work.

    • George William Herbert (History)

      Mark:
      They don’t have a nuke warhead ICBM yet either.

      They have not demonstrated a nuke warhead ICBM; they have demonstrated a launch vehicle that is a ICBM clone, needs a big silo or pad, and finally more or less succeeded. They do not appear to be deploying Unha-2 ICBMs in quantity. They have rolled a bunch of possible road-mobile ICBMs on TELs without test flight. They’ve fired a number of nuclear test shots.

      It’s fair to say there are gaps in demonstrated capability, but the credible case includes the possibility that the current apparent KN-08 deployments are real missiles with nuclear warhead.

      Countermeasures that can defeat present US BMD, at least to the level that we won’t know if it will work (if it works even without the CMs) are easier to do than the warhead and the rocket.

      It took longer time and more money for the west to do good countermeasures than the first gen ICBMs took, and about what the first MIRV buses took.

      Those are more complex than the rocket or warhead.

      It’s trivial to put chaff and IR flares and cheap balloons on everything if you don’t care about positioning and relative velocities in the cloud. That’s about what we tested against. Going beyond that is the big development programs.

      This has all been laid out, and no specific reason has been given why what has been laid out won’t work.

      Time and money.

    • John Schilling (History)

      Mark,

      Accusing people here of being con artists is in my opinion well over the line.

      There are many legitimate reasons why people who have the answers you are asking for, will not give them to you. First off, neither you nor anyone else in the public domain is paying for those answers, yet you are asking for a level of effort well beyond what anyone can reasonably be demanded to generate pro bono. Second, a public discussion of missile-defense countermeasures at the level of detail you are implicitly asking, even including no officially classified information, could cause non-trivial damage to national security.

      Third, you are not even being polite about it.

      Based on my substantial professional expertise, credible decoys capable of effectively defeating even the present US national missile defense system, would be a multi-billion dollar, 10+ year development program that North Korea has not yet seriously embarked on and would have difficulty integrating into the missiles they are presently deploying. That’s about all the answer you are going to get here, for good reason. If you aren’t willing to trust my judgement, and that of my colleagues, without the underlying details, that is your right. If you are going to accuse us of wrongdoing for not meeting your demands, that is not your right.

    • Mark Gubrud (History)

      It’s not that I think you guys think you are simply trying to fool everyone with something you know is not true. It’s that everything you are saying on this (CMs) still comes down to “This is our judgment and we know what we are talking about.” Each round is another attempt to affect a more authoritative, or more offended, pose. It seems to be a script that all you guys follow, and its obvious intent is to impress and intimidate without informing and without answering fact and logic. You throw in some jargon, you hark back to your days working on this or that, you knock off a few hits on big, controversial issues as if you and everybody else in the know just plain knows, and those who don’t are dilettantes and fools. In fact, I expect that’s actually how you see it. Seeing it that way is probably part of the admission ticket.

      In response to Jeffrey’s suggestion that we need to let the measure-countermeasure race run its course, since we otherwise seem to have an epistemological impasse here. But it would be far, far less expensive to resolve this by funding unclassified academic research into the matter. No, no, not another blue-ribbon panel stacked with “insiders.” Open, peer-reviewed research and an orderly process of inquiry and debate. I’m sure we could settle this once and for good, for far, far less than a billion bucks.

      But for that matter, we’ve spent more than a hundred billion and that’s bought a lot of loyalty from those who’ve been a part of this enterprise, or boondoggle. I’m sure you guys believe what you’re telling us. But I don’t. You haven’t given me any reason to, and I’ve got plenty of reasons to think otherwise, which you have not engaged or answered.

    • George William Herbert (History)

      Mark, you are still not addressing what we’re saying.

      Do you not believe that Chevaline – an equivalent program to what you’re proposing to first approximation – cost $2 billion in 1980 dollars, about $5 billion today, for a nation which already had most of the component technologies and was working with another that already had decoys and MIRV buses. With project inception / low rate R&D in 1970 until full scale development in 1975, first field units in 1982, so 12 or 7 years depending on how you count project start. With a lot of test flights.

      Do you believe that starved 24 million person $32 billion GDP NK will be able to do the job faster than the 56 million person, $1.3 trillion GDP 1980 UK was?

      As far as I am aware, I have not seen any of the skeptics advance an analysis of the development program scope and demographic and economic impact and environment in which NK would have to operate to build these decoy systems. Please forward references if you are aware of any.

      This is not assertion of special knowledge. The publicly disclosed costs, time, and difficulties of Chevaline are a perfect example of the scope of a first decent decoy program. The UK declassified the bulk of the data here in the mid-1990s for the program costs, time, basic issues, even photos of the hardware and mass budgets and performance summaries for components (the equivalent of spacecraft payload users’ guides for current expendable space launch vehicles, in terms of environment of the launch etc). It’s an excellent information base for analyzing the issues involved in someone else’s potential or theoretical program.

      Please, if Chevaline is not convincing to you as to the scope of difficulty here, tell us why.

    • George William Herbert (History)

      And, for the record, I have never been employed nor a contractor to nor paid in any way by ballistic missile defense programs.

      I got interested in it as an outgrowth of finding MAD morally offensive and repulsive in the early 1990s.

    • Mark Gubrud (History)

      Chevaline was specifically designed to ensure penetration of the Moscow ABM system, including two layers of high and low altitude interceptors with nuclear warheads. The Chevaline decoys had to be heavy enough to withstand early reentry and included gas thrusters to compensate for atmospheric drag. They were bus-released so that they could be placed on chosen trajectories and sufficiently separated so that they would not all be destroyed by one nuclear detonation. Besides these decoys, a number of other pen-aids were combined in the Chevaline package. The British wanted to be very sure of actual penetration and destruction of the Soviet leadership and central authority. Their budget for Chevaline was measured against the cost of more expensive alternatives such as doubling the total size of their offensive force. So yes, they conducted a lot of tests.

      North Korea or Iran do not need anything so elaborate and refined in order to confound current US BMD sensors. The defense definitely needs one interceptor per target object, and the targets do not need to be widely separated. The intercepts occur in hard vacuum and drag will not provide discrimination. Very thin balloons can be used, they can be highly randomized, and decoy thrusters or a maneuvering bus probably are not needed. The warhead can be enclosed in another randomized balloon, or coated with Eccosorb and hidden in a cloud of chaff and balloons of varying shapes, temperature, emissivity spectra, etc. MDA has never tested against anything like this and there is no evidence they ever intend to. Granted, the objects will have different signatures. The bad guys will not email MDA any info about which one corresponds to the real warhead.

      In the end, these threshold states that are only trying to achieve deterrence really do not need high assurance of penetration. Their objective is not to hold at risk the political integrity of the US, but only to prevent the US from willingness to bet that not even one American city (or for that matter a major city of a major ally) will suffer a nuclear detonation as the price for the kind of aggression witnessed 10 years ago this month.

      So they do not need to conduct lots of tests to prove that their CMs will work reliably time and again. They can just make sure we know they have something, and that there is a risk, too large to ignore, that it will work at least once.

      As of today, we can’t have confidence that our BMDS won’t fail catastrophically in case of even one North Korean ICBM with no CMs. If they conduct one or a few tests with CMs, we will probably know that the BMDS is useless.

      In the long run, CMs mean we can’t win even against North Korea any more than we could against the USSR. The more missiles, the more CMs, the more complicated this all gets, the more certain it is that they have us deterred because of the likelihood that one or more warheads will get through.

      Neither North Korea nor Iran can use their still hypothetical nuclear deterrent to enable aggression of their own. If they launch conventional warfare, we and our allies will respond, and beat them. If they use nuclear weapons, we will respond with nuclear weapons. And if they use a nuclear weapon against one of our cities, and they succeed, we will show them no mercy.

      The Countermeasures report has been out for 13 years. No technical rebuttal has ever been produced as far as I know. Instead of another round of posturing, how about some specific criticism of the specific analysis and proposed CM systems described therein. Please, not another round of “we know better but we can’t tell you why.” That card has been played, and your bluff has been called.

    • George William Herbert (History)

      Please stop waving the CM document as if it were delivered religious truth. The study it reports on was extensive in a number of technical areas, but also had quite a number of gaps and blind spots.

      For one, it did not adequately perform a cost/time/difficulty analysis of decoy development programs. The cost/time/difficulty analysis is more or less limited to the observation that the US was spending $3-400 million a year by the 1960s ($1.8-2.4 billion in 1999 dollars) (pp 35 and 145).

      There is considerably more written about that problem above than in the report.

      For two, its description of Chevaline as overly complex due to its necessity to penetrate endo- and exo-atmospheric interceptors with nuclear warheads is embellished. (pp 148). The nuclear interceptors problem is more of a warhead hardening issue (to prevent multiple warhead kills or far proximity kills from intercepts of nearby decoys) than decoy problem. The endo-atmospheric intercept decoy function required that the four high quality decoys include a small liquid fuel rocket motor to simulate real warhead upper atmosphere drag profiles; this adds complexity but not a billion dollars worth. The maneuvering bus, to configure the decoy cloud to arrive in a simultaneous time-on-target salvo (all warheads and decoys from all 16 SLBMs simultaneously), is similar to the problem of adequately dispersing “good decoys” in a cloud (requires 3-D maneuver and stability, multiple separation points, complex dynamics and kinematics), which is a lower delta-V bus of equivalent complexity to a MIRV bus, as I said above. They do note that the UK program had little bus experience prior to Chevaline (as they’d bought not designed Polaris) – this is similar to the difficulty for NK or Iran without prior experience.

      Again, the dynamics and kinematics of the bus, which is a rocket vehicle which is going to carry then separate a number of subpayloads, are a big deal, which are difficult to get right. As we stated repeatedly earlier, relatively simple space launcher shroud and upper stage / payload separations and stage separations go awry quite often even today for western launchers.

      If you want to get a feeling for the type and magnitude of challenges for spacecraft launchers and launch vehicle / payload separation, the relevant chapters of “Space Systems Failures” by Harland and Lorenz are a great education (minor disclaimer – I was at International Space University in Kitakyushu in 1992 with Ralph Lorenz, but have not worked with him since). There’s also “Disasters and Accidents in Manned Spaceflight” by David Shayler, though it’s more focused on manned missions. There are voluminous technical reports on separation and shroud issues and dynamics failures (less this week, with NTRS down due to Wolfe, but I digress).

      This is all stuff anyone can talk about. It’s discussing the particulars (MIRV or decoys bus) by analogy (maneuvering space launch vehicle upper stages, stage / payload dynamics for separations, separation mechanisms, spinup mechanics and dynamics etc). The analogy is close enough for the sake of argument, certainly within the precision of how closely the CM report addressed other problems they did look at closely.

      Again – it’s not that the CM report is wrong in particulars. It’s that it’s incomplete, and in an area that provides an important gap in serious strategy discussions. The time and cost of countermeasures programs bound the rate of deployment and risk they pose to our interceptor capability. The CM group did not attempt to assess that in depth.

      I’m describing parameters of such an assessment by analogy to Chevaline (a good, but not perfect match in terms of complexity and mission). I’m not in a position to stare inside North Korea’s program in depth – I have no special access to intelligence or the like here. But the analogy is at least of use to characterize the problem. We can use it to hypothesize about the characteristics and difficulties of a NK decoys program.

    • kme (History)

      I don’t think you’ve addressed Mark’s point that the North Koreans might not need to achieve a rock-solid, well-tested capability.

      For example, if you knew that they’d designed a complex warhead bus to deliver the decoy payloads but hadn’t actually flight-tested it, you might expect that it would have a high likelihood of failure – but would you be prepared to bet Seattle on it?

    • George William Herbert (History)

      KME wrote:
      I don’t think you’ve addressed Mark’s point that the North Koreans might not need to achieve a rock-solid, well-tested capability.

      For example, if you knew that they’d designed a complex warhead bus to deliver the decoy payloads but hadn’t actually flight-tested it, you might expect that it would have a high likelihood of failure – but would you be prepared to bet Seattle on it?

      NK is batting 0.00 on first-test successes of complex systems. Nuclear, missile, launch vehicle.

      From western program difficulties, they did poorly in early tests as well.

      From knowing principals at SpaceX, Xcor, Armadillo Aerospace, Masten Space Systems and others going back 20+ years, nobody gets things that complex right on first try.

      So, no. I do not hold credible that which they cannot successfully demonstrate via test.

    • Mark Gubrud (History)

      Ooops, I made one mistake. I wrote that, with our amazing HTK intercept capability, as compared with nuclear warheads, just one of which could wipe out an entire “threat cloud” if it were not widely dispersed, the “defense definitely needs one interceptor per target object.” That is incorrect, of course, since per the testing record, even when everything has been carefully polished, tweaked and tuned in preparation for a $150M test that will be postponed if a problem is found, the success rate is only about 50%. So, as we’ve discussed, to get 97% effectiveness, you need 5 interceptors per target object (i.e. warhead or undiscriminated decoy).

      But of course, one may ask whether 97% effectiveness is really good enough, if what we are considering is a war of choice against a nuclear ICBM armed Iran or North Korea.

      The Bush gang justified the 2008 $112M American ASAT demonstration on the basis that the hydrazine ice ball posed some kind of a threat to human life, citing (inappropriately) a NASA rule that you don’t launch if you pose a 1-in-10,000 or greater chance of killing someone, and in fact the actual risk of a human fatality if the ice ball had somehow landed intact was of that order of magnitude.

      So, should we be willing to launch a war of choice against J. Random Tinpot Dictator if it carries a 3% chance of a nuclear weapon detonating above a US city (something like 100,000 fatalities, to begin with)? And that’s with 5-on-1 targeting and assuming only one ICBM with no CMs. If they have two, it rises to a 6% risk; with 10 they exhaust our GBI magazine and it’s nigh certain we lose a metropolis or two. You can do these numbers more carefully, kidding yourself that they are that precise, but the qualitative conclusion is the same.

      CMs are just a cheap way to multiply the effectiveness of a small number of missiles as a deterrent to a war of choice. And again, it is only a deterrent to a war of choice. It does not provide a credible shield that enables free aggression. We will respond proportionately to a tinpot, as we would have to the Sovs but with much less to worry about.

      I wave the Countermeasures report because it is better than just waving hands. The report lays out very specific technical analysis and proposes simple countermeasures which it convincingly argues a) will be effective against the type of BMD sensors and interceptors we now have, and b) can be implemented at small cost relative to the cost of developing and producing ICBMs with nuclear warheads. And it has never been rebutted, by anyone, other than with the kind of rhetoric pasted up and down this column. In contrast, your argument that the CMs would be too difficult for the Norks, or would require too many tests, are just more hand-waving.

    • George William Herbert (History)

      Mark Gubrud writes:
      I wave the Countermeasures report because it is better than just waving hands. The report lays out very specific technical analysis and proposes simple countermeasures which it convincingly argues a) will be effective against the type of BMD sensors and interceptors we now have, and b) can be implemented at small cost relative to the cost of developing and producing ICBMs with nuclear warheads. And it has never been rebutted, by anyone, other than with the kind of rhetoric pasted up and down this column. In contrast, your argument that the CMs would be too difficult for the Norks, or would require too many tests, are just more hand-waving.

      a) Is correct. The report says that and is generally acknowledged as accurate in those regards.

      b) Is wrong. The report asserts that, that it could be implemented at low cost compared to nuclear and missile programs. Without providing anything approaching a detailed analysis either from first principles or a detailed historical comparison with, for example, the now-largely-declassified Chevaline program in the UK. The assertion in passing is not a priori credible, for the reasons John and I have brought up here.

      Your assertion that Chevaline is not a valid comparison is hinging on a couple of technical details (bus delta V and atmospheric penetration for the advanced decoys) that are down in the noise of actual technical analysis of the development cost and time envelope. We have explained why in more detail above.

      You are ascribing accuracy to the report in an area it did not investigate in detail. This is in your head, not the report.

      I understand why you want it to say that, but it doesn’t.

      We may not be able to convince you what you’re doing wrong here, but to the other readers here, please consider that John and I are experts in closely related fields, and we think these are hard problems. Not impossible, but long and expensive to solve and requiring open tests, not short and easy and something realistically done in secret.

      Perhaps we can get Jeff to host a study at CNS. I would be up for participating.

    • Jeffrey (History)

      Ah, well, that takes money. Not that it is such a long drive from where you live …

      I would only reiterate that my understanding is that much of the cost of Chevaline was driven by rad hardening the warheads. I need to find the source for that dim recollection.

    • George William Herbert (History)

      Jeffrey writes:
      Ah, well, that takes money. Not that it is such a long drive from where you live …

      Not that far, and I end up down in Monterey fairly often, my brother lives down thataways. I actually bought my current car at one of the used dealerships in Monterey.

      I understand the money thing. The question to me is whether there are sufficient complicated issues and a topic of enough importance to justify it, versus some dueling NPR or Bulletin papers.

      I would only reiterate that my understanding is that much of the cost of Chevaline was driven by rad hardening the warheads. I need to find the source for that dim recollection.

      I don’t know how the total accounting went. My impression was that the total program cost included a lot of warhead hardening, but it’s not clear if the warhead hardening was accounted against Chevaline, as it was part of the overall warhead development (which was, in general, not accounted against Chevaline costs, unless I misunderstood the accounting information).

      However, I admit not to having focused on that precise information previously, so I could be wrong. Ah, well, back to the primary and secondary sources…

    • Jeffrey (History)

      Well, we should have a coffee at some point at the very least.

      You’ve now triggered me to go through my reference collection and make one or two strategic purchases.

    • John Schilling (History)

      “So, should we be willing to launch a war of choice against J. Random Tinpot Dictator if it carries a 3% chance of a nuclear weapon detonating above a US city (something like 100,000 fatalities, to begin with)? ”

      Who is “we”? The decision will be made by One Guy Called POTUS. POTUS is extensively protected against the consequences of nuclear war; neither he nor his immediate family will be killed by anything less than a full-scale superpower exchange. About the worst threat that can be levied against POTUS is that he loses his job, perhaps serving the rest of his current term as the lamest of ducks, and his presidential legacy.

      Consider that this threat has never been carried out against a president who allowed hundreds of thousands of Americans to die on his watch (FDR, Lincoln), and only to one of the three who is seen to have seriously risked global nuclear war (LBJ but not JFK or Reagan).
      On the other hand, Carter’s handling of the Iran hostage crisis almost certainly earned him the “presidential death penalty”, and Bush the Elder’s failure to take down Hussein at the end of Gulf War I likely contributed to his single-term status compared to his son’s two terms.

      If the choice is between certainly failing to defeat a tinpot dictator who has sufficiently outraged the American people, and a 3% chance of the destruction of Seattle, well, one of these plans has a much greater 3% chance of leading to the highest possible penalty for POTUS and one of them a much less than 3% chance.

    • j_kies (History)

      The following should be considered when looking at Chevaline costs / complexities.

      1) The Brits were fully supported by the US data and design packages for everything below the separation plane. Their program was the add-on for the UK unique Penaids effort and other aspects of what came off the bus.

      2) Due to the fact of the Soviet A35/A135 system, pretty much any warheads had to be relatively rad-hard. I miss doubt that hardening testing was a Chevaline unique cost (but I could be wrong).

      3) You could reasonably compare the Chevaline costs / timeline to the basis Polaris A3 missile under it for that relative complexity argument.

      – Finally do recall Mr Montague’s comment ” they had a lot of resources and a lot of desire. And it didn’t work very well.” …based on observations during flight test. (yes that’s how I read Mr Montague’s comment, that his view was data driven but I cannot swear that’s what he meant unless someone was to ask him directly.

      So a first-world top-shelf science and engineering effort with informed US support cost ___ pounds over ___ years and provided a less effective countermeasure set than the British desired?

    • Jeffrey (History)

      Well, I think what is in order is to take a very close look at the Chevaline project and its cost drivers. I am going through my modest library and making one or two additional purchases.

      I’ll send you a note about why I said what I did, which is not the same thing as asserting that it is necessary correct. Time will time.

  11. j_kies (History)

    I wish my skills were up to being an effective con-artist… Effective deterrence is about perceptions, if we could manage appearances without the outrageous level of cost and man-power that we commit without gaining that perception of effectiveness we desire as a national tool then life would be better…

    I have diligently attempted to respond to every aspect I can with material pointing to open source / calculations you can perform. I am not asking for trust, I merely point out the historic rationale as to why the critics left the conversation when they chose to focus on things that cannot be discussed at the engineering level.

    • Mark Gubrud (History)

      I see no such pointers on the countermeasures issue. Others have looked at whether Unha-3 can carry a nuke to WDC, and they don’t think so. I won’t make an independent claim about that one.

    • John Schilling (History)

      I have been remiss in updating my models of Nork missile performance with what we have learned from the last Unha launch, but have now remedied that failure. Using best-guess performance and engineering data for the successful Unha-3 launched last December (with thanks to both David Wright and Markus Schiller for delving into details of both the recovered wreckage and observed trajectory of that launch), I think I have a pretty good model to work with.

      The third stage of the Unha-3 is clearly not suited for warhead delivery, but a two-stage Unha launched from Musudan-ri could probably deliver a 1,250 kg payload to Washington, DC. The North Koreans can almost certainly fit a small fission warhead and a low-beta RV into that mass budget; even the old Chinese DF-2 warhead (peddled far and wide by the A.Q. Khan network) weighed just 1,300 kg complete with RV.

      The Unha is too cumbersome and vulnerable to make for a particularly effective weapon. I am reasonably certain that the mobile KN-08 will be North Korea’s primary strategic nuclear deterrent. The KN-08 probably can’t reach Washington with a nuclear warhead, but if the Norks aren’t satisfied with burning out west-coast cities and insist on a shot at the White House, then yes, keeping an Unha on the pad with a nuke would be a way to accomplish that.

    • Mark Gubrud (History)

      John’s speed in modeling and drawing conclusions about Unha-3 range/payload capabilities is admirable. David Wright and Markus Schiller seem still to be working on theirs. Readers interested in what they have to say about the matter may view these links.

    • John Schilling (History)

      Turns out my speed was slightly less than admirable; I just found a typo in the original run file. Using a stock Unha with just the first two stages, the correct payload to Washington, DC, is only 200 kg. It can be pushed to just over 1,000 kg with the existing third stage, so the Unha is still a theoretically credible ICBM.

      But even more obviously a highly suboptimal one – that third stage, while it does offer substantial performance gains in that far corner of the range/payload envelope, comes with a ten-minute burn time that practically begs for boost-phase intercept from an Aegis ship offshore.

      If the Norks had been trying to build an actual ICBM masquerading as a satellite launcher, higher-thrust engines on the second and third stages would have given them better performance in the ICBM role while still allowing a token satellite deployment for plausible deniability. The present Unha design makes more sense as a satellite launcher / ICBM technology demonstrator than as an operational ICBM.

  12. Gregory Matteson (History)

    I have found somewhat of a dark amusement in watching this row. The credibility of both the North Korean ICBM posturing and the US GBI response is so low that the less thethered blogosphere and global media have been running conspiracy stories that ascribe totally extraneous motives to the parties. Example, http://www.russiatoday.com/op-edge/us-amd-korea-arctic-371/

  13. George William Herbert (History)

    Mark writes up a bit:
    The Bush gang justified the 2008 $112M American ASAT demonstration on the basis that the hydrazine ice ball posed some kind of a threat to human life, citing (inappropriately) a NASA rule that you don’t launch if you pose a 1-in-10,000 or greater chance of killing someone, and in fact the actual risk of a human fatality if the ice ball had somehow landed intact was of that order of magnitude.

    Minor factual correction – this is the Expected Casualties calculation (Ec) and the numerical threshold is 30×10-6.

    http://www.faa.gov/about/office_org/headquarters_offices/ast/licenses_permits/media/Ac4311fn.pdf

    • Mark Gubrud (History)

      The policy I was referring to was NASA’s policy; if you google “nasa policy 1 in 10000 risk” you will find a number of references to it. The rule you cite is FAA and it applies now to licensing commercial space launches, and is somewhat more stringent than the rule that officials cited as justification for the ASAT demo back in ’08.

  14. Mark Gubrud (History)

    “Your assertion that Chevaline is not a valid comparison is hinging on a couple of technical details (bus delta V and atmospheric penetration for the advanced decoys) that are down in the noise of actual technical analysis of the development cost and time envelope.

    No, not bus delta v but the need to use a bus at all. Atmospheric penetration also implies a heavier decoy and thus a more limited number of them. Plus they have thrusters to simulate mass.

    There are other factors that would be appropriate to consider in a formal study but aren’t worth it here.

    The main point, though is that the British wanted to be very sure of effectiveness against a two-tier nuclear tipped terminal phase intercept system, very sure of high effectiveness, and not just sowing enough doubt in the minds of the Sovs to keep them from launching a needless war of choice.

    We have explained why in more detail above.

    Nope.

    to the other readers here, please consider that John and I are experts in closely related fields, and we think

    ’nuff said.

    • George William Herbert (History)

      Mark:
      No, not bus delta v but the need to use a bus at all. Atmospheric penetration also implies a heavier decoy and thus a more limited number of them. Plus they have thrusters to simulate mass.

      What’s going to dispense all these things, if not a maneuvering bus… ? …

    • Mark Gubrud (History)

      Possibly just a dispenser that ejects the very light objects with some small velocity relative to the massive, shall we call it, platform. That could be gas pressure, a mechanical thrower, or the objects could be spring-loaded… many possibilities here. Or, they are dispersed by a small explosive charge. Or, the platform maneuvers, but randomly or with very little precision. One problem I see is that at some point you have to drop off the actual warhead. It may need its own small thruster so it isn’t the one big dim object right in the middle of the original course, and so the defense can’t take a clue from how the platform moves. But one can readily think of ways to address that. If it is a (random, low delta-v) maneuvering platform (instead of a “bus” how about we call it a “van”), the problem is relieved as long as the delta-p impulses are sufficiently randomized and there is an appropriate shift in their distribution when the warhead dropoff occurs so that there is no detectable shift in the delta-v distribution. This is still not exactly rocket science.

      Again, there are many, many ways this could be approached without doing anything that compares in difficulty to building an actual nuclear warhead that can reenter and explode, and an actual rocket that can carry it halfway around the world. If we were exploring this at all seriously, we’d go back and forth, and you’d undoubtedly find problems with some of the things I might propose but ultimately it is the defense that has the harder problem and the capabilities of our current BMDS are quite severely limited.

  15. Mark Gubrud (History)

    I’ll admit to having gotten pretty exasperated here, and I’m dissatisfied with my own responses, so for the sake of posterity, at the risk of reopening this endurance contest, after rereading the whole thing, here is an attempt at a (hopefully last) substantive reply.

    On Unha-3, again, I don’t know, and modeling this is not trivial, so I’ll defer to David and Markus and wait for their analysis, because I doubt I could do a better one. Also, I don’t doubt that North Korea can produce a rocket that can carry a ton of nuke to DC one day if they keep at it and don’t collapse first.

    On more reasons why invoking Chevaline as the answer to Countermeasures is like comparing Ferraris to Fiats: The Norks would not need to use a bus at all, or certainly not a high delta v, precision guided bus. The Norks would not be concerned with bringing a bunch of warheads down simultaneously on precise aimpoints, and when the interceptors are HTK the decoys can be very close together and each one must be engaged individually.

    One possible scheme would be a bunch of self-inflating balloon decoys, with randomized characteristics, which are launched with their inflation gas already in them, but they are kept collapsed by gas pressure under the shroud. The petals of the shroud burst and out come all these objects.

    Or, instead of balloons, you have self-unfolding wire frames that stretch the mylar films.

    Or, the objects are dropped off a kind of bus, but it’s essentially unguided and it just makes a random delta v maneuver between dropping off each object. The warhead is a fission device aimed at a big fat city, so up to a few hundred meters of extra aimpoint randomness doesn’t change the equation.

    Another thing they could do is just release a big cloud of radar chaff and some IR/optical obscurant, and now we can’t see where the warhead is in that cloud. Coat it with Eccosorb so the radar can’t see it, even without the chaff. Throw in some balloons of varying IR and radar brightness just for fun.

    Against all the claims of expertise I am going to claim common sense that such tricks are much lower tech than building the rocket and warhead, and simple enough that they might go without any flight testing at all, which would make it especially unlikely that the BMDS would be prepared to cope with them. Of course the mechanisms would be tested on Earth but there is probably not even any reason why they need to be tested in zero-G.

    The deterrent value would be lost if we had no inkling they were there. But there are so many possible variations that the Norks could let us know they were doing such things and keep us guessing as to the exact details.

    Now, of course you can come back with the usual replies, “We’re experts and it’s harder than you think,” or “ROTFL,” or some equivalent. Yawn. Good night.

    Maybe some day there will be enough funding for a serious sound-off between proponents and critics of BMD to dig down and settle this. The mind reels at the comparison between what that would cost and the amounts being flushed down the missile defense hole every year.

  16. Gregory Matteson (History)

    I continue to think that the non specialist can intelligently evaluate and comment on these matters, especially in my case by looking at history, and the old stuff that has been declassified. I have, however, also spent much of my life around engineers.

    Blue Sky, arm-waving engineering will only get you so far, as will the advantages of PCs and modern technical information. The MTCR would be a meaningless exercise if stringent testing and development could be bypassed.

    In regard to history, I refer back to Sputnik and the Soviet Space Challange, and, The Soviet Space Race With Apollo, both By Asif A. Siddiqui, both histories written for NASA. Reading these histories helped give me the gumtion to make comments here.

    Those histories make it clear, among other things, that the early ICBMs were stupendously inaccurate, and only became accurate after years of actually flight experience, including in the manned space programs. Basic, truly ‘ballistic’ ICBMs have CEP in the tens of Kilometers, not hundreds of meters, and more things can foul out an untested design than you can easily imagine.

    As Kennedy said “We do these things not because they are easy, but because they are hard”

    One of the stated reasons the Japanese gave up on their small shuttle was because the major powers, under the MTCR, would not help them with the re-entry dynamics.

    • Mark Gubrud (History)

      Gregory, you are missing the point. I am talking about things that are really, really simple. Much simpler, and much less constrained and demanding in terms of precision, than a rocket or a nuclear warhead.

      You are right that the early CEPs were kilometers and we might not expect much better of a North Korean or Iranian ICBM. That was my point! Against our HTK BMD systems, the objects in the “threat cloud” do not have to be separated by many kilometers as they do when the interceptors are nuclear. And they can be separated by hundreds, even thousands of meters, and the real warhead’s trajectory can be randomized by that amount, without it making a significant difference, because they weren’t going for high accuracy in the first place. Therefore you do not need a high precision bus that can carefully control the trajectories and spread them out. Dispersing the objects with a small explosive charge, or an unguided, fairly random thruster, is good enough.

      What George et al. are arguing here is that North Korea or Iran can’t equal what Britain did with Chevaline, but that is irrelevant. What they can do, yes, possibly even without any flight testing, is good enough to throw a monkey wrench into the exquisitely fragile engineering that allows MDA to stage its “hitting a bullet with a bullet” demos.

      Again, it’s true that the laws of physics are the same, but the engineering challenges for the defense are much harder, and the defense has to work perfectly every time. In this context, the offense doesn’t really need to even work at all. Look how freaked out everyone is about North Korean ICBMs that probably don’t even exist (yet). All the “offense” needs is to plant a good healthy dose of self-doubt in the mind of POTUS, so that whatever brouhaha erupts can be settled short of total war.

  17. ara barsamian (History)

    Have we lost all sense of reality? Pissing away a billion dollars on a bunch of savages that can’t feed their own people, particularly when we don’t have money to spend on our returning veterans or retirees on social security, or on fixing the third rate education for our kids is an abomination.

    We have a tendency in US to look for a technical fix when none is needed. How about completely ignoring DPRK? This includes everybody, government, diplomats, news media, not providing any aid (food, fuel, etc, and stopping all contacts?

    Also making it explicitly clear to DPRK that they will be incinerated…which our politicians don’t always have the “gonads” to do?

    • John Schilling (History)

      Aside from the dismissive and counterproductive stereotyping here, there seems to be a flaw in the logic.

      What, exactly, is the value of threatening to incinerate people who in this conception of reality are already starving? Dead is dead, and of the specific choices offered, most people would probably take death in battle.

  18. Gregory Matteson (History)

    Savages? Unfortunately both history and etymology demonstrate that ‘civilization’ is about flush toilets, public water supply and paved sidewalks; and not about being nice. We too can be quite savage.

    Also unfortunately North Korea can not be ignored. Incompetents and ignoramuses do not engage in a strenuous program of nuclear testing and rocket development. It seems to me that the point of this thread is to try to sort what is posturing from what is meaningful. A billion dollars isn’t really a whole lot if it reassures our allies and puts pressure on the Chinese. Granted putting pressure on the Chinese deterrent might have unintended consequences. And, we may secondarily get other benefits from an Alaska buildup.

    There can also be questions about our GBI program. It is well to remember that the Obama administration acted rationally in backing away from the original plan for East European ABM deployment. The only ABM we have that we have demonstrated works, and that we have reasonable confidence works, is the SM-3, which is a terminal area defense weapon, not suited to midcourse interception. Awkward.

    • Jeffrey (History)

      “Savages? Unfortunately both history and etymology demonstrate that ‘civilization’ is about flush toilets, public water supply and paved sidewalks; and not about being nice. We too can be quite savage.”

      Ain’t that the truth.

    • George William Herbert (History)

      Gregory Matson writes:
      The only ABM we have that we have demonstrated works, and that we have reasonable confidence works, is the SM-3, which is a terminal area defense weapon, not suited to midcourse interception. Awkward.

      I believe GBI works, but at a reliability level that is atrocious.

      Pk of 50% – including all the launch and in-flight failures – is not in the greater scheme of things acceptable. But it’s the hammer we have.

    • John Schilling (History)

      I don’t think “Pk = 50%” really captures GBI, and while I am sometimes guilty of the same oversimplification I have to cringe a bit every time I see proposals that, OK, we just fire 5 GBIs at every target and that gives us PK = 1-0.5^5 = 97%.

      What we’ve really got is an unknown, inadequately tested Pk hiding under a bimodal distribution. One thing the US aerospace industry does really well is, even when it is building crappy designs, building them exactly the same way every time. And BMD engagements are relatively clean; there isn’t much environmental or circumstantial variation between one warhead and the next in a North Korean strike.

      So, 50% chance we have a GBI design that actually will work quite well (80+% Pk) under the circumstances it will actually face on The Day, and 50% chance that we screwed up and have something that will be lucky to deliver a 20% Pk in the field. If that coin comes down tails, won’t matter how many shots we take.

      What I want, given the present force structure, is GBI backed up by SM-3. We of course can’t get full CONUS coverage with SM-3, but it looks like the Norks are going with an ICBM that can’t give them full CONUS targeting, either. It may be that an Aegis cruiser off the Pacific Northwest, can give us an independent backup against the targets the Norks can actually hit.

      THAAD sites in the prime target areas might be worth considering as well.

    • Gregory Matteson (History)

      The sticking point on the question of GBI performance that I see is the calculation that Chinese planners have to make. Since entering the nuclear deterrence game, they have engaged in a strategy of minimum credible deterrence. Now, just as we have to ask ‘what if the North Korean missiles work?’, they have to ask, ‘what if GBI works?’

    • John Schilling (History)

      The quasi-good news on that front is, China is vastly richer and more technically sophisticated than North Korea. GBI as presently deployed has a fair chance of working against half a dozen KN-08s. It plausibly might, barely, work against two dozen DF-5s. It won’t work against two dozen MIRVed DF-5s, or DF-41s, or JL-2s.

      Even if American missile defenses are substantially increased in a full-up arms race with North Korea, China would need to make relatively modest force changes to maintain the status quo, and the net effect would be only modestly annoying to the US and China both even as it pushes the DPRK towards the same end as the USSR.

      If we have to do this, serious effort should be made to diplomatically engage the Chinese along the way. Modest annoyances shouldn’t be left to fester, and I can see some diplomatic approaches that should be fairly effective.

  19. Anon2 (History)

    First, this was an excellent thread. Thank you all for the contributions.

    It seems to me that no one has mentioned the unmentionable, a scenario for minimizing risk may be a preemptive attack by the United States before North Korea is certain of its ability to wage a nuclear war. I believe that this previously radical alternative can no longer be ruled out.

    As John Shilling has pointed out, faced with death by starvation or death by battle, most men would choose battle. If that is the long term situation for most of the North Korean population short of a complete cave in by the Allies to the nascent nuclear threat (i.e. ransom tribute), then war is highly probable. It would be better to stage the war when NK has only a few warheads, limit range missiles, with a low probability of delivery on target due to internal failure of the missiles and a still effective BMD shield. NK most certainly has a crash program to build sophisticated (i.e. Chevaline) and unsophisticated (i.e. balloon) decoys. It is unlikely that they are weaponized. NK has inadequate (short range KN-08) and untested means of delivery. While reasonable national intelligence would show the possible number of weaponized warheads, even without this, we (the non-privy to intelligence community) know that they are building more as we speak, and doing their best to make them more effective and with longer range.

    Following Shilling argument — if sanctions continue and it is true that millions of North Koreans will die anyway of starvation, for the United States it is better if they die without NK successfully delivering a city buster to a U.S. city as they have already said that they would do, and made multiple videos showing an imagined results. These videos and the “propaganda” cannot be for internal consumption (because no one outside of the elite have internet to view it). In fact, with effective propaganda for internal consumption, it was entirely unnecessary to launch a missile or test a nuclear device at all — it can all be “made up” for the low-res TV news, radio, and Pravda newspapers that the typical North Korean citizen and soldier uses for their “Why We Fight” purposes.

    Imagine the following two scenarios:

    Scenario one is:

    Say that the probability of NK launching an attack in the distant future is 50%. Say that the probability of successful delivery of multiple NK city busters is 75% in the future. This means a 37% chance of the destruction of multiple city busters.

    Scenario two is:

    U.S. launches a preemptive attack. Probability of NK delivering at most a Hiroshima size weapon +/- 25 miles from target is < 3%.

    Scenario three is that North Korea by magic becomes a peaceful member of the community of nations. Probability what is that?

    What is the least risky choice?

    North Korea with the bluster have made the POTUS choice clearer IF we take them at their word. Their word is that they will attack when they have the means.

    This is an inherently unstable situation. I believe that the B-2 and B-52 overflights are a type of signalling. It does not appear to be working.

    If there is any back channel through to North Korea to defuse the situation, now is the time.

    I repeat my earlier statement — allowing NK to successfully test a weaponized warhead and to flight test a 3 stage Unha as a proxy for a two stage KN-08 with essentially the same engines was an epic fail by the United States.

    The final potential epic fail if we believe John Shilling's model is accurate: allowing North Korea to assemble anywhere a 2 Stage Unha (with a 1200 kg WDC throw-weight) on the pad with a hidden surprise under the payload shroud. We may have made it to the point where the assembly of a missile on the pad, or the erection of a KN-08 in the field is an act that demands prompt neutralization of the missile.

    I know that if I were the United States I would have every "national intelligence means" monitoring the shell game with KN-08 being constructed and then rolled into tunnels, with a prompt part of the war plan to place laser guided bombs repeatedly at each such tunnel entrance in the wack-a-mole. I'd also have the equivalent of airborne seal-teams to manually neutralize the equipment buried deep in the tunnels.

    I am extremely concerned about this. I live in a target. I wish I lived in Monterey.

    • George William Herbert (History)

      Anon2 writes:
      It seems to me that no one has mentioned the unmentionable, a scenario for minimizing risk may be a preemptive attack by the United States before North Korea is certain of its ability to wage a nuclear war. I believe that this previously radical alternative can no longer be ruled out.

      Both this idea and Ara’s “ignore them” are clearly “in play”.

      My reason for focusing my discussion on the defensive approach is that I find spending any reasonable amount of money ($3,5,10 billion) on defenses a more ethical approach than a war of choice or the risk associated with ignoring them.

      I believe that ignoring them risks disaster – their recent history of cross-border attacks certainly indicates that they are willing to consider violence for pure political theater, as well as internal stability purposes. Disaster with a nuclear ICBM (or potentially ICBM, and certainly regional range missiles) opponent strikes me as something to spend significant effort avoiding.

      I believe that a war of choice here and now would risk China intervening on the other side, and even once the situation evolves to where they would prefer no NK rather than NK, the Afghanistan and Iraq wars show the potential difficulties. We can clearly defeat them, but long term instability would not be good. China does not want long term instability on their border. SK does not want long term instability internally or on their border.

      The advantage of the defensive option is that it’s just money – not us taking an aggressive stance or offensive responsibility – and if NK choses to try to play the “more warheads and missiles and decoys” game they can spend themselves into the ground and be unable to keep going, a la USSR in the late 80s. This might be the smoothest path to regime change. They are taking steps down the road, so we just spend the money to keep them going down it until they collapse.

    • Jeffrey (History)

      I support the notion of providing a defense at a reasonable cost. We got a bit sidetracked by figuring out how to calculate cost-effective at the margin, but what I really worry about is the GMD system serving as a form of technological escapism from dealing with North Korea. That’s undesirable both because we may fail to take other actions that might more directly address the threat or may act recklessly on the basis of a false confidence conferred by the existence of a midcourse defense. All the while allocating scarce defense dollars suboptimally.

      So, I still struggle to assess our missile defense efforts in terms of matching investments to clearly defined outcomes. I don’t really care if we have to spend more than the North Koreans, though I do worry about being budget constrained in the grand scheme of things. What I care about is that we are clear about what we are, and are not, purchasing.

      If we had a broad North Korea policy that I saw as competent and clear-eyed, I would not object to spending money on defense at an unfavorable rate to marginally reduce North Korea’s capacity to harm us — as long as we are clear about what we are purchasing. I do, however, object to spending that money absent a plan that envisions subsequent moves, largely as a short-term political gesture that substitutes for thinking strategically about this problem.

      It’s a strange answer, but I am trying to express why the very logical arguments you’ve made leave me somewhat unpersuaded.

    • John Schilling (History)

      Is there a competent and clear-eyed North Korea policy you think we should be pursuing? Because, long term, I don’t really see one. I can sometimes see merit in forgoing marginal improvements in one’s military defenses in order to force emphasis on more productive diplomatic approaches, but only if I am highly confident that the diplomatic approach can really be made to work.

      That the DoD is going to squander most of the money we give it for national defense, is unfortunately a given. Fortunately, we’re filthy stinkin’ rich, and our enemies aren’t paragons of efficiency either.

    • George William Herbert (History)

      The disadvantage of my plan is that it allows us to not resolve the grand strategy problem.

      Its advantage is that it allows us time to resolve the grand strategy problem.

    • Mark Gubrud (History)

      1. Stop wasting money on missile defense. Deterrence works.

      2. Respond to any North Korean provocations with measured, in-kind responses.

      3. Meet them at the negotiating table for a peace treaty.

      4. Kill them with kindness, or at least trade.

      Distasteful as it may be to accept that the regime will only fall from within, and that the best way of promoting that end is to engage and do business with them, it is still better than any proposed alternative. Also, it has worked before. In many places, it is still a work in progress. But overall, it has a better record than the alternative of endless sustained hostility.

    • George William Herbert (History)

      Mark writes:
      1. Stop wasting money on missile defense. Deterrence works.

      The assertion “Deterrence works” is a canon of certain arms control factions, however, anyone who has followed deterrence studies in depth for the last 20+ years (much of which has been repeated on this blog in some form) should be aware of how many times Deterrence has Very Nearly Failed.

      Between adversaries who were much much more calm and sophisticated than NK is.

      We’ve very nearly had more than one nuclear war by accident. It is not OK to rely on that assertion.

      It’s horribly impolitic to insist on this point because it argues either for the hated defense systems or the impractical near-complete disarmament, but the point must be made (and again, and again, and again if necessary).

    • FOARP (History)

      “2. Respond to any North Korean provocations with measured, in-kind responses.”

      You mean, bombard their islands, sink their warships, assassinate their leaders, raid their territory?

  20. j_kies (History)

    Passions and concerns are appropriate, Ara expresses dismay at the expenditure of funding that he does not see a reason for, Anon2 has concerns as to the immediacy of threats based on proximity and perhaps familiarity with those making threatening statements.

    I see three threat ‘facts on the ground’ that inform choices: First, NK has demonstrated a relevant ballistic missile capable of ranging the US mainland with significant payload (as of 12/12/2012). Second, NK has demonstrated (from seismic data on 2/12/2013) a nuclear explosive with similar yield to first generation fission devices, which they claim to have miniaturized. Third, NK is now apparently threatening to employ nuclear weapons against the US as a preemptive attack. While ‘clumsy’, this appears a credible threat (capability and intent) that we should respect as none of the other elements of an operational ICBM would be seen unless they chose to show them to us.

    As an American taxpayer, I share Ara’s dismay at the concept of throwing money at problems, especially without some reasons to believe that isn’t equivalent to just ‘pissing it away’. Missile Defense was selected by previous elected leaders as a preferred means to address missile threats; it can be seen as an insurance policy and a means to select less catastrophic responses to attack. I wish that such fundamental decisions as to what risks and what responses we choose to accept could have been popular poll as it’s a stretch to trust partisan politics to stay out of making such issues a ‘Pork’ question.

    I doubt the US has ever been serious enough to assign national importance to BMD. Missile interception is hard; the overall problem has been compared with the original difficulty of nuclear weapons development in the Manhattan Project. BMD had many contenders for the ‘Les Groves’ role but no similar persons exist for the scientific and engineering role of Oppenheimer. While correlation is not causation, BMD seems to have the worst test record of any current weapon system and that is without discussions of sophisticated threats. The best arguments against BMD would seem to be ‘don’t do a weapon system if you aren’t serious about having the people needed to make it work’.

    At the same time, it’s inappropriate to react to ‘10’ tall North Korean’ threats involving mobile ICBMs of a type that neither the US or the Soviets ever did, with ‘easy to do’ countermeasures, all with no testing whatsoever with starving scientists and engineers. Yep, can I buy that bridge please?

    Make sensible informed choices on where to expend resources, I get paid with or without BMD so I want to see a national policy choice, whichever way it goes.

    • Mark Gubrud (History)

      NK is now apparently threatening to employ nuclear weapons against the US as a preemptive attack.

      A candidate for the most certain thing in this entire discussion is that if the DPRK launches a nuclear weapon at the US, whether they think it’s preemption or not, and probably without regard to whether it gets intercepted or whether it works, the history of the DPRK will promptly be brought to a close.

      How exactly that will be brought about is somewhat less certain; as someone pointed out, with China next door, and South Korea and Japan, too, we can’t just go all nukespaz. But we will take them out. For sure.

      Now why would a 26-year old who has just inherited a dictatorship want to do that?

    • George William Herbert (History)

      Mark, the essential nature of this problem is whether we trust that traditional deterrence – which we find stable-ish with Russia (and the predecessor USSR) and with China – will be stable with NK.

      The USSR and China do not on an annual basis bombard neighboring territory or sink ships, nor do they threaten war on a daily basis.

      That is it self-evidently (to us) suicidal behavior does not exclude it from happening. This is the essential failure of “deterrence theory” as applied to the real world.

    • Mark Gubrud (History)

      The DPRK’s inimitable over-the-top style of brinksmanship and bluster does not persuade me that the leadership there is actually suicidal. And I beg to differ with you on whether Russia, China, or for that matter the USA don’t “threaten war on a daily basis.” Why, just the other day I believe I heard that our Nobel Peace Prize-winning POTUS made some comment about options and a table. Sure, that’s milquetoast compared with Pyongyang rap, but then again, his immediate predecessor invaded not one but two UN member states which had not, nor had they threatened, nor were they in any position to militarily attack the United States, though one of them was being less than fully cooperative in our pursuit of a wanted criminal. Maybe I shouldn’t open that last can of worms, but the fact is, by any reasonable standards, the US threatens war 24/7/365, anywhere on Earth, and actually practices it in a seemingly ever-expanding list of locales, which makes the soft-spoken words by the man with the big stick a lot scarier to a lot more people than all the shrill rhetoric from the little-stick guy.

    • George William Herbert (History)

      There is a world of difference between “actively suicidal” and “sufficiently reckless or dangerous as to accidentally trigger a war”.

      Again – looking back at the wide history of deterrence failures or near-failures – competent stable leaders have nearly tripped over the edge. Repeatedly.

      One does not need to ascribe insanity or malice to NK to qualify them as at the very least far less competent and stable by larger world standards. Ergo, the risk seems much higher.

      It would be bad enough if they were just more at risk in normal times, but they on a regular (annual or more often) basis are intentionally *starting* the sort of incidents which are conceivably starting points for a tip-over scenario.

  21. Anon2 (History)

    WSJ:

    “In North Korea, Mr. Kim ordered rockets to be on standby to strike U.S. bases in South Korea and the Pacific, as well as the U.S. mainland, state media reported. The order came after Mr. Kim and senior military officials held an emergency meeting in the early hours Friday, according North Korea’s official KCNA agency.

    The B-2 flights were “an ultimatum that [the U.S.] will ignite a nuclear war at any cost on the Korean Peninsula,” Mr. Kim said at the meeting. KCNA said the leader “said he has judged the time has come to settle accounts with the U.S. imperialists in view of the prevailing situation.”

    ————

    “Time to settle accounts.” Who in their right mind makes that kind of threat in state media? They appear to have put their nuclear missiles (to the extent they have them ready) on Defcon-3.

    Now they should have POTUS’s complete attention with hourly briefings.

    It would be negligent not to follow this with all the resources the national intelligence apparatus can muster.

    If POTUS has not sent a back channel ambassador, a real diplomat, POTUS is negligent. The ambassador needs to take any personal risk (including being held hostage by NK) to prevent the worst outcome by engaging Kim Jong Un himself. Send Biden. Send Clinton. Send someone.

    • George William Herbert (History)

      Can’t be anyone in our current chain of command / succession; Bill or Hillary Clinton would potentially be OK, Bill Richardson, or any number of others. Not Biden (or Kerry).

    • Mark Gubrud (History)

      There is no new issue between the US and DPRK. The only crisis is the DPRK succession crisis. Kim is trying to show that he is in command of the whole useless military machine which is the DPRK’s main enterprise. Recent US moves have been to underscore what would happen if Kim made good on all these ridiculous threats. Of course he responds by upping the volume, but nothing will happen.

      I’m all for sending high-level diplomats, but this might not be the best moment for that.

  22. Daryl Press (History)

    I think where Mark and other committed opponents of defenses go wrong is in their common claim that an imperfect system is close to useless, because whatever the residual risk of leakers remains, that risk will be enough to deter the U.S.. As they say, a U.S. President would not risk the loss of a couple U.S. or allied cities…so we’ll be deterred.

    The flaw in the argument is that a conventional war may erupt whether we want it to erupt or not, and NK might escalate such a war — just as NATO planned to do when NATO was outgunned. In those circumstances, we would all be very happy to have defensive capabilities, even if imperfect. (Whether it’s worth spending $X billion for that capability is an argument that turns on the technical debates those above were arguing about.)

    But the trope of: “our missile defense is only helpful if we can be sure of getting them all” — which is not just something Mark writes, but others too — is wrong. The US doesn’t want a war; the US doesn’t want a nuclear war. But we are not the only ones with a vote. Imperfect defenses can be quite valuable if war is thrust upon us.

    • Mark Gubrud (History)

      Daryl,

      I know you are familiar with the concept of “escalation dominance.” We have it, the North Koreans don’t.

      If a conventional war between the US/RoK and DPRK should “erupt”, North Korea’s problem is going to be to terminate that war without the termination of its own existence. Ours will be to terminate the war without too much damage to South Korea and with no use by anyone of any nuclear weapons. The incentives, nay, imperatives, for both sides to avoid too great an escalation are clear.

      Should our side fail to communicate that it is willing to return to the status quo ante, leaving the Pyongyang regime in place, the danger of their resort to nuclear use becomes real. However, should they do so, or even unambiguously attempt to, there will be no possibility that we would stop short of destroying the regime.

      I think the leaders on both sides understand this. I see no issue for the two sides to fight over that would override these basic considerations.

      Kim’s problem is to hold on to power, and his regime’s problem is not to fall apart in the slightly longer run. They can’t be thinking they are going to take over the South. Likewise, we should not be entertaining any notion of starting any war of choice with North Korea, nor kidding ourselves that missile defense is going to give us that option.

    • George William Herbert (History)

      Mark:
      I think the leaders on both sides understand this. I see no issue for the two sides to fight over that would override these basic considerations.

      I find your faith in deterrence disturbing and frankly obsolete.

    • Daryl Press (History)

      Mark,

      Earlier you were asserting that we must acknowledge the reality of MAD vis-a-vis the North Koreans. Now you’re saying that we have escalation dominance. Which is it?

      Earlier you said that if they used NW during a war, we (or China, or someone) would do something to end their regime. (Perhaps. Though Pakistan thinks escalation is a good strategy for stalemating India. Russia says the same thing about conventionally superior enemies. And NATO thought it was a good strategy for stalemating the Soviets.) But suppose you’re right: when we’re doing that thing to end their regime, won’t you hope we have missile defenses — even imperfect ones — to help us with any NK nukes that they fire at our allies?

      Again, there is a trope that steadfast opponents of defenses often assert: namely, that imperfect missile defenses are irrelevant because they will never provide enough assurance to free us to do…something. But if that something is retaliation — or more precisely, damage limitation — then aren’t imperfect defenses valuable?

      Daryl

    • Mark Gubrud (History)

      Daryl,

      If you are challenging my use of jargon, I think you should take care not to accuse me of using jargon that I didn’t use (MAD). North Korea certainly does not have the ability to destroy the USA, and we do have the ability to destroy the DPRK. However, the potential costs of doing so are so high that we are certainly deterred from doing it as long as we don’t have to.

      Under what circumstances will we (at least believe we) “have to”? A full-scale conventional assault on the South might qualify, although we would more likely try to repulse it without any nuclear detonations. Any use of a nuclear weapon, or even just the attempted use of a nuclear weapon, by the DPRK probably qualifies. A nuclear attack on a major city in South Korea or Japan, or a major US military base, almost certainly qualifies. A nuclear attack on any of the 50 states of the US absolutely certainly qualifies. Even if the Norks have a few in reserve and might manage to get in another blow, and even if that is irrational. In such a situation we would be acting as much out of emotion as of reason.

      If you want to know what I mean by “escalation dominance” as well as by “deterrence,” in this context (I think the jargon is blunt enough and the concepts fuzzy enough that their full and exact meaning must be a function of context), reread what I wrote about what each side’s problems will be in the event that something we might call a war “erupts”. Let me know which part of that, if any, you disagree with.

    • Mark Gubrud (History)

      Oops, I overlooked your other important question.

      The problem with the argument that, all else being equal, it’s better to have “imperfect” missile defenses which, who knows, might just work in case the Norks launch and we’re lucky that day, is that all else is not equal.

      The pursuit of missile defense has a significant effect on the level of investment by all parties involved in pursuit of this otherwise pointless confrontation, and thus on the level of hostility, and on the complexity of the confrontation and possibilities for war to “erupt.”

      Then there is the possibility that a future POTUS will take greater risks if she thinks the missile defense has a good chance of working; I’m not sure if I believe that but it is plausible, especially in an otherwise difficult situation such as where the smoke and volcanic ash of war are already spewing forth.

      Missile defense undermines deterrence, because it signals an unwillingness to accept the status quo and the reality of deterrence, requires a multiplication of effort by the “offense,” and involves the deployment of encircling radars, interceptor missile sites/ships, space systems, etc.

      Possibly the most dangerous proposal that has been put forward is to base interceptors on manned or drone aircraft which would be flown just outside Iranian or North Korean airspace, and possibly even penetrating their airspace on the assumption of stealth. Such weapons would make inviting targets for an incident in peacetime, and even more so in a crisis, and might well trigger the war and the nuclear attacks we’re trying to avoid. Worse, they would obviously be candidates to go in with the first wave, along with strike aircraft or missiles seeking to destroy the nukes on the ground. Movements of such weapons could thus be mistaken as a signal of imminent attack. To some extent, this risk also exists with ship-mobile interceptors.

      You may ask why I would raise that, since the US isn’t currently planning to deploy air-based interceptors. But if we keep chasing the White Rabbit, which hole will we not eventually fall down into?

  23. Ara Barsamian (History)

    I’m afraid history is the best guide to what can be done re DPRK “threat”. Recent dictators, from Hitler to Stalin to Mao were appeased by the West, and what did it get us? Peace in our time?

    On the other hand, while they killed tens of millions of their own people, they also valued their survival above everything else…I am pretty sure that little Kim and his family’s highest priority is their survival…and they make high-decibel bellicose noises just to be “noticed”, and maybe get a gullible West to pay ransom in form of fuel, food, etc….

    DPRK’s little incursions on SK are similar to the probing we got from the USSR to see how far they can push and get away with it without threatening their existence, e.g. Berlin blockade…

    The point is that appeasement does not work, whether Hitler or Stalin or little Kim, but overwhelming brute force does….like the threat to incinerate them…

    For the technophiles, if we’re going to waste billions on interceptors and technical “fixes”, might be worthwhile to at least revisit SDI and particle and beam weapons and lessons learned, although I believe that Russia and China will feel threatened…

    The bottom line seems to be that we’re paying way too much attention to DPRK and its barking while neglecting our society’s needs…

  24. j_kies (History)

    Ara

    “…if we’re going to waste billions on interceptors and technical “fixes”, might be worthwhile to at least revisit SDI and particle and beam weapons and lessons learned,…” – err as someone that tracked onto all of that from the SDI days on – that’s a considered ‘HELL NO’ the real lesson learned is those certainly don’t work.

    Directed Energy happens to have the unique aspect of being more dangerous where you make it than where you send it.

    O’Dean Judd used to have a sign outside his office, it said “If it involves a Laser it won’t Work”, “If it involves a Plasma it certainly won’t work”, the final statement involving discrimination had to do with the deficiencies of sensing available at the time.

    • George William Herbert (History)

      j_kies wrote:
      O’Dean Judd used to have a sign outside his office, it said “If it involves a Laser it won’t Work”, “If it involves a Plasma it certainly won’t work”, the final statement involving discrimination had to do with the deficiencies of sensing available at the time.

      That was in the era when lasers were largely chemical lasers. The solid state ones are remarkably reliable, and big enough for weapons purposes now for at least tactical ranges.

    • j_kies (History)

      George

      They considered all the means of pumping a medium that you might wish. ‘Tactical ranges’ for what purposes against what targets? Shooting down ballistic missiles are certainly not within the credible applications according to the NAS/NRC (I strongly agree) and I think they were pretty generous as the real-world issues of contamination control render laser weapons exceedingly dubious especially if you consider maintaining them with military crews on military platforms.

      Lasers are essentially remote spot heaters,

    • Mark Gubrud (History)

      I like to measure laser beam power in hair dryers. One hair dryer = 1.5 kW. So now we have solid state lasers with beam powers of up to around 100 kW. That’s like focusing the heat of 67 hair dryers on an object that’s designed to pass through the atmosphere at supersonic speed, or reenter at hypersonic speed. Unless, of course, the object has, say, 90% reflectivity (not hard to do), in which case it drops to 6.7 hair dryers.

      Also, since said lasers are maybe 20% efficient, the laser itself has to dispose of the heat of 333 hair dryers. Which it can do, even though it is not much bigger than the object it’s shooting at.

      Lasers may have some utility as soft-target weapons, at moderate range and in fair weather.

      BTW, chemical lasers are still the only thing available in the MW class, and no, they don’t work very well. They’re big, fragile, spew huge clouds of toxic exhaust and run out of fuel quickly, unless they are ground-based and have huge tanks. A little bit of dirt on the optics will cause them to self-destruct. Then you have all the by now familiar beam propagation problems.

    • George William Herbert (History)

      There are two issues with lasers; throughput and optical range (Rayleigh range for the wavelength and effective aperture). (Pointing and sensors and the like are solved, if not always cheap).

      For reasonable cleanliness (industrial / astronomy equipment / etc) hundreds of KW and up through small apertures is unwise.

      For tactical use, “laser searchlights” where you give up trying to combine it through a single small laser but instead use large arrays of mid power lasers (100s of watts, KW range) work just fine, because rayleigh range out of most diode laser units with adequate front end optics is tens of miles. You just need to point all the individual units in conjunction. Lest you argue that this is not practical or too expensive, Jordin Kare (long formerly of LLNL and Intellectual Ventures, now at Lasermotive) has been working on this for spacecraft propulsion out to hundreds of km range (past individual unit rayleigh range) and has a pretty good workup of unit costs. You can do this with a single optical table mounting all the units and a single-axis distance/convergence adjustment if you want.

      For hundreds of KM it’s harder. For thousands, a fair amount harder, but the A=r^2 of the apertures required for rayleigh range for those distances make the power density problem fade…

    • George William Herbert (History)

      Also:

      Mark writes:
      Also, since said lasers are maybe 20% efficient, the laser itself has to dispose of the heat of 333 hair dryers. Which it can do, even though it is not much bigger than the object it’s shooting at.

      Modern laser diodes are a lot more efficient than 20%, Mark. 40% was 10 years ago; 50% is common commercial now, 70% is nearterm.

      That is not to say that the thermal waste heat at transmitter end is not formidable, but it’s getting far better…

    • Mark Gubrud (History)

      Diode lasers are the only type that can have efficiency approaching unity. The 70% you cite is in the upper range of what is possible now for “wallplug efficiency” defined as the ratio of DC volt-amps in to photon power out, not taking into consideration the power needs of cooling and inefficiencies of power conversion, let alone the ~20-30% efficiency of power generation if mounted on a vehicle. DARPA’s 2003-2006 SHEDS program was aiming at 80% “wallplug efficiency” but seems to have fallen short of that goal. It’s not surprising, as you are pumping large amounts of current through a small piece of semiconductor that gets quite hot.

      That said, I am not aware that anyone is proposing to build a high-power laser weapon by incoherently combining laser diode beams. This would only work at short range. It isn’t just “harder” at long range. It’s fundamentally impossible. Diode laser coherence is pretty poor to begin with, and combining them incoherently gives you (spatially and temporally) incoherent light.

      Which is why the leading candidates for high-power solid state laser weapons are diode-pumped slab and fiber lasers.

      Fiber lasers are rugged but limited by the amount of power you can put through a single narrow fiber. IPG Photonics is the industry leader; their single mode fiber lasers at 1070-1080 are limited to about 1 kW. They offer lasers up to 50 kW but these are multimode. Coherent combination of single mode fiber laser beams is theoretically possible but has not been demonstrated.

      IPG advertises system “wallplug efficiency” at “>30%” but I’m not sure if that includes the cooling. It isn’t going to get much higher than that due to the inherent inefficiency of optical pumping.

      Slab lasers can have higher power and better beam quality but at the cost of large optics which must be aligned, and somewhat lower overall efficiency. According to this 2012 NRL presentation, NorthGrumm’s JHPSSL diode-pumped Nd:YAG slab laser puts out 100 kW with 20% wallplug efficiency.

      And remember, that’s assuming you’ve got a wall to plug it into. On a vehicle, you are looking at something around 5% overall efficiency, i.e. you have to dissipate around 20 times as much heat as your target. This only makes sense if you are big and powerful, and your target is small and vulnerable.

      Which is sometimes the case. But that doesn’t change the fact that guns and missiles are usually smaller, simpler, cheaper, tougher, more lethal, and can traverse things like fog, rain, smoke, sea spray, and long distances.

  25. Ara Barsamian (History)

    It can’t be done!!! I heard that one before. That’s what folks said in the 1920’s, and 30’s until Hahn and Strassmann discovered the fission of Uranium…Then everybody and their mother in law came up with fission explosive designs within the same year (French patent in 1939! albeit with the wrong isotope of Uranium).

    The Germans also thought enriching isotopes was a phantasy, until they heard about Hiroshima…

    If we can motivate scientists to work on a problem, we can solve it, but today people are “shamed into” not working on weapons; instead we have LANL and LLNL and Sandia work on “green” projects at the tune of hundreds of millions to ward off criticism for weapons work…totally ridiculous…Yeah, “peace in our time” cost tens of millions of lives during WW II, and I, for one, don’t care to repeat that.

    SDI showed many promising paths, but we gave up the research work too quickly under pacifist pressure, and the crumbling of USSR was the final nail in the coffin.

    But there is no “last word” in science, and many of the issue raised are legitimate but also solvable with enough creative brainpower if there is a will to do so…

    • Gregory Matteson (History)

      Ara Barsamian

      Correction of a historical error and injustice you may be perpetuating unintentionally: The correct description of nuclear fission was discovered by Lise Meitner, with the assistance of Otto Frisch, and provided to Hahn, who did not give due credit for whatever reasons.

      The claim that the Germans discounted the possibility of isotope separation is novel to me. The reason the Nazis went down a rabbit hole in their A-bomb project is well documented; an unaccountable error in simple arithmetic by Werner Heisenberg, who should have been able to get the numbers right in his sleep, led the Germans to believe it would require unobtainable quantities of fissionables to build a bomb.

      Careful checking of information is important, especially when the stakes are high.

      A couple of points regarding the nature of scientific progress; first, it often gets forgotten that fission itself was discovered by civilian scientists working on ‘pure’ science, with no military goal in sight, as were many other technological advances that we think of as WWII military advances.

      Second; last I heard DOE claimed the work of the National Laboratories on green projects had actually turned a profit and they were struggling to advance their progress in ways that would not cause ideological wrath to fall on them over competition with businesses.

    • j_kies (History)

      Ara
      As to these promising technologies that the SDI abandoned, can I ask for your list? I trust you aren’t talking to Neutral Particle Beams, charged particle beams or anything of such ilk. No shame accrues to pursuing effective military technologies; marketing B.S. that claimed effectiveness one viewfoil deep was much more the stock in trade. SDI ‘opened the aperture’ to consider many possibilities; their success was in trimming back the non-physical and impractical concepts. My annoyance with the SDI period was the establishment of a giant ‘stovepipe’ on something we hope to not use in earnest thus making the problems far less tractable.

  26. ara barsamian (History)

    Greg:

    Regarding accuracy…

    I said discovery of fission, not description of the process…The description of the process after the fact of discovery was by both Frisch and Meitner after the actual Hahn and Strassmann experiments (see Nature, 1939)

    You’re also confusing Heisenberg’s erroneous calculations of critical mass with the then incredible and fantastic “isotope separation” for enrichment. That’s why the Germans eventually chose the element 94 (Plutonium) path.

    Finally, the BS about “green” paying for itself is a myth…from the rip-off of the US taxpayers with the Renewable Fuels Standards “quotas” that each of us pays at the gas pump in a crashed economy, with the idiocy that EPA required the use of non-existent Cellulosic Ethanol as part of the quota for which the US refiners and blenders had to pay hundreds of millions of dollars in “fines” which the court ordered the EPA to refund. And what does the EPA do? They triple the quota for 2013 for obligatory use of the non-existent fuel. Yeah, the “green” pays for itself with ordinary workers paying for it…like Obama’s “green” jobs (replacing an incandescent bulb with a CFL at minimum wage), yeah, you can raise a family on that…

    • Gregory Matteson (History)

      Hahn will always be under a cloud for his omission. Lisa Meitner was Hahn’s assistant. A 1939 Nature submission simply ignores the subsequent zusammenbruch, http://www.atomicarchive.com/Bios/Meitner.shtml

      Heisenberg is also under a cloud, http://www.ucpress.edu/book.php?isbn=9780520229266 No confusion on my part, Heisenberg’s calculational errors greatly sidetracked and delayed the Nazi bomb, the Nazi Plutonium effort was classic Nazi fail, too little, too late.

      So it goes with totalitarian regimes. Kim Jong Un should take note.

      We were talking ‘National Laboratory’ efforts being diverted, and suddenly there is a rant about the EPA?