Jeffrey LewisSanger and Shanker Confusion

They don’t look anything alike.

Well, this is irritating. David Sanger and Thom Shanker have published a really interesting article that is irredeemably fouled up by their profound confusion about North Korean missiles.

The lede of the article is that “North Korea is moving mobile missile launchers around the country, some carrying a new generation of powerful rocket…” What I presume this means is that the US IC has satellite images of shiny new TELs at missile units around North Korea, which is usually a sign that new missiles are being deployed.

What kind of new missile, however, is impossible to fathom because Sanger and Shanker conflate two different missiles — North Korea’s KN-08 ICBM and the Musudan IRBM (~3,000 km range). What mess!

Here are the important parts of the article:

The discovery by American intelligence agencies that North Korea is moving mobile missile launchers around the country, some carrying a new generation of powerful rocket, has spurred new assessments of the intentions of the country’s young new leader, Kim Jong-un, who has talked about economic change but appears to be accelerating the country’s ability to attack American allies or forces in Asia, and ultimately to strike across the Pacific.

The new mobile missile, called the KN-08, has not yet been operationally deployed, and American officials say it may not be ready for some time. But the discovery that the mobile units have already been dispersed around the country, where they can be easily hidden, has prompted the White House, the Pentagon and intelligence agencies to reassess whether North Korea’s missile capabilities are improving at a pace that poses a new challenge to American defenses.


The more immediate mystery for the administration, however, is what North Korea may intend with the intermediate-range KN-08, which was first shown off by the North in a military parade last April. At the time, many analysts dismissed it as a mock-up. In fact, it has never been test-flown. But parts, including the rocket motors, have been tested separately, according to officials familiar with the intelligence reports, who described the missile developments on the condition of anonymity because they were not authorized to discuss the assessments.

Officials familiar with North Korean missile technology say the KN-08 weapon is designed with a range capable of striking South Korea, Japan and parts of Southeast Asia — although with uncertain accuracy.

Let’s tease these statements apart.

The KN-08 ICBM.  In April 2012, North Korea paraded what appeared to be a multi-stage ICBM through Kim Il Sung Square. Secretary Gates had gone out of his way in 2011 to mention such a missile, directly calling it a road-mobile ICBM. It was this missile that was carried by Chinese TELs, exported in apparent violation of Security Council sanctions.  And it was this missile that was the subject of a debate on this blog about whether it was a fake or not.

This is as good a time as any to plug three papers — an assessment of the KN-08 prepared by reader John Schilling, Nick Hansen’s contribution at 38North and Markus Schiller’s new RAND monograph, Characterizing the North Korean Nuclear Missile Threat.

The Musudan IRBM.  Two years prior, in 2010, North Korea paraded a new intermediate-range mobile missile through Kim Il Sung Square.  The so-called Musudan just happens to be “intermediate range” and “capable of striking South Korea, Japan and parts of Southeast Asia.” (NASIC described it as a one-stage IRBM with a 2,000 mile or ~3,200 km range, which perfectly corresponds to ” parts of Southeast Asia.” A Wikileaked US cable to MTCR states put the range at a bit more with a smaller payload — 4,000 km with a 500 kg payload.) Markus Schiller and Robert Schmucker have their doubts about this one too, but I am rather more convinced.

The two missiles have some things in common — neither has been flight-tested — but, as Sanger and Shanker were told, there have long been reports of ground-testing for Musudan (SS-N-6 type) engines that might power one or both missiles. (I wrote a very long post on the Musudan, including claims of static engine testing, in a post titled, Origins of the Musudan IRBM.)

I suppose the simplest explanation is that North Korea has deployed the Musudan IRBM, not the KN-08, but that Sanger and Shanker simply confused the two missiles.  This is such a simple story, though, that it is hard to imagine Sanger would screw it up this badly.  And South Korean officials have long asserted that the Musudan is deployed, even though the US IC has been more cautious.  How the KN-08 got into all this, I cannot fathom.

On the other hand, perhaps it is KN-08 units that have been seen in the field.  In this case, Sanger and Shanker simply misstated the range of the missile. That requires the least rewrititing of the original text, but it requires Sanger and Shanker to bury the lede — RED KOREAN MOBILE MISSILE CAN STRIKE U.S.  I can’t imagine they would let that one pass unremarked.

Finally, it’s possible that the IC has concluded the KN-08 range is simply not much better than the Musudan and isn’t an ICBM at all.  That also requires burying the lede: U.S. REDUCES ESTIMATES OF NORTH KOREAN MISSILE.

Really, it’s hard to tell.  I bet there is a story in there somewhere, but I can’t fathom what it is.


  1. George William Herbert (History)

    They do look reasonably alike, if you Camoflage the Musudan and are at a view angle where you can’t count the road wheels on the TEL and the aspect ratio of the launcher is ambiguous, and if you don’t have enough resolution to see the smaller diameter second stage clearly.

    I would hope the US government has better resolution and can get better view angles than that.

    • Jeffrey (History)

      Me too.

  2. Andy (History)

    Yeah, very weird. I suppose for completeness sake we should consider the possibility the anonymous “officials” don’t know what they are talking about and provided incorrect information.

    • Jeffrey (History)

      Or after four beers its just the Musudong-o-8.

  3. Jonathan McDowell (History)

    Another interesting thing in the article is that Panetta and other officials are still apparently describing last month’s Unha-3 launch as a ‘missile test… that reached the Philippines’ rather than, say, ‘a satellite launch that advanced NK missile technology’. Purely spin, or a certain amount of denial as well??

    • Jeffrey (History)

      The framing as a missile test is about maintaining pressure on North Korean for violating UNSC sanctions. I suspect the Obama Administration is afraid that were the launch to be described as a space launch, it would be harder to secure another round of sanctions.

    • Magpie (History)

      Shouldn’t it be a “missile test that reached *everywhere*” then? I mean, it went into polar orbit.

      I’ve got to get into lazy journalism. Every slow-news-day hack on earth shoulda submitted a “NK launches missle that could hit us!” story. Sure, it’s a pretty flexible value of “could”, but hey…

    • Jeffrey (History)

      Let’s not encourage the “slow-news-day-hacks.” Getting them to properly explain the difference between different ballistic reentries seems unlikely.

  4. simorgh (History)

    Another issue that struck me while reading a lot of articles about the NK missile programm is the cited lack of warhead miniaturization.
    Of course there is no proof they managed it, but there is no proof to the contrary neither. We only know, that Pakistan has a nuclear equipped Nodong in its arsenal, that plans of a Chinese warhead reached Libya, which has had less extensive cooperation with Pakistan than NK, that the Norks had pretty much 20 years to think about miniaturization and that NK now has an advanced uranium enrichment programm that provides it with HEU useful for copying an early Pakistani or Chinese warhead.

    So what exactly makes analysts so sure they do not have a miniaturized warhead yet? In my opinion there is simply no evidence and too few indications to stress the likeliness of the best case scenario in every single article dealing with NK’s missiles and nukes.

    • Jeffrey (History)

      They’ve also conducted two nuclear tests.

    • George William Herbert (History)

      A lot of people tent to assert and act like all nuclear bomb programs would have to start out with very large bombs.

      The key technical lessons for miniaturization are now open, though a lot of details are not fully public. Those of us who dabble in the field don’t want to help NK or Iran if we can avoid it.

      The technology and techniques to shrink from 54 inch implosion systems to 22 inch systems is obvious. They were known and documented in the Manhattan Project, just not implemented until the early 1950s.

      The technique to shrink from there to say 12 inch diameter is in Wikipedia, though Howard Morland got a couple of subtle details wrong I am not correcting…

      There are nuclear material behavior details that one would need to test fire devices to confirm, once you get that small. But that’s different than it being impractical to jump way down the miniaturization curve for a modern program…

    • Jeffrey (History)

      The fact that the DPRK probably attempted a miniaturized device in 2006 and failed is not unrelated to the fact that they got a bigger yield the next time around which is not unrelated to my suspicion that the third time will be a charm — even if it is an HEU device.

  5. John Schilling (History)

    Looking back over the analysis I put together last year, I did suggest that the earliest possible deployment of the KN-08 would be right about now, but that additional flight testing would be required beforehand. The DPRK has never been big on rigorous flight test programs; if they consider the recent Unha-3 success to have validated the KN-08 design as well, it is possible that we are seeing initial deployment of the latter missile.

    If so, they are likely to be unpleasantly surprised the first time they try to launch one of their not-an-Unha ICBMs. And it couldn’t happen to a more deserving group of people.

    But we should be able to pin this down with even half-decent satellite imagery. Commercial one-meter resolution should suffice to get the overall length of the TELs and probably to count the wheels.

    • George William Herbert (History)

      On the technical front… I’m away from a decent computer until later tonight, but… Is there any credible KN-08 analysis which could have as bad a range as down to IRBM threshold?

      I couldn’t see one BOTE but that’s just BOTE.

      The only thing I can think of would be a masrikova with the parade missiles and the real one only being two stage, or something like that… Again, BOTE not an engineering analysis…

    • John Schilling (History)

      Looking at my original analysis, if you assume Scud/Nodong technology for the first two stages, and a nominal (IMO needlessly heavy) 1000 kg warhead, the range will come in just shy of the traditional 5500 km threshold for an ICBM. Even then, one could push it well into ICBM territory with a reduced payload, but 500 kg for a complete warhead may be beyond North Korea’s abilities.

      The third stage I think has to use R-27/SS-N-6 technology because the stage is too short to fit a Scud engine bay and reasonable propellant tanks, and the performance gain from using an R-27 engine for the second stage as well is substantial, so that is my best guess as to how an operational KN-08 would be constructed, but a different analyst might reasonably come to different conclusions and call it a high-performance IRBM.

      All of which reminds me I need to redo that analysis to incorporate what we have learned from the latest Unha launch. At this point, it looks like we have seen all we are going to see of the wreckage.

    • Jeffrey (History)

      In 1999, DIA assessed that “An early-generation warhead weighing 650-750 kg is the best Pyongyang could achieve.”

      In the mid-2000s, a defector claimed the DPRK North Korea “built a one-tonne nuclear weapon with 4 kg of plutonium”

      My best guess is that the DPRK tested something like the Mark 7 in 2006.

      I hate linking to things I wrote six years ago.

    • George William Herbert (History)

      Jeffrey writes:
      My best guess is that the DPRK tested something like the Mark 7 in 2006.

      The question is, why on earth would someone test something like the Mark 7 in modern times?

      To go from Mark 1 to Mark 7 (Thor) you do:
      1. 32 -> 92 point initiation
      2. probably non-newtonian explosive lenses
      3. probably levitated pit
      4. probably system without wave-shaper layer (explosives -> tamper/reflector -> pit directly)

      To go from Mark 7 to Mark 12 (Brok) you do:
      5. DU to Be (tamper)/reflector
      3.a. reasonably assuredly a levitated pit

      The idea of using a Be reflector and the implosion system energy inputs requirements reductions that implies were well documented in public long long before the first NK test.

      During the same timeframe, their ally Iran was allegedly working up towards the alleged R-265, which is far far more compact than a Mark 7, though based on an entirely different implosion initiation system and is a HEU device.

      The idea that by the early 2000s the best North Korea could do was a Thor pretty much argues that there’s no way North Korea’s entire nuclear weapons program is a fraction of as smart as the public, open source nuclear weapons engineering community. While it’s possible that the handful of us are in fact individually and collectively better than whole national weapons programs in some countries, it seems difficult to credit absent better technical detail. I do not assume I am smarter than the people doing this full time in NK or Iran. This is near the tippy-top of their national priority levels, and presumably has their best and brightest. I am smart, but better than one in a million smart?…

      R-265 was surprisingly simple but in an elegant way, and one that matches their delivery vehicles. A Thor sized NK weapon does not.

  6. George William Herbert (History)

    We can get one Unha tank mass ratio reasonably well, yes. If they used that tech for the first two stages. They’re smaller but could be the same structural concept / designers / mass per unit volume / etc.

  7. Markus Schiller (History)

    I see some flaws in the analysis approaches here, but I could be wrong.

    1. I never heard of Pakistan testing a nuclear equipped Ghauri. I am sure that they want to have one, but we do not know if they have one that actually works. Just because “everybody knows” doesn’t mean that it is true.
    2. Plans reaching anyone are pointless. It is not about some secret plans or designs.
    3. From what I heard, Sig Hecker was shown a large room with a collection of metal cylinders that looked like centrifuges. If that is your only indication for an “advanced uranium enrichment program”, I would be careful to base all the further analysis on that. Every Home Depot market could create the same show over night. Looking back in history, it sometimes turned out that the other side pretended to be more advanced than it actually was.

    But the most profound mistake everyone seems to keep making: If someone wants to have something, it is assumed that they just design and build it. According to this view, the North Koreans go straight for the state-of-the-art miniaturized nuke design. And they use SS-N-6 (or more advanced) tech because Scud technology has lower performance. And they design and build missiles from scratch right into deployment because they don’t want to do these expensive and annoying test flights.

    That it took them 20 years to successfully get the Unha off the pad – whatever. That the Unha uses Scud tech – no matter. There seems to be no contradiction to them quickly building and deploying several road-mobile ICBMs with advanced SS-N-6 tech in parallel.

    Again: It is not some secret formula, or plans for a device, that enable access to a technology. The theory of jet engines is well known and understood, and taught at universities all over the world. Everybody knows how to build jet engines, everyone has access to both knowledge and the final product. So why is China still importing Russian engines for their fighter jets? Because knowing how something works and actually successfully doing it are two very different things. Knowing how Golf should be played doesn’t win you the Masters. And knowing how a Porsche works, you will have a hard time building one. You start with a simple design that works, and slowly move toward refined and more complex versions. This is true for cars, trucks, trains, airliners, fighter jets, nuclear power plants, aircraft carriers, milling machines, printing presses, gearboxes, and every other machinery – yes, even for nuclear weapons, warheads, and rockets.

    By the way, a conventional KN-08 design solely based on Scud technology cannot offer ICBM range. Simple rocketry: The KN-08 launch weight is about one third that of the Unha. Therefore, with access to the same tech base, the range has to be much lower. Sanger and Shaker might know that and could therefore have labeled the KN-08 as an IRBM, preserving the ICBM name tag for the Unha. In that case, the article makes sense.

    • rwendland (History)

      Marcus, I agree with most of that, and was impressed by your NK Missiles paper.

      But a problem with the Home Depot pretend/exaggerated enrichment plant possibility, is that building the pretend 25MWe experimental LWR next door, under satellite observation, would be a whole lot more expensive – implausibly so I’d say. They must be very confident of the enrichment working to have gone so far with the LWR.

      One interesting take-away from Sig Hecker’s visit report, is that the enrichment plant (and I guess LWR) is being largely implemented by a much younger generation of engineers than the Pu/GCR guys. Probably engineers well versed in modern Chinese hi-tech, possibly educated there. That could explain some rapid modernisation, but possibly some unexpected failures in the future as well.

    • George William Herbert (History)

      Marcus writes:
      You start with a simple design that works, and slowly move toward refined and more complex versions. This is true for cars, trucks, trains, airliners, fighter jets, nuclear power plants, aircraft carriers, milling machines, printing presses, gearboxes, and every other machinery – yes, even for nuclear weapons, warheads, and rockets.

      You do NOT get into rocketry cloning V2s, nor cars with Model Ts.

      The suggestion that someone would start simpler and more robust is almost certainly true. But mistaking the early designs or design concepts for simpler or more robust is a mistake.

      Modern weapons concepts may have tighter tolerances, but they have many fewer parts and are inherently simpler and more robust.

    • Magpie (History)

      So there’s a good question: if you were starting a nuclear weapons program, what sort of design would you start with?

      Obviously there will be a balance between simplicity and effectiveness, resources and outcome – but if I can put in an extra 10% of effort to get a MK12, which gives me an erection, over the effort to get a Mk1, which is of no practical use to me, which do I build? Of course that hand-wavy metric makes no sense – but seriously, what would you shoot for, if you were North Korea?

    • George William Herbert (History)

      Magpie writes:
      Of course that hand-wavy metric makes no sense – but seriously, what would you shoot for, if you were North Korea?

      That is an excellent question I will not answer in public, for reasons I hope are obvious.

    • Magpie (History)

      …you could lie…

      Fair enough.

  8. John Schilling (History)

    “Everybody knows how to build jet engines, everyone has access to both knowledge and the final product. So why is China still importing Russian engines for their fighter jets?”

    Because a 1960s-vintage fighter jet is pretty much useless in the modern world, where it would be met by an adversary’s much more advanced and capable aircraft. Only the most technically advanced industrial nations can build *modern* jet engines; it does not follow that smaller and less developed nations could not build *any* jet engines if they had some perverse reason to do so.

    Unlike jet fighters, even a 1960s ICBM would be of some use in the 21st century, because it would be fielded not against enemy missiles but against enemy cities, which have improved only slightly in their ICBM-resistant properties over the past fifty years.

    “According to this view, the North Koreans go straight for the state-of-the-art miniaturized nuke design”

    Which view is this? The view expressed by just about everyone here, is that the North Koreans would probably go straight to sixty-year-old pure-fission weapons an order of magnitude or so heavier than lightweigh modern boosted-fission primaries. This is, to be sure, an “optimistic” assessment; much of the national security community believes the North Koreans would have to start with even heavier seventy-year-old designs.

    “It is not some secret formula, or plans for a device, that enable access to a technology”

    Who has mentioned secret formulas or plans?

    Schiller, you are tilting at straw men. You are clearly knowledgeable enough to contribute to the discussion here, but the first requirement is to pay attention to what other people are actually saying.

  9. George William Herbert (History)

    Whoops. CNN reporting the NK are threatening another nuclear test imminently.

    • AnonNL (History)

      When the Norks conduct another nuclear test, how will we know whose weapon they are testing? One can differentiate Pu from HEU weapons by measuring the airborne by-products. So if it’s Pu, it’s a Nork weapon. But if it’s HEU, is it Nork, or is it Iranian? Are there ways to determine the origin of the HEU based on differences of concentrations of by-products, or are such differences too small to be useful? I’m willing to assume that the West knows the exact trace-element contaminants in both unenriched and 20% enriched Iranian U & can therefore predict these contaminants in Iranian HEU, but I wouldn’t make that assumption for Nork U.

  10. Magpie (History)

    No reason to think Iran has any HEU, but NK does. If it is an HEU test, though, then I reckon the scariest possibility would be a test of the fabled Iranian R-265 technology loaded with NK HEU, which would seem a logical point of cooperation between the two countries – especially if some Iranians are considering a run to breakout in the near future.