October 23, 2022

In Defense of Missile Defense

Probably the most curious gap in the current debate around defense issues is the subject of missile defense. Despite the possibility to meaningfully reduce the danger from nuclear war, it is largely ignored, and when it does come up it is widely misunderstood. I've referenced this before, but figured it was time to lay out the argument in more detail.


USS Shiloh fires an SM-3

The basic case for missile defense is quite simple: nuclear weapons aren't something you want detonating in your country, and shooting them down seems like a good idea. But as appealing as this is, it is usually countered by a pair of counterarguments, that it's far too hard to shoot down all incoming missiles, and that it would be destabilizing if it was possible. But this relies on the basic premise that any missile defense system which can't shoot down all missiles is useless, and there's no reason that would be true unless someone was smuggling it in to defeat missile defense for other reasons. (We'll come back to that in a bit.) A system which stops 50% of incoming missiles means that only half as many people will die in a nuclear war,1 obviously a desirable result.


GBI, the interceptor for GMD

The difficulty of shooting down an incoming missile is broadly overstated. We routinely play sports (baseball, cricket, skeet shooting) where an unaided human has to intercept a ballistic target, and while those humans don't have a perfect success rate, they also don't have radar, computers, or multiple chances. More practically, the US has invested a fair bit in this field over the last 30 years, and those efforts have paid off, leaving us with an array of anti-missile systems. The big three are GMD, which is designed to protect the continental US from ICBMs, Aegis BMD, a ship-based system intended for use against theater missiles that has gained anti-ICBM capability in recent years, and THAAD, a mobile ground-based system intended to counter shorter-range missiles. All of these have performed well in testing, getting over 50% of hits, sometimes well over, with failures having far more to do with hardware issues than with the inherent difficulty of hitting an incoming warhead. Obviously, these are one-off tests in ideal conditions, and there's at least a chance that the testing setup obscures some significant systemic issue that will render the whole thing useless when missiles actually start flying. I can only point out that the defense industry has generally proved to be pretty good at building robust systems given enough time and money, although it usually takes more of both than anyone would like.


THAAD is launched during an exercise

The other thing usually brought up as a counter to missile defense is the potential use of decoys, flooding the system with targets at low cost. In practice, this seems to be far more difficult than it first appears, although details are elusive and the best source of information is what missile designers actually did to try to penetrate missile defenses. The basic problem with decoys is that they only work so long as they are indistinguishable from the warhead, and thanks to modern computers and sensors, this is much harder than it used to be. Even back in the 60s and 70s, there were serious concerns about the effectiveness of decoys, prompting the designers of Poseidon to decide that the best decoy would be the same size and weight as an actual warhead, in which case they might as well put the nuclear payload in as well. The British Chevaline decoy system was a major technical challenge in the late 70s, and it seems doubtful that the techniques used then would still be effective. The last half-century has seen very little in the way of better decoy technology, and massive improvements in discrimination.


Chevaline

But the uncertainty of any missile defense system brings us naturally to the second prong of the argument of missile defense, that it is inherently destabilizing. The theory goes that a 100% impenetrable missile defense would undermine deterrence, allowing the nation that owned it to strike with impunity, and thus prompting the opponent to launch before the system is ready. While this is true as far as it goes, we should instead examine what happens in the real world, where any missile defense system is necessarily going to be imperfect. In that case, it's far from obvious that missile defense is destabilizing. Even if we assume that the US builds a system capable of shooting down 99% of all incoming missiles, that's still going to mean that China can land 3 or so warheads on US territory, and Russia probably 10-15. In practice, no leader this side of Stalin is likely to court those kind of losses unless they really need to, no matter how much they stand to gain from it.2 People are remarkably scope-insensitive when it comes to nuclear weapons, with North Korea deterring the US with 6 warheads (one for every 7 American interceptors) nearly as effectively as Russia does with 1550. A strong ABM shield would also provide a great deal more security against accidental launches. A few ICBMs inbound for whatever reason are less concerning when most or all of them can be dealt with before impact, giving decision-makers more time to respond and reducing the risk of escalation.


The Nike family of missiles, with Nike-Zeus on the left

But if the case for ABM is fairly straightforward, why is the case against it so common in discussions? This goes back to that root of all kinds of evil, Robert McNamara and the Kennedy/Johnson Administration. When they entered office, they inherited the Nike-Zeus ABM system, which was almost ready for deployment. McNamara didn't think it was a particularly cost-effective system, but chose to justify his cancellation on the basis of general ineffectiveness instead of poor value for money. This critique was later picked up in the 80s by critics of Reagan's Strategic Defense Initiative, whose opposition often seemed to flow more from political opposition to Reagan than from a technical evaluation of the merits of the system.3


The Aegis Ashore facility in Romania

But the obvious conclusion of all of this is that the US is in a position where we can easily build a missile defense shield, the biggest problem being the political will to buy the weapons in sufficient quantities. Procurement has lagged for years, with only 44 interceptors for the GMD system currently installed to protect the US mainland and no more planned. The situation is slightly better for theater missile defense, with the Navy having about 265 SM-3s and a few hundred more THAADs for the Army. The obvious solution, underscored by recent events in Ukraine, is to expand our missile defense net. The lowest-risk way to do is to buy more SM-3s for the Navy, as SM-3 has proven the most effective of the systems and the line is still hot. The SM-3 Block IIA's capability against ICBMs should also make us take a close look at installing Aegis Ashore near our major cities and military bases. In addition to providing proven ABM protection with minimal development costs, this would also give the areas around the installations a world-class air defense system, capable of dealing with atmospheric threats, which American cities have had no reliable protection against since the Nike system was shut down in the 1970s. As the GMD interceptors were rather rushed into service, it probably makes sense to wait on expanding that system until the new interceptor currently in development is ready.

Nor should the US limit this system to its own use. We've already partnered with Japan on Aegis BMD, installed Aegis Ashore in Romania and Poland and sold THAAD to the Saudis and the UAE, who used it to shoot down a missile in early 2022. Ensuring that these systems are widely distributed will not only allow us to spread the costs across a larger buy but also significantly decrease the impact of any potential nuclear conflict. The ballistic missile is a threat that can be mitigated, and for both strategic and humanitarian reasons, it should be mitigated.


1 Assuming, of course, that the number of warheads remains constant. This was probably the best argument against missile defense during the Cold War, when it was at least theoretically possible that the Soviets could double their arsenal in response to the threat. It's a lot less feasible today, given arms control treaties and the shaky state of the Russian economy. It's plausible that China could pull it off, but they've long taken a rather more minimal approach to deterrence than Russia has, which means a reasonable missile defense system probably wouldn't worry them too much. Also worth noting that yes, it's possible to overcome a 50% effective system by shooting two warheads at each target, but the number of targets is cut by 50% either way.

2 To put this another way, imagine the American President is presented with a button to launch a first strike on another nuclear power. If they do, the other power will be devastated, and between the strike and American missile defenses, only one warhead will reach an American city. Stalin would absolutely take that option, and Putin might well, but no US president's hold on power is firm enough for that to work unless they do it to preempt something worse for the US. And there's no way they'll have the absolute knowledge that only one warhead will get through.

3 This shouldn't be read as a strong defense of the technical merits of SDI as laid out in the 80s. It was definitely compromised by the need to work around the ABM treaty, and a great deal of what was proposed was intended to send the Soviets on a wild goose chase, spending money all the way. It's not entirely clear what falls into that category, and what was serious. Reagan also seemed to push the "getting rid of nukes" position pretty hard, which I am less sure is a good idea.

Comments

  1. October 23, 2022John M said...

    I could almost here you boo when you wrote Robert McNamara. You were too kind to him.

  2. October 23, 2022bean said...

    Yes, but I try to be careful in only making claims about him that I can back up with sources.

  3. October 23, 2022Humphrey Appleby said...

    How soon and how far could said capabilities be expanded? Are there any ongoing efforts to do so?

  4. October 23, 2022bean said...

    There's talk of Aegis Ashore in Hawaii and Guam, but I don't think either have been funded. Don't know exactly what timelines would be like, but I'd guess 3-5 years.

  5. October 23, 2022ike said...

    It’s plausible that China could pull it off, but they’ve long taken a rather more minimal approach to deterrence than Russia has, which means a reasonable missile defense system probably wouldn’t worry them too much.

    HUH?

    Maybe I am missing something, but if the Bean Defense Initiative is big enough to make a dent in the Russian Broadside. Surely it would work even better against a much smaller adversary.

  6. October 23, 2022Humphrey Appleby said...

    At a time when we seem determined to play nuclear chicken with Russia over Ukraine (and plausibly soon with China over Taiwan) why on Earth is it not a crash priority to immediately build Aegis ashore near every major US city?

  7. October 23, 2022Humphrey Appleby said...

    Also, those 42 American interceptors you reference - is that 42 platforms each with multiple shots (so, defenses exist at 42 locations)? Or is that 42 total shots, so that we could intercept at best 42 launches, if every single shot was a hit?

    And could the answer to my oft-repeated question above be that our missile defense planners are relying on the power of narrativium, and on 42 being the answer to life, the universe, and everything (the last category presumably including missile defense?)

  8. October 23, 2022echo said...

    Any idea what sort of coverage SM-3s have? I'm confused how "range" works for midcourse interceptors rather than terminal ones.
    You make it sound like Aegis ashore would be regional defense rather than being able to hit any missile that arcs over its general location.

  9. October 23, 2022David W said...

    Suppose we're cost-constrained, and need to give something up to afford BMD? What would you propose?

    Personally I'd lean toward getting rid of something offensive. It's a lot harder for someone to make the argument they have to hit us first if we're obviously drawing down the forces that we'd use to hit them. Maybe halve the Marine Corps and remove three carriers from the fleet, along with the B-2's and B-52's?

  10. October 23, 2022Chipsa said...

    Range for the SM-3 is approximately a tear drop shaped area such that a ship of the west or east coasts could each cover about 50% of CONUS. Some of that would be overlapping, so some of it also wouldn't be covered (southern Texas). Thus, we wouldn't actually need many Aegis ashore sites, probably just 3 in CONUS.

    The 42 interceptors are total, so if every one of them were launched and hit, we could get at most 42 enemy warheads.

    If we had to trim to pay for the Bean Defense Initiative, I'd suggest cutting active duty Army.

    Something worth talking about at the lower levels of missile defense is virtual attrition: if the offense doesn't know which warheads are going to make it to which target, they need to put more warheads on the targets they consider more important to make sure those get destroyed. Thus, less important targets won't get hit at all, and some of the more important ones will get hit multiple times, effectively wasting the warheads.

  11. October 23, 2022Emilio said...

    Regarding the Nike family, a Nike Hercules base in northern Italy has been transformed in a museum, for those interested.

    https://www.basetuono.it/en/

  12. October 23, 2022Alexander said...

    I'm still a bit ignorant about how you distinguish a warhead from a decoy during the midcourse phase. Are they still in the (very) upper atmosphere, so the lighter decoys are slowed by drag? If not, what other means are there to tell them apart?

    Because of this, as well as the greater expense of the larger midcourse interceptors, aren't terminal phase interceptors a better bet? Especially as they more easily double up as defence against shorter range ballistic missiles, and even aircraft. They are also a more proven technology. I suppose the US is so much larger than the UK, and with a less concentrated population, that it might be less practical to have enough Aegis Ashore installations to cover everywhere.

    Either way, I agree that they are worth investing in, especially since it looks like fewer resources need to bet spent on deterring Russia from starting a conventional conflict with NATO.

  13. October 23, 2022Humphrey Appleby said...

    48 total interceptors? That seems pitifully small. Maybe it gives us a chance against a crazy North Korean dictator (or equivalent), but against Russia (or even the PRC) it rounds to `we have no ballistic missile defense.'

    Why so few interceptors?

  14. October 23, 2022Jade Nekotenshi said...

    Because as I understand it, the kooky tin-pot dictator with a handful of nukes is exactly what we designed the system against.

  15. October 23, 2022Jade Nekotenshi said...

    Because as I understand it, the kooky tin-pot dictator with a handful of nukes is exactly what we designed the system against.

  16. October 23, 2022Humphrey Appleby said...

    So what would it cost to build a couple of hundred times that number of interceptors, and actually have a chance at stopping the Russian (say) broadside?

  17. October 24, 2022Echo said...

    Thanks Chipsa.
    Last part of your post made me wonder: do interceptors multiply the value of decoy sites against a counterforce first strike?

    1) You have 1 ICBM base, he has 4 missiles. He has 3 missiles free for railyards and ports.

    2) Same but you have a decoy ICBM base which soaks up 1 missile, leaving 2 free. Not a great use of money, even if you make both cheap "dense pack" affairs.

    3) You have 1 ICBM base and interceptors. He throws all 4 missiles at the high priority target. You need 4-12 interceptors to defend.

    4) You have 1 ICBM base and the decoy site. He throws 2 missiles at each, you defend only the real one. The decoy soaks up two missiles, and you only need 2-6 interceptors. Some cows in Montana are briefly inconvenienced.

    Seems like decoys would be better multipliers the lower your interception chance is, and the worse his sat recon capabilities are. So back in the 1960s.
    It also wouldn't do anything against a retaliatory counter-value strike by NK or China.

  18. October 24, 2022Echo said...

    Re. Alex's comment, do midcourse interceptors hit a bus before the warheads separate? That would be a huge multiplier on effectiveness vs terminal defense.

  19. October 24, 2022John Schilling said...

    @Echo: Ground-based midcourse interceptors will not hit the bus before the warheads (and decoys) separate. A system with space-based interceptors might do so, depending on where the orbital platform is, but it still isn't likely. Warhead/decoy deployment occurs as quickly as possible, for several reasons - you can spread the warheads across a larger target set, you can deploy warheads and decoys while still out of view of enemy radar, and of course you don't give the enemy a twofer or a tenfer as a target.

  20. October 24, 2022chipsa said...

    @Echo & @John Schilling: Mid-course interceptors may hit the bus before the warheads separate. There's a balance on when debussing occurs, as the warheads don't have the ability to maneuver afterward. So, the earlier you do it, the greater cross range you have for them, but the poorer accuracy you have. And the reason why the warheads can't maneuver is that the stuff you need for that is expensive and heavy, as it's effectively most of a warhead bus that you're adding on. For MMIII or LGM-35, where only one warhead is actually fitted, this may be moot, but Russian ICBMs are nominally multi-warhead, and SLBMs are all multi-warhead. But part of the point of the GBIs is to force them to debus before the interceptor can just kill the bus, which creates more inaccuracies, and makes it more likely they'll miss, which brings back to: have to toss more warheads at a target to be sure to kill it.

    Also, your decoys will have to maneuver, because if they don't, that's another part that's not the same between the two, but then if they do, then you can discriminate based on the fact that the decoy is lighter (because if it isn't, you're wondering why not just add another warhead).

    The radar horizon is an issue, but that's part of why the DSP and SBIRS satellites exist: you can see the launch before they come in view of the radars. SBIRS Low is currently cancelled, replaced by the STSS demo, but it's probably feasible to continue to develop the tech for IR tracking of launches from LEO. Starlink may be involved in this, using the satellite bus developed from them for a new tracking layer.

    And yes, WRT decoy sites, they're probably more effective, if they're in the target plan. And one of the ways you can do the decoy site is just have more launch sites then you actually fill. This was a real idea, using railroad based, or truck mounted launcher, where the launch points are pre-plotted, but which launch point actually has a missile in it, and which just has an empty launcher shell is a mystery. However, there's security issues with just moving live ICBMs around, so they prefer silos, which end up being expensive enough that you might as well make sure your silo is filled with a missile. Additionally, since they're all currently single warhead missiles, they act as decoy sites regardless, since if you drop two warheads on a silo, then you only manage to kill one at the cost of two warheads.

  21. October 24, 2022Alexander said...

    @chipsa: Can you practically tell the mass of a warhead/decoy by observing it burn to manoeuvre? It seems like it would be much harder than a satellite launch. You can't expect much notice, there will likely be multiple near simultaneous launches, giving you probably hundreds of targets to watch, with SLBMs you don't know the launch site in advance, and you need the answer in minutes or it's useless. It sounds very challenging to me.

    The one thing I can imagine working is if the velocity of the bus changes notably when a warhead separates, it a way that lets you distinguish it from a lighter decoy separating. Even then, the bus might dispense several decoys together with the warhead along a similar trajectory.

    Does anyone (who is free to talk about it Ü!) know if atmospheric drag is significant enough in the midcourse phase to noticeably slow a decoy? I expect it would be much easier to spot even a small change in velocity via radar, and tricky to give the warheads a tiny and continuous deceleration with a rocket so they slowed along with the decoys.

  22. October 24, 2022Naomi said...

    You mention procurement has lagged, yet you fail to disclose why. Specifically, why would you spend 70-100 million on an interceptor that has failed many tests nor operationally tested?

    No, building a shield is not an easy task. It’s truly remarkable it’s effort has progressed this far, but easy? Nope!

  23. October 24, 2022Naomi said...

    You mention procurement has lagged, yet you fail to disclose why. Specifically, why would you spend 70-100 million on an interceptor that has failed many tests nor operationally tested?

    No, building a shield is not an easy task. It’s truly remarkable it’s effort has progressed this far, but easy? Nope!

  24. October 24, 2022Hugh Fisher said...

    I (Australian) suggest that maybe the US government and people are not enthusiastic about missile defense in the 21st C because there's already a proven way to reduce the number of warheads that might get through: negotiation.

    According to Wikipedia "Historical nuclear weapons stockpiles" article the largest number of atomic warheads that could have hit the USA (and allies) was in 1985 when the Soviets had 39,179; because of treaties and negotiation that is down to just 6,375 as of 2022. That's an 83% reduction.

    (Yes I'm ignoring any differences in accuracy and yield between the two sets of warheads. At this scale, I don't think it matters.)

    No doubt the collapse of the Soviet Union and decline in economic potential also has something to do with the reduction on the Russian side, so let's consider the US nuclear arsenal from the viewpoint of Russia.

    Largest number of US nuclear warheads was 27,519 in 1975, today it's 5,800. That's a 79% reduction. No missile defense system required or even proposed.

    Diplomacy isn't perfect, but so far it has a much better track record on reducing the potential for casualties in your own country.

  25. October 24, 2022Directrix Gazer said...

    "...because of treaties and negotiation that is down to just 6,375 as of 2022. That’s an 83% reduction. [...] No doubt the collapse of the Soviet Union and decline in economic potential also has something to do with the reduction on the Russian side, so let’s consider the US nuclear arsenal from the viewpoint of Russia."

    Cause of Death: Twisted Ankle and Leukemia, but mostly twisted ankle.

  26. October 24, 2022Asher ty said...

    Why not just buy or co-producethe israeli hetz system. In a short time we could have a very effective system - closer to 100% in tests - covering the entire USA. And its mobile so not as vulnerable as the aegis ashore.

  27. October 25, 2022dood said...

    But one of the main arguments against SDI was that SDI ultimately had little to do with defending against a Soviet first strike, as for all it's advances, there were simply too many warheads to be destroyed in so little a time period. Instead the main problem with a semi effective system like you describe is that it gives the US a realistic first strike capability, by attacking the Soviet missile silos, leaving only a few prepared to respond and which the US might be able to bear

  28. October 25, 2022Anonymous said...

    Alexander:

    Can you practically tell the mass of a warhead/decoy by observing it burn to manoeuvre?

    How bright is the rocket plume?

    Even if you artificially brighten the rocket plume looking at the spectra will give that away.

    Asher ty:

    Why not just buy or co-producethe israeli hetz system. In a short time we could have a very effective system - closer to 100% in tests - covering the entire USA. And its mobile so not as vulnerable as the aegis ashore.

    But also significantly less capable.

    Israel isn't defending against ICBMs but mostly unguided short range rockets which is just not a credible threat for pretty much any US territory (unless you're worried about Canada shooting rockets south of the border).

    dood:

    Instead the main problem with a semi effective system like you describe is that it gives the US a realistic first strike capability, by attacking the Soviet missile silos, leaving only a few prepared to respond and which the US might be able to bear

    Had you bothered to read the article you would have realized that a semi-effective system does nothing of the sort, only a near 100% effective system would make a first strike a viable option because even a 99% effective system still means millions dead on your side in the event of a full scale exchange.

  29. October 25, 2022bean said...

    On decoys, I think it's probably possible to tell a balloon from an actual warhead at surprisingly high altitude, because modern radar/computer systems are a lot better at detecting small changes in trajectory than they used to be, and satellites see quite a bit of influence from drag over long periods. But I'm not sure. Beyond that, no, the decoys do not have to maneuver. We're dealing with ballistic missiles.

    @Hugh

    Yes, but that's a strategy with built-in limits. Also, your numbers aren't that close to reality, because most of those aren't deployed. Still would rather make a potential nuclear exchange better rather than not doing so.

    @Asher

    Hetz/Arrow is a good system, but I doubt it's that much better than the existing American systems (for no other reason than that we funded a lot of it and could make use of any fancy technology developed) and it's not clear how capable it is against ICBMs. I know GMD and the latest SM-3s can handle that, so I'm going to stick with recommending those. Do note that the USMC seems to be planning to buy some missiles for their new medium-range air defense stuff.

    @dood

    If the President launches a first strike without an exceptionally good reason and even one warhead gets through, he's gone. I doubt we'll ever be in a place where he's confident enough to take that risk except in the most extreme cases.

  30. October 25, 2022DampOctopus said...

    @Anonymous: you're thinking of Iron Dome, the Israeli system optimised for defending against cheap, unguided rockets. Hetz/Arrow-3 is a different system, intended (I assume) to defend against Iranian medium-range ballistic missiles like the Sejjil family. As Bean said, it may or may not be effective against full ICBMs.

    For the US it makes sense to use domestic systems (GMD, SM-3) but Arrow-3 has some potential in the export market. There was a spree of stories a few months ago about a possible sale of Arrow-3 to Germany, though the US, as a partner in the development programme, has the right to veto this.

  31. October 25, 2022Chipsa said...

    @Bean, Anonymous, Alexander: My discussion on maneuvering decoys is only in relation to the idea of having maneuvering warheads. If the warheads maneuver, but the decoys don't, then you can see that. If the decoys maneuver, then you can see the thrust plume and how it doesn't match the real warhead (wrong brightness, or doesn't match the actual acceleration, or your doppler doesn't match because you're thrusting in two opposite directions). How do we see this happening? SDA's Tracking Layer.

    If we're talking only classic ballistic missiles, then yeah, decoys don't maneuver, but neither do the warheads. I would suspect that a bus would maneuver a bit differently if releasing a heavy warhead compared to a light decoy, but that's hard to know based on open-source info.

  32. October 26, 2022AlexT said...

    tell a balloon from an actual warhead

    How about making the decoys really streamlined and the warheads really blocky, such that they'll fly effectively the same?

    I don't think that would significantly reduce the real warheads' range, since if it would, the same effect could be used to maneuver aerodynamically, which I understand isn't being done.

    And anyway, maintaining the pretense as long as possible (eg by deploying drag-inducing parts on the warheads and jettisoning them as late as possible) would probably mess with interception.

    a bus would maneuver a bit differently if releasing a heavy warhead compared to a light decoy

    Presumably the bus could throttle its engine or flare its nozzle or something to a slightly different degree after releasing a decoy vs a real warhead.

    If you're worried about the plume, inject something that burns bright (and in the right spectrum) into the exhaust when appropriate. Or shine some LEDs into it, with the right colors.

    Ideally, handle the bus so that it maintains the same acceleration between releases. I suppose that's much easier said than done, but it seems worth the trouble.

  33. October 26, 2022bean said...

    How about making the decoys really streamlined and the warheads really blocky, such that they’ll fly effectively the same?

    Then they don't look the same to the radar, which rather reduces the effectiveness of the decoys.

    I don’t think that would significantly reduce the real warheads’ range, since if it would, the same effect could be used to maneuver aerodynamically, which I understand isn’t being done.

    MaRVs have been looked at, but they usually impose tradeoffs that aren't seen as worth it.

  34. October 26, 2022Matt said...

    Given that our own ICBMs are built for MIRVs but now carry only one warhead due to treaty limitations, I'm wondering if an actual dummy re-entry vehicle (identical to the ones with warheads, except filled with ballast weight instead of a functional warhead) would be feasible. Or is that just as much a treaty violation as fielding extra warheads, given that inspectors probably couldn't readily prove the difference either?

  35. October 26, 2022bean said...

    I'm not sure what the treaties say about that, but it's the only form of decoy that would be really effective.

  36. October 26, 2022Anonymous said...

    AlexT:

    How about making the decoys really streamlined and the warheads really blocky, such that they'll fly effectively the same?

    Along with the different RCS issue there's also the part about warhead RVs being streamlined for good reason, namely accuracy and to reduce the time the enemies missile defense system has to engage it (a blunt RV will slow high up in the atmosphere providing a nice target after all the decoys have burnt up).

    AlexT:

    Presumably the bus could throttle its engine or flare its nozzle or something to a slightly different degree after releasing a decoy vs a real warhead.

    Which could be noticed.

    AlexT:

    If you're worried about the plume, inject something that burns bright (and in the right spectrum) into the exhaust when appropriate.

    If such a something existed, which it almost certainly doesn't.

    The spectrum of something burning is determined by the energy levels of the electrons in the atoms and how much energy they release when they change states, you can't arbitrarily select them.

    AlexT:

    Or shine some LEDs into it, with the right colors.

    LEDs get their color from the semiconductors they are made out of (again, electron configuration) with some also having a layer of phosphor turning some of the (usually blue) light into other wavelengths (say yellow) but that also gives LEDs a very distinctive spectrum that is not anything like a rocket plume.

    Matt:

    Or is that just as much a treaty violation as fielding extra warheads,

    Probably not, but it would mean that the rocket is much larger and more expensive than it needs to be, also the silos housing them and all the support infrastructure to get the rockets to the silos.

    Matt:

    given that inspectors probably couldn't readily prove the difference either?

    I suspect it would be pretty easy to determine that the decoy really doesn't contain anything that produces a mushroom cloud.

  37. October 26, 2022Hugh Fisher said...

    When I saw the "Missile Defense" heading on the post my first expectation wasn't possible nuclear war. Nobody has ever fired ICBMs or IRBMs at the US or allies. Yeah, past performance is not a reliable indicator of future results, but to me the "obvious solution underscored by recent events in Ukraine" would be that the US and allies need missile defense against cruise and theatre missiles, long range rocket artillery.

    Not an expert, and about everything I know comes from the YouTube channel Perun who also cheerfully admits he isn't an expert either. But it seems that in Ukraine there are a LOT of missiles flying back and forth. The Russians are using everything from Kalibr to S-300 SAMs in surface to surface mode to 1960s KH-22. The Ukrainians have their own Soviet Tochka missiles and now US GMLRS, and they really want ATACMS.

    From your past posts the US Navy seems to be on top of this with three layered long / short / last ditch missile defense for major combatants. I'm curious (as Australia is often an ally of the US on the ground) as to what you think of US Army tactical missile defense.

  38. October 26, 2022ike said...

    @Hugh Fisher

    My understanding of why missile defense is topical is that, Russia has used the threat of her strategic rockets to deter NATO from sending volunteers to shoot Russians directly. (One of) Bean's point(s) is that with better missile defense such threats would be empty and we could send a few Condor Legions to help.

  39. October 27, 2022Chipsa said...

    @Anon:

    Probably not, but it would mean that the rocket is much larger and more expensive than it needs to be, also the silos housing them and all the support infrastructure to get the rockets to the silos.

    At least with the US, the infrastructure for silos is all pretty much built, so GBSD has the throw weight for MIRV, even if we aren't planning on using it for MIRV at this time. So, it's quite possible that we are going to just stick dummy RVs on top. This goes into the US ICBM force being a strategic decoy, in that it has nukes, and therefore should be targeted, but it'll probably take more to destroy them than actually exist in the silos.

  40. October 27, 2022Placid Platypus said...

    People have said "if you're going to the trouble to include a sufficiently realistic decoy, might as well just use another real warhead instead." Am I understanding correctly that this means that the cost of warheads is trivial compared to the cost of missiles?

  41. October 27, 2022EngineOfCreation said...

    @Placid Platypus It's not only about the cost of the warhead relative to the delivery system. The problem is that decoy missiles/rockets have to function exactly like the real thing in every way.

    Say you want a 50% decoy rate. You have to plan for all-out nuclear war where you want to launch your whole arsenal at once. Then you need twice the silos, twice the SSBNs etc., twice the facilities to produce and maintain everything. At that point, saving on the warheads really does start to look like penny pinching.

  42. October 27, 2022EngineOfCreation said...

    @ike

    My understanding of why missile defense is topical is that, Russia has used the threat of her strategic rockets to deter NATO from sending volunteers to shoot Russians directly. (One of) Bean’s point(s) is that with better missile defense such threats would be empty and we could send a few Condor Legions to help.

    Most if not all of what Putin says on the war, including the nuclear threats, is domestic policy, not foreign policy. The whole war in Ukraine is domestic policy. That is why he can't lose to Ukraine: he would lose all legitimacy in the eyes of Russians, especially since Russian propaganda has painted Ukrainians as Nazi Untermenschen that yearn to return to Mother Russia and Zelensky as a western puppet that would run on the first day of the "special operation". Losing against Ukraine alone would be the political end of Putin.

    Given that context, NATO sending ground troops to Ukraine would be the best thing that could happen to Putin at this point. The war is lost either way, but a direct defeat against NATO would be a much easier sell domestically than a defeat against Ukrainians - it might even boost his popularity almost as if he had won the war. Putin wants to survive this war, both politically and literally. He's not Hitler. He's not going to start a nuclear war over Ukraine.

  43. October 29, 2022AlexT said...

    How about making the decoys really streamlined and the warheads really blocky, such that they’ll fly effectively the same?

    Then they don’t look the same to the radar, which rather reduces the effectiveness of the decoys.

    Not necessarily, you could make a balloon with a plastic nose in front of a radar-reflecting shape approximating a warhead. How accurate is a radar's picture anyway?

    MaRVs have been looked at, but they usually impose tradeoffs that aren’t seen as worth it.

    Weren't marvs meant to maneuver during/after reentry, not during the space-borne ballistic flight?

    The topic here is maneuverability in the thermosphere/exosphere and how badly would being blocky affect a warhead's flight there. The assumption being that you can tell a decoy from a warhead by how strongly its trajectory is affected by drag.

    But I think the real question is whether/how decoys would work for FOBS. My guess is the deorbit burn would be much harder to fake.

    @Anonymous

    a blunt RV will slow high up in the atmosphere providing a nice target after all the decoys have burnt up.

    I really don't think that's how it works, any of it.

    Presumably the bus could throttle its engine or flare its nozzle or something to a slightly different degree after releasing a decoy vs a real warhead.

    Which could be noticed.

    Notice a nozzle flaring? Seems improbable. It's radar, not the eye of Sauron.

    For engine throttle, see below.

    If you’re worried about the plume, inject something that burns bright (and in the right spectrum) into the exhaust when appropriate.

    If such a something existed, which it almost certainly doesn’t.

    The spectrum of something burning is determined by the energy levels of the electrons in the atoms and how much energy they release when they change states, you can’t arbitrarily select them.

    You can select what you're throwing into that plume, up to and including bypassing the nozzle and outputting a fraction of the content of the combustion chamber, since the purpose is to throttle down unobtrusively.

    Or shine some LEDs into it, with the right colors.

    LEDs get their color from the semiconductors they are made out of (again, electron configuration) with some also having a layer of phosphor turning some of the (usually blue) light into other wavelengths (say yellow) but that also gives LEDs a very distinctive spectrum that is not anything like a rocket plume.

    Yes, and the warhead owners get to choose what semiconductors they make those LEDS of. There's a decent number to choose from, are you quite sure nothing can produce the required spectrum?

  44. October 29, 2022bean said...

    Not necessarily, you could make a balloon with a plastic nose in front of a radar-reflecting shape approximating a warhead. How accurate is a radar’s picture anyway?

    The radar's picture is very accurate, and I am extremely doubtful that something like that would work. Modern radars are capable of all sorts of tricks, many classified, for telling things apart.

    Weren’t marvs meant to maneuver during/after reentry, not during the space-borne ballistic flight?

    Correct. Maneuvering in space would require making the warheads vastly more complicated, which is why nobody does it except for FOBS.

    I really don’t think that’s how it works, any of it.

    It definitely is. RV shapes are driven by the need to get through the atmosphere as quickly as reasonably possible. A blunt RV is likely to get too slow. Polaris was subsonic at the ground, making it in theory able to be shot down by normal SAMs.

    Re the monitoring of maneuvers, I'm pretty sure (drawing on old stealth in space discussions) that if the IR satellites are capable of monitoring that effectively, they can distinguish between real and decoy launches, and there isn't much you can do to disguise it. But I'm not sure the satellites are up to it in the first place.

  45. October 29, 2022Anonymous said...

    AlexT:

    I really don't think that's how it works, any of it.

    See Coming Home: Reentry and Recovery from Space by Roger D. Launius and Dennis R. Jenkins (available for free from NASA). Chapter 1 has sections called "Nuclear Warheads and the Blunt-Body Theory" and "Better Warheads".

    AlexT:

    Notice a nozzle flaring? Seems improbable. It's radar, not the eye of Sauron.

    Satellites could see it.

    AlexT:

    You can select what you're throwing into that plume, up to and including bypassing the nozzle and outputting a fraction of the content of the combustion chamber, since the purpose is to throttle down unobtrusively.

    Do that and spectral lines that don't show the few kilometer per second Doppler shift of rocket exhaust will show up and give away that you're doing that.

    AlexT:

    Yes, and the warhead owners get to choose what semiconductors they make those LEDS of. There's a decent number to choose from, are you quite sure nothing can produce the required spectrum?

    There are not an infinite number of semiconductors so it is very unlikely that someone can approximate a rocket exhaust using LEDs that can fool a spectrometer.

  46. November 07, 2022AlexT said...

    @bean @Anonymous sorry for the semi-necro, but this is interesting.

    The radar’s picture is very accurate, and I am extremely doubtful that something like that would work. Modern radars are capable of all sorts of tricks, many classified, for telling things apart.

    Surely not if the detail fits to within a wavelength, which would be 1cm or more, which seems doable for a ghost warhead. And of course the surfaces would be powdered with the correct metal to make the reflection "look" right.

    It definitely is. RV shapes are driven by the need to get through the atmosphere as quickly as reasonably possible. A blunt RV is likely to get too slow. Polaris was subsonic at the ground, making it in theory able to be shot down by normal SAMs.

    Switching goal posts here, I think. This was about decoys in exo-atmospheric flight. The point was that, if the decoys experience significant drag resulting in trajectory changes that give them away, then the RV can also be made more draggy temporarily, eg inflate some external bladders and jettison them on reentry. Is the thermosphere so thick that this would significantly cut into the warhead's range? Honest question.

    See Coming Home: Reentry and Recovery from Space by Roger D. Launius and Dennis R. Jenkins

    Thanks, will take a look! Same objection though, this discussion was about warheads and decoys flying the same-ish in the upper thermosphere, and how during that phase there's little drag to speak of, and what there is should be counterable by very slightly increasing the real warhead's drag, which should not affect the flight path. Are either of those "should" wrong?

    You can select what you’re throwing into that plume, up to and including bypassing the nozzle and outputting a fraction of the content of the combustion chamber, since the purpose is to throttle down unobtrusively.

    Do that and spectral lines that don’t show the few kilometer per second Doppler shift of rocket exhaust will show up and give away that you’re doing that.

    Are the tracking optics good enough to spot that instantly, for a fraction of the exhaust, seen through distorted atmosphere?

    Notice a nozzle flaring? Seems improbable.

    Satellites could see it.

    Same as above, minus atmospheric distortion, plus vulnerable, dazzle-able, volume- and mass-constrained platform.

    There are not an infinite number of semiconductors so it is very unlikely that someone can approximate a rocket exhaust using LEDs that can fool a spectrometer.

    Doesn't need to be identical, just close enough to make discrimination chancy. Bear in mind, the job of the LEDs would be to make up the dimming from throttling down, not fake the whole plume.

    drawing on old stealth in space discussions

    Was late to that party, but the leftovers are yummy.

  47. November 07, 2022bean said...

    sorry for the semi-necro, but this is interesting.

    Naval Gazing's policy on necros is quite clear. They're encouraged if you have something to say.

    Surely not if the detail fits to within a wavelength, which would be 1cm or more, which seems doable for a ghost warhead. And of course the surfaces would be powdered with the correct metal to make the reflection “look” right.

    Every year or two, I discover some new capability that I didn't think radars could have, much less that they do. In practical terms, I wouldn't rule out some ability to analyze details under 1 cm using, I don't know, how the surface's roughness affects the phase of the returned signals or something like that.

    Switching goal posts here, I think. This was about decoys in exo-atmospheric flight. The point was that, if the decoys experience significant drag resulting in trajectory changes that give them away, then the RV can also be made more draggy temporarily, eg inflate some external bladders and jettison them on reentry. Is the thermosphere so thick that this would significantly cut into the warhead’s range? Honest question.

    Ah. Clever idea, although I don't know how well that would work. The basic problem is that you're not looking at just a little extra drag. To get the ballistics to match, you need (Drag Coefficient * Area)/Weight to be the same, or at least very close. A typical warhead is at least 300 lb, and I don't think you can change Cd more than maybe threefold with this scheme. Which means you're going to need either a pretty heavy decoy or a lot of extra surface area. That will also add weight, and there's the potential for things to go wrong. Not to mention the risk of the other guy using means that aren't radar to tell the decoy and real warhead apart. An orbital IR camera could probably discriminate between the two pretty easily, although I'm not sure how well current systems work for that. But that's solvable.

    Was late to that party, but the leftovers are yummy.

    Technically, so was I, even though I showed up a dozen years ago.

  48. November 09, 2022DampOctopus said...

    In practical terms, I wouldn’t rule out some ability to analyze details under 1 cm

    It's possible to resolve details smaller than the wavelength of a radar, if you have a high enough signal-to-noise ratio. Conceptually, imagine adding some noise to a sine wave, then trying to reconstruct the position and amplitude of its peaks: you'll generally be able to locate them more precisely than the wave's period, but just how precisely depends on the amplitude of the noise relative to the underlying wave.

    This rapidly runs into diminishing returns, though. The rule of thumb for a radio dish - or optical telescope mirror - is that the surface should be precise to ~1/20th of the wavelength, and the practical limit for radar resolution is probably similar.

    If I had to hide the flaring of a nozzle from a radar, though, I'd just deploy a shroud around it.

    Are the tracking optics good enough to spot [non-Doppler-shifted spectral lines] instantly, for a fraction of the exhaust, seen through distorted atmosphere?

    We've reached the point in this discussion where I think we'd need a quantitative analysis. Something like: if we have an optical instrument that can detect a rocket plume in a blind whole-sky survey with a reasonable cadence (once per minute, say), and we use this same instrument to look at the spectrum of that plume, how long would it take to detect a spectral line of width X containing Y% of the energy of Z% of the exhaust? The absolute brightness of the rocket plume falls out of the comparison, and atmospheric distortion thankfully doesn't come into it: the atmosphere smears emission around spatially, but doesn't shift it in frequency, so it doesn't have any additional effect on a spectral measurement.

  49. November 10, 2022AlexT said...

    In the context of warhead decoys, would an opponent even have a chance to observe the bus before it finishes releasing its payload? AIUI, separation occurs in the early stages of flight, which would be below the horizon for any intercontinental target. E.g. the horizon for 1000km height is 3500km ish, right? Never mind something like FOBS, where most of the flight path is below the horizon of the target's continent. So separation is only really visible to satellites, which would be the first targets anyway in any significant exchange.

  50. November 10, 2022bean said...

    The SBIRS satellites (which I would expect to be the most likely candidates for this) are in GEO, and thus will give hours of warning if you try to take them out with direct attack. You might be able to launch coorbital ASATs way ahead of time and get them in position, but that runs a lot of risks of being found out.

  51. November 11, 2022Doctorpat said...

    Any technique can be overcome. The question is always: At what cost?

    The USA probably has some very sophisticated tracking systems. Though even there, there are many examples of long term, well thought out military projects that suddenly lost their way, for reasons technical, political and/or financial.

    But everyone else? A military that is resorting to weapons from World War 2 may not have radar systems much beyond the 1970s. And even if they do, are there enough to cover the whole hemisphere?

    Or did the super advanced tracking systems get announced, get half funded, only 3 stations ended up being finished, and 1 of those is down for lack of parts.

Comments from SlateStarCodex:

Leave a comment

All comments are reviewed before being displayed.


Name (required):


E-mail (required, will not be published):

Website:

You can use Markdown in comments!


Enter value: Captcha