June 23, 2024

Military Spaceflight Part 8 - Nuclear ASAT

Satellites have found a number of military uses: reconnaissance and intelligence, communications and navigation. But all of these require the satellites to be intact and functional, and that raises the possibility of denying the enemy these capabilities by degrading or destroying their satellites. That's right. It's finally time for ASAT.1

Bold Orion

First, a few words of definition. We can divide ASAT weapons into two fundamental categories: direct ascent and coorbital. Direct-ascent weapons, as the name suggests, ascend directly from the surface to intercept their target, while coorbital ASAT involves launching into the same plane as the target, then approaching and using some sort of warhead at close range. Direct-ascent is conceptually simpler and cheaper (as it can be executed by a fairly small missile if it's in the right place), and can use the target's own speed in place of a warhead if it can get a hit.2 The downside is that it requires incredibly precise guidance, with a typical launch window being a second or less, even if the system is fitted with a nuclear warhead. Coorbital ASAT, on the other hand, needs a full-size booster, but once the price for that has been paid, the guidance problem is greatly simplified, removing the need for a nuclear warhead even with 60s technology, and it is theoretically possible to leave the ASAT shadowing the target, then kill it at a moment's notice when war breaks out. The US has always preferred direct-ascent systems, while the Soviets and Russians often went for coorbital ones. Also worth noting early on is that not all satellites are created equal from the perspective of ASAT systems. An imagery satellite a couple hundred kilometers up is a far easier target than a GPS satellite or a comsat in GEO that are tens of thousands of kilometers up. This is more true for direct-ascent systems than for coorbital systems, but it affects both of them, and news stories about LEO systems that bring up the danger to GPS and the like should not be trusted.

Nike Zeus

The first American ASAT efforts were driven not by concerns about conventional space capabilities but by fears that the Soviets would place nuclear weapons in orbit, where they could be used as part of a surprise attack.3 As a result, all three services began proposals for anti-satellite weapons, each based in the environment closest to is heart. The Army's proposal was a straightforward adaptation of its nuclear-tipped Nike Zeus ABM, while the Air Force put forward a variant of the Bold Orion air-launched ballistic missile, which was even tested in the ASAT role in 1959 and managed to get within what would have been lethal radius of the target satellite had it been carrying a warhead. The Navy's plan was to mount an interceptor on top of a Polaris missile, although this was quickly scrapped because Polaris tubes were seen as too valuable to give up, and a follow-on plan for an ASAT based on surface ships didn't get very far. They also ran a test using an air-launched weapon under the codename Hi-Hoe, although the program was cancelled shortly thereafter. Ultimately, the Nike Zeus system was given the go-ahead under the codename Project Mudflap, the lower technical risk offsetting the fact that it could only fire from one site (Kwajalein), giving it at most two shots a day on a given satellite.

Even then, performance of the Nike-Zeus was fairly marginal,4 and the Air Force put forward a proposal to make use of the Thor IRBM, which was being withdrawn from service in the early 60s and had the performance to be genuinely useful.5 The Thors would be based at Johnston Island, an isolated speck of land in the central Pacific that is probably best-known as the launch site for Operation Hardtack and Operation Fishbowl, a series of high-altitude nuclear tests. One of these, Starfish Prime, famous for producing an EMP that caused damage in Hawaii, 1,400 miles away.6 This test also revealed that nuclear detonations in space would cause various forms of radiation damage to satellites, and with the increasing importance of space capabilities to the US, people began to question the importance of nuclear ASAT. But the Thor ASAT system, under the codename Program 437, continued towards service.7 It was first tested in early 1964, when a Thor was fired at a spent rocket stage, passing within the lethal range of its W49 nuclear warhead. To make sure that the rather complicated Thor was able to fire in the narrow launch window, the system actually involved two missiles counting down in parallel, a feature that proved useful during the second test, when the primary missile had a failure and the backup missile ended up carrying out the test.

The result of a high-altitude nuclear weapon

Concerns about nuclear warheads also drove plans for satellite inspection at close range, although there was frequent overlap between inspection and killing of satellites. The Air Force proposed a system known as SAINT, which was intended to rendezvous with and inspect hostile satellites, and maybe even destroy them, although that functionality was deemphaisized fairly early on.8 SAINT was too ambitious, and there were concerns about the need to rendezvous giving the target time to respond, so Robert McNamara cancelled it in 1962. But the potential need to get a camera close to a hostile spacecraft remained, and in 1964, work began on Program 437AP, which replaced the nuclear warhead with a camera payload that could photograph the target from within a few miles. Most tests were successful, but it was never used operationally for fear of the diplomatic fallout that would follow a close inspection of a Soviet satellite. The base Program 437 soldiered on for a few more years, although the signing of the Outer Space Treaty, which prohibited positioning nuclear weapons in orbit, along with greater knowledge of the Fractional Orbital Bombardment System, a big ICBM that would briefly put a warhead in orbit so it could attack from the south outside the view of most US missile defense radars. With the major role gone, In 1970, the system was taken off alert, with many components moved to the Continental US, meaning it would take about a month to launch. Despite declining interest, damage to the facilities on Johnston Island from Hurricane Celeste in 1972 was repaired, and it soldiered on until 1974, when the whole thing was shut down.

But the demise of Program 437 was not the end of the ASAT story in the US, and work began on a replacement, one without the problems that came with the use of a nuclear warhead, almost immediately. We'll pick up that story next time.

1 The standard abbreviation for anti-satellite weapons.

2 Rick Robinson's First Law of Space Combat: An object impacting at 3 km/sec delivers kinetic energy equal to its mass in TNT. As a typical object in LEO is doing about 7.5 km/s, the interceptor will hit with energy equal to at least 6.25 times its mass, which should be more than enough to destroy the target.

3 To be clear, this wasn't the most rational fear. First, satellites in the late 50s and early 60s were notoriously unreliable, so this wouldn't be good as a deterrent and might not work all that well even in a planned attack. Second, the fears of a sneak attack were significantly overblown, because of the orbital mechanics stuff we discussed at the start of the series. Each orbiting bomb would only be in range of a given target about twice a day, although this could in theory be coordinated around if you wanted an attack at a specific time. But the US could watch for such launches, and might be able to back out the time of attack and hit first. On the whole, it was easier to just use ICBMs.

4 Exactly what this means is complicated, and I don't want to spend a bunch of time working out the performance details from the records I can find. Astronautix credits it with a maximum altitude of 560 km, although this has to be if fired straight up, and another source says 200 miles (not specified which one) and a third says 174 miles. It is of course worth noting that something like this can only achieve maximum altitude if fired straight up, and attempting to hit something at a greater distance will decrease altitude capability.

5 My main source on this says 700 nautical miles of altitude and 1,500 nautical miles of crossrange. Given that most sources agree that the Thors used were the standard ballistic missile version, these two are mutually exclusive, and I'd guess that altitude at 1,500 nm would be no more than 150 or maybe 200 nm.

6 To be clear, the threat of EMP is heavily overrated. Starfish Prime knocked out a few streetlights in Hawaii, not power across the whole island.

7 It is worth noting that the selection of Thor over Nike-Zeus came against the background of McNamara battling the deployment of Nike-Zeus ABMs for deeply stupid reasons, and it is very possible that he wanted to make sure all traces of it were dead.

8 That said, there remains a lot of commonality between inspection and ASAT, even today.


  1. June 24, 2024redRover said...

    Each orbiting bomb would only be in range of a given target about twice a day, although this could in theory be coordinated around if you wanted an attack at a specific time. But the US could watch for such launches, and might be able to back out the time of attack and hit first.

    Couldn't they cluster all of their in-orbit weapons? Like, if they slowly launched a warhead a month (or whatever), but put them all in an 11AM/11PM ET orbital plane, it would mean that any "surprise" attack would have to be coordinated for 11AM/11PM, and they would have great difficulty striking outside that window from orbit. However, because it would be a gradual build-up, it also seems like they wouldn't have a clear escalation point to counter it, short of either preemptively blowing them out of orbit, or going into maximum readiness twice a day.

    (Though I suppose the bigger deterrent would be that it would be destabilizing for both sides...)

  2. June 24, 2024Alexander said...

    Depending on orbit and how much ΔV the bombs have, deorbiting them might take an comparable amount of time to an ICBM flight anyway. And they are presumably more vulnerable in orbit than in a silo, and harder to maintain. I blame panic over Sputnik for making people worry about satellites rather than the rockets that could put them into orbit.

  3. June 25, 2024bean said...


    I think the problem is that it's going to be pretty obvious what you're up to. And at that point, the obvious thing for the US to do is launch a strike just after the bombs go over, and hope the 12-hour window, plus some work by Project 437, can thin them a bit. In practice, just easier to use ICBMs. Also, yes, you're probably going to have about 15 minutes between when you start deorbiting the warheads and when they hit. That's enough to let a lot of the bombers and some ICBMs get off.

  4. June 25, 2024Emilio said...

    On Amazon Prime there is a French series about an orbital Soviet nuclear bomb, called Totems.

  5. June 25, 2024EngineOfCreation said...

    How long does it take a coorbital to get from launch to kill range at the minimum? And if it shadows its target first, would it not be identifiable as a threat and even counterable by direct ascent?

  6. June 25, 2024Coffeebean2017 said...

    RE, footnote 7; do you plan to ever do a full write up on the McNamara DoD? I've read bits and pieces in books and your other articles but a dedicated article tying it all together would be an interesting read.

  7. June 25, 2024bean said...

    It's on my long-term list. The problem is that doing it right would involve a lot of work to make sure that my sourcing is nailed down, which is not something I've wanted to do just yet.

  8. June 25, 2024bean said...

    How long does it take a coorbital to get from launch to kill range at the minimum?

    I have references to the Soviets doing it within a single orbit, but that took a fair bit of development, and 2-4 was common in the early version.

  9. June 25, 2024John Schilling said...

    If you put a whole bunch of satellites in approximately the same Low Earth Orbit, people are going to notice, and they're going to get suspicious as to what you might be up to. Then they'll use some of the tricks from the previous article in this series, at which point they'll know what you are up to.

    And what you are up to, is deploying an obvious first-strike arsenal in a highly provocative manner in plain view and easy reach of the intended target. Of course they're going to blow them up pre-emptively, if you don't back down and take them out of orbit first.

    This is straight out of the Cuban Missile Crisis playbook. If you've got a choice between deploying your nuclear missiles in hardened silos deep in the interior of your own nation, or on submarines lurking somewhere beneath the world ocean, or out in the open in a proxy nation right on the enemy's border, exactly one of these plans results in the other side saying "yeah, that's not going to happen" and making it stick. And Low Earth Orbit is worse than Cuba or Turkey, because it is literally hanging over the enemy's head and it doesn't have a fig leaf of national sovereignty to hide behind.

  10. June 26, 2024john.schilling@alumni.usc.edu said...

    First-orbit rendezvous (within ~90 minutes of launch) is definitely possible; it was I believe first demonstrated by Pete Conrad and Jim Gordon on Gemini 11, in 1966. That requires very tight control of the original launch, and as bean notes it's normal to settle for rendezvous within a few orbits. Or even a few days with e.g. visits to ISS.

    But if you're worried about the enemy interrupting your play with a direct-ascent anti-anti-satellite weapon, you bite the bullet and do the math for first-orbit rendezvous. At that point, the enemy would need an ASAT launch site maybe 15 degrees west of your own to make the counter-intercept.

    There would be some interesting geopolitical implications if anyone were serious about that. 15 degrees west of Baikonur would put the hypothetical American site in one of the Persian Gulf nations, or Yemen or Somalia. 15 degrees west of Vandenberg, Russia would pretty much need a launch site in French Polynesia.

Comments from SlateStarCodex:

Leave a comment

All comments are reviewed before being displayed.

Name (required):

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


You can use Markdown in comments!

Enter value: Captcha