An “ASAT Test Ban” is Not Arms Control

Writing in his regular column in Foreign Policy, @armscontrolwonk Jeffrey Lewis has recycled an old proposal that Bruce Macdonald and others have advocated for years: seek a global ban on tests of “debris creating” anti-satellite (ASAT) weapons, particularly hit-to-kill ASATs that impact their targets at thousands of meters per second and create thousands of small pieces of debris which can stay in orbit for years or centuries. The debris creates hazards for satellites and spacecraft, and poses the threat of a “chain reaction” of collisions which might make orbital space unusable.

Lewis summarizes his proposal as “no blowing things up in space that leaves a bunch of space junk.” However, while purporting to ban tests of hit-to-kill ASAT weapons, this rule would permit testing of exactly the same weapons against non-orbital targets such as ballistic missiles. China did just that in 2010, testing the same or an improved version of the weapon they’d tested three years earlier against a satellite, this time calling it a ballistic missile defense (BMD) test. Unlike China’s 2007 test, which earned worldwide condemnation for its massive contribution to orbital debris, the 2010 test was clean because of its low altitude (250 km), and because neither the interceptor nor the target missile were on orbital trajectories. The US in 2008 tested an SM-3 missile, nominally a BMD weapon, against a satellite, on an orbital trajectory but at the same low altitude, and has conducted at least 15 more tests of SM-3 against missile targets since then.

The Tuesday Test Ban

The crucial point is that testing against missiles or other non-orbital targets is completely sufficient to validate a weapon as an ASAT. Like other “ASAT test ban” proponents, Lewis seem to be pretending that if a weapon is only tested against missiles and not “against live targets in orbit,” then the weapon might have “a latent capability to threaten space assets” but somehow that “latent capability” would not be an actual capability. However, Lewis does not actually claim that not testing against satellites would imply some uncertainty about whether the weapon would work against a satellite. If that were true, it would confer some real arms control value to his proposal. So I am guessing the reason he doesn’t say it is because he knows it isn’t true. It would be like saying that if the US only tests its ICBMs against Kwajalein, that leaves some doubt about whether they could hit Beijing.

Still not convinced? It would be like saying that if missiles are only tested on Thursdays, there would remain some doubt about whether they would function on Tuesdays. Trust me, I’m a physicist. When an interceptor and target approach each other out in space, there are various factors which may make a successful intercept more or less likely. One is the relative speed, aka closing speed, of the interceptor and target. This is not a strong function of any difference between orbital and suborbital targets. It is a strong function of the geometry chosen for the intercept and, in general, especially for satellite intercepts, the closing speed can be chosen to be arbitrarily slow. Also relevant is whether the sun is shining on the target and at what angle, whether the interceptor is approaching the target from below and viewing it against the blackness of space (or whether the Moon or Sun or some other interfering space object is behind the target) or approaching from above and viewing the target against the Earth, or against the limb of the atmosphere, and myriad other factors, none of which is dependent, in any systematic way, on whether the target’s velocity vector (relative to Earth’s center) happens to be one that, at given altitude, constitutes an orbit.

To put it another way, all objects in near-Earth space, other than when they are being boosted by a rocket, are always on elliptical trajectories with the Earth’s center at one focus. It’s just that if the ellipse happens to be one that crosses the atmosphere and intersects the Earth, we call it “suborbital.” If it misses the atmosphere, we call it an “orbit.” Missiles and missile targets are generally on suborbital trajectories. But out in space, in terms of the physics of an intercept, whether the target is destined to hit or miss the Earth simply does not matter at all.

People who really don’t understand this issue, which I assume does not include Lewis, may know that missile trajectories are described as “suborbital” and the speed of an ICBM, relative to Earth and at apogee (highest point) will be a bit less than that of a satellite. They may think, therefore, that hitting a satellite is a bit more demanding. But that is not true, either. As Lewis acknowledges, “China’s so-called missile defense tests represent a big threat to U.S. satellites. While shooting down missiles may be hard, shooting down satellites is easy.” This is mainly a function of target size and visibility, and the absence of countermeasures which might accompany a missile attack. It’s also due to the fact that satellite orbits are known in advance, while a missile defense must respond to a sudden attack with incomplete knowledge of where the missiles may be coming from, where they may be going, and when. It has nothing to do with generic differences in the trajectories of satellites and missiles.

In reality, the range of closing speeds and intercept geometries for ASAT and BMD missions are broadly overlapping, and the capabilities of an interceptor for ASAT use, under the full range of relevant conditions, can be thoroughly explored in tests that do not involve orbital targets and do not leave debris in orbit. Not only is a test against an “orbital” target not needed; it adds zero information that can’t be gained in “suborbital” tests.

The intercepts can even be limited to low altitudes, as the US does and as China did in 2010, in order to ensure that no debris is left in orbit. A system that is intended to intercept satellites in high orbit (LEO, MEO, GEO, etc.) can simply be tested at low altitude to verify its intercept capability, and flown to high altitude to verify its reach; in fact, both exercises could be combined in a single test.

Thus, as arms control, such an “ASAT test ban” would have absolutely no technical meaning, and could at most have political value. It would vaguely suggest a repudiation of using hit-to-kill weapons as ASATs, and it would prevent another highly provocative explicit demonstration of the capability to do so. It might also have some norm-setting value, although the fact that nations continue to develop, test and deploy weapons which have actual (not just “latent”) ASAT capability, and that may be intended for that role, might cast doubt on the value of whatever norm we’re talking about. I think the best one can say for this idea is that it would be space environmental protection. But then, it might only spur the further development of non-debris creating Green ASATs.

Real arms control limits weapons or places material obstacles in the path of their development, testing, production and deployment. This proposal does neither.

A Better Idea

What would have some meaning is a complete ban on any collision tests above the atmosphere. That means no “missile defense” tests involving an actual intercept, either. System tests and flight tests that include “flybys,” where the interceptor is instructed to pass by a target at close range but not hit it, could still be permitted, but the lack of an actual collision would leave some genuine uncertainty about the readiness of a system, whether for ASAT or BMD missions.

American analysts, including arms control proponents who are unwilling to challenge the reigning political assumptions, reflexively reject the notion of any limits on missile defense activities, despite the strategic destabilization these activities cause and despite the ultimate futility of any attempt to provide a meaningful shield against nuclear weapons.

But ironically, it is BMD critics who have historically demanded increased testing, while the successive agencies in charge of the program have resisted such demands. We know that these systems can hit targets; they’ve done it many times. When they fail, it is usually not just a narrow miss, but rather some catastrophic system failure, and until the source of the problem has been located and fixed, there is little point in staging further failures.

Under a collision test ban, testing and debugging of existing BMD systems could continue, without having to worry about embarrassing failures to actually hit something. The hardest (and almost certainly insoluble) problem facing BMD is the discrimination of warheads from decoys, or of decoys that contain warheads from decoys that don’t. Work on that problem can continue and does not require any collision tests.

If BMD has any value at all, it is to plant doubt in the mind of a potential attacker that a (possibly limited) first strike would achieve predictable results. On the other hand, the greatest danger of BMD is that decision makers will take risks because they believe the BMD can be relied on to work if needed. Uncertainty about how well the BMD would work is therefore stabilizing in the same way that uncertainty about how well offensive missiles would work is stabilizing. We don’t want anyone feeling too confident.

An exoatmospheric collision test ban would be stabilizing and would put a real obstacle in the path of the space arms race. An ASAT-only test ban would limit the creation of orbital debris, but would not create any impediment to the further development, deployment and proliferation of ASATs either nominally disguised as BMD or explicitly deployed as ASATs.

The Real Space Arms Race

Another point not mentioned by Lewis is that the latest (non-collision) Chinese test took place just after the US launched GSSAP, an advanced space surveillance system consisting of two satellites which will travel around the geosynchronous belt and photograph and probe everything that is up there. They could be directed to interfere with or crash into other satellites, although this is unlikely, or they could be carrying weapons, although this is also unlikely. What is likely is that when China gets around to launching a system like this, some Americans will make all kinds of accusations.

Lewis rightly warns that in addition to the US and China, Russia, Japan, Israel and India are among the nations “investing in hit-to-kill systems that can be used for missile defense or anti-satellite applications.” We are indeed at the threshold of “A world filled with hit-to-kill anti-satellite missiles” and indeed that “should be very disturbing.” But this is inescapably the result of United States policy, which for the past 3+ decades has rejected space arms control and pursued the technology of space weapons. By now we have deployed a substantial hit-to-kill ASAT arsenal, nominally as missile defense, and are pursuing more advanced and more usable Green ASATs, on top of our considerable capabilities for jamming and electronic/cyber warfare against space systems. You reap what you sow. Maybe arms control proponents should begin to sow some better ideas.

I have a paper covering much of this in greater detail, posted here.

UPDATE: Regrettably, Lewis has declined to respond to this critique, and blocked my comments about it on (I can’t comment on FoPo because I don’t subscribe). In a Twitter exchange, he was dismissive and claimed that the arguments above had been “preempted” by his column. I don’t know exactly what he meant, but my best guess is that he admits his proposal is like a Tuesday test ban, but maintains that somehow it would still be meaningful — the vague suggestion, which I wrote about above, that in this case, the lack of testing on a Tuesday really does sow some doubt about whether a weapon will work on Tuesdays. Maybe he meant something else. I’d love to have him explain. Comments are open.

Leave a Reply

Your email address will not be published. Required fields are marked *