Published by The Lawfare Institute
in Cooperation With
Editor’s Note: How vulnerable are U.S. submarines in the event of a nuclear war? Some analysts have argued that, after firing their missiles, submarines would be sitting ducks for adversaries to target. RAND's Austin Long takes a hard look at this argument. He finds submarines are tough to target and highly likely to survive.
The 2018 Nuclear Posture Review’s decision to deploy a low-yield submarine-launched ballistic missile (SLBM) has produced intense criticism. One issue this presents, which I have addressed elsewhere, is the so-called “discrimination problem”—the challenge adversaries face in determining whether a launched SLBM is carrying a single low-yield warhead or a massive nuclear strike with several warheads. Another criticism of the review’s decisions is that the launch of a single SLBM with a single low-yield warhead will compromise the location of the submarine launching it. This revelation will then subject the submarine to counterattack through either conventional or nuclear anti-submarine warfare. Having spent a large part of my professional career assessing targeting options—ranging from terrorists to buried nuclear facilities, and including mobile missiles and submarines—I am sympathetic to this argument. But prosecuting an attack on a very quiet submarine is quite challenging, so it is worth analyzing potential attack options to determine the actual risk to the submarine launching a low-yield SLBM.
There is one place where ballistic missile submarines are clearly vulnerable: in port. Some experts argue that in wartime the United States could upload additional warheads on to SLBMs and then load those missiles on to submarines. This would heighten the discrimination problem, and this time-consuming procedure could only be done at one of two facilities: Strategic Weapons Facility Atlantic in Georgia or Strategic Weapons Facility Pacific in Washington state. These facilities can be readily monitored and potentially targeted. In an extended conflict, the United States is unlikely to risk the survival of its ballistic missile submarine force to attempt this upload. Indeed, it is far more likely that any U.S. submarines in port during a major war will be headed out to sea as quickly as possible.
Once at sea, submarines are elusive, so the first step for any adversary seeking to target the launching submarine is to detect the launch of the SLBM. The fastest means to such detection is a space-based system, such as the U.S. Space Based Infrared System (SBIRS) satellites, which detect the heat of rocket motors. At present, none of the three states the Nuclear Posture Review mentions (Russia, China, North Korea) have a robust space based early warning system. North Korea has no space based early warning. China is rumored to be deploying a space based infrared early warning system, but even if true the system is still at an early or even experimental stage. Russia has been deploying early warning satellites since the 1970s but they have had serious problems. At present, Russia only has two modern infrared satellites, greatly limiting their utility for detecting SLBM launches.
Instead, most potential adversaries rely on ground based radars for early warning, which can only detect an SLBM launch after the missile reaches a certain altitude. However, SLBMs can be launched on “deceptive trajectories” that complicate determination of the point of origin using radar. This delay and possible uncertainty about the missile’s point of origin limits (though does not eliminate) the utility of these early warning methods for targeting a ballistic missile submarine (SSBN).
For long range planning purposes, though, it is prudent to assume at least Russia and China could build effective space based systems capable of detecting SLBM launches even if they lack the capability today. It is also prudent to assume that these systems could be able to pinpoint the point of launch with reasonably high fidelity. Then the question becomes: How will the adversary attack the submarine they have located?
First, it is important to note the submarine will not sit still. While there is a brief post-launch period where the submarine, which must be near the surface to launch, will not be at full speed, it can quickly dive and accelerate. At the official quiet cruising speed of 20 knots, the submarine can travel six miles in any direction in about 15 minutes.
This presents significant challenges unless the adversary has anti-submarine warfare assets located very close by. Neither China nor North Korea has any meaningful anti-submarine warfare capability capable of quickly reaching far in to the Pacific, where U.S. ballistic missile submarines are likely to patrol. Only Russia, which maintains maritime patrol aircraft and a fleet of both surface combatants and nuclear attack submarines that can patrol the Pacific and North Atlantic, might be able to relay information to forces ready to prosecute an attack.
Russian maritime patrol aircraft would be challenged to reach the likely patrol areas in the North Atlantic of a U.S. ballistic missile submarine, given they would have to cross airspace dominated by NATO air defenses and fighter aircraft. Even if this were not true, Russian maritime patrol aircraft travel less than 500 miles per hour—at best they would need at least two hours to travel from a forward location like Kaliningrad to an area south of Iceland and west of the United Kingdom where a U.S. submarine might have carried out a launch. In that time the submarine could have transited fifty miles in any direction, yielding a search area of more than 7500 square miles. More realistically, the flight time would be from the Kola Peninsula and would be longer than three hours, expanding the search area to more than 20,000 square miles. This would be a nearly insuperable search problem even without NATO air defenses. If Russian maritime patrol aircraft were already patrolling within a few hundred miles of the SSBN when it launched, the search problem might be manageable. But this would require having a refined sense of where U.S. SSBNs patrolled, good timing to be on station at the time of launch, and again being left alone by NATO air defenses.
Russian surface combatants (e.g. destroyers) could be patrolling closer to U.S. SSBN launch points in the North Atlantic than maritime patrol aircraft. These combatants can be armed with anti-submarine rockets armed with nuclear depth charges or homing torpedoes to attack submarines at a distance. The United States had a limited version of such a system, the ship-based Anti-Submarine Rocket (ASROC) with a 10-kiloton warhead. Russia currently deploys a similar system with a range reported at about 100 kilometers. Yet a Russian surface combatant is unlikely to survive very long in the middle of the North Atlantic during a shooting war with NATO. These ships would be subject to attack from NATO submarines, surface ships, and aircraft. Moreover, U.S. SSBNs could readily be warned of the presence of such ships and avoid them.
Russian attack submarines already in the North Atlantic might have a better chance than maritime patrol aircraft or submarines. Yet Russian attack submarines are not substantially faster (if at all) than U.S ballistic missile submarines at speeds at which they remain quiet. So unless one of the Russian attack submarines was already very close to the launching submarine it would have to be very lucky to close with and track the U.S. submarine with sonar. While sonar detection ranges are variable, it is probably the case that, unless the Russian attack submarine could get to about twenty to thirty miles (one sonar convergence zone under nearly optimal conditions) from the U.S. submarine, it would not be able to prosecute an attack even if it knew where the submarine was at time of launch. In significantly worse conditions for sonar (which are not uncommon in the North Atlantic) the Russian submarine would need to close to within a few thousand yards of the U.S. submarine. Of course, like the surface combatants, the submarine could be armed with the Russian version of ASROC (with a nuclear warhead), but it would still need to close within the missile’s range of 100 kilometers.
Given their similarity in speed, this means that the Russian submarine would already have to be quite close to the U.S. submarine when it launched in order to be able to close the distance—about 100 kilometers at the outside. This is certainly possible but statistically quite unlikely unless the Russians already had some means to cue their submarines to where U.S. submarines patrolled. In other words, unless Russia can detect U.S. submarines prior to launch, they probably have little ability to make use of the location information provided by detecting a launch. Moreover, in a war Russian attack submarines will likely have many other missions, ranging from launching cruise missiles to protecting Russian SSBNs from U.S. (and probably U.K.) attack submarines. Only a few, if any, Russian attack submarines might be available for this mission.
Of course adversaries could use other means to attack a submarine that has revealed its location. One obvious means would be to develop a system to launch nuclear weapons on long-range ballistic missiles back at the submarine. As noted, both the United States and Russia have deployed nuclear-armed anti-submarine rockets. There is no reason an adversary could not make a longer range and larger version of this system.
But even nuclear weapons have limited utility in attacking underwater targets moving at high speed. Let us assume Russia wants to build such a system to attack U.S. submarines. First it would have to modify ballistic missile warheads to become depth charges; basically this would entail ensuring the warhead can survive impact with the ocean and then detonating at a depth of a 100 or more feet. This is not impossible but would be challenging—the U.S. ASROC had a much shorter range and so could easily use a parachute to slow the warhead before it hit the ocean. A ballistic missile traveling thousands of miles would be moving at many times the speed of sound, so surviving impact (perhaps by adding multiple parachutes) and fuzing at the right depth would be a costly engineering challenge.
Even then, the destructive radius of a nuclear blast underwater is less than one might imagine. Based on the U.S. Wigwam nuclear test, it would require about 700 pounds per square inch (psi) overpressure to significantly damage or perhaps sink a submarine of 1950s vintage. Other estimates conclude roughly 1500 psi would be required for a submarine of contemporary vintage. A 1-megaton nuclear warhead produces 1500 psi at a distance of about 2.25 miles underwater. This means a 1-megaton warhead could cover about 16 square miles with sufficient blast pressure to severely damage or sink a submarine. Russia could, in theory, convert the roughly 1-megaton warheads on its SS-18 heavy intercontinental ballistic missile for this purpose.
The best case for the Russians would be this: They detect launch and then retarget SS-18s almost instantly. Even in this best case, the SS-18s would probably take at least 20 minutes to reach their targets. In 20 minutes, a U.S. submarine could travel about eight miles in any direction, creating an area of approximately 200 square miles where the submarine could be. The Russians would thus need to deliver about 13 1-megaton warheads to cover this area with 1500 psi. The SS-18 can carry 10 warheads, so this might actually be a good exchange for the Russians—two SS-18s to sink or at least damage a U.S. submarine carrying 19 Tridents (20 minus the one fired) with probably four or five warheads per missile.
But that’s Russia’s best-case scenario. Assuming there is roughly a 10-minute delay from the time of U.S. SLBM launch to the time of Russian missile launch, due to the challenge of detecting the launch (e.g. if there is cloud cover over the launch point) and then a lag for retargeting, the U.S. submarine would have thirty minutes before Russian missiles arrived and could move 12 miles in any direction. This expands the area Russian warheads would have to cover to about 450 square miles. It would then require about 29 warheads to cover the area with 1500 psi. If the delay is 20 minutes, the requirement increases to about 50 warheads.
Under some unfavorable sets of assumptions—Russia’s best-case scenario—experts can argue that a low-yield SLBM would not be worth deploying as it would put U.S. ballistic missile submarines at unacceptable risk. But the costs to the Russians of developing the capability to target U.S. submarines with nuclear weapons are substantial. It would require the successful deployment of additional satellite systems, warhead modifications to detonate underwater, and rapid retargeting systems. In contrast, the costs to the United States are low, requiring only modification to an existing warhead.
When realistic conditions are taken into account, it is therefore arguable that a low-yield SLBM is at worst a successful competitive strategy, requiring the Russians to spend massively to develop an effective counter to a modest investment. Should they choose to do so, these resources will be spent whether or not the United States ever uses the low-yield SLBM. This is ideal for the United States; should the Russians (or in the future other adversaries) spend the resources required to counterattack, then the United States could simply choose to never use the low-yield SLBM. If the Russians don’t choose to expend these resources, then the United States has a low-risk option for prompt low-yield nuclear use with a high probability of penetrating defenses.
In this context it is important to recall U.S. ballistic missile submarines have been committed to various limited nuclear options in the past. In other words, there have been scenarios in which a submarine would launch less than in its entire complement of missiles. For example, the recently fully declassified Carter administration presidential directive on ballistic missile submarine commitments to NATO notes that SSBNs assigned to NATO will carry some “NATO-pure” missiles, totaling 400 warheads, committed to Supreme Allied Commander Europe (SACEUR) targets. The remainder of the missiles on these SSBNs would be assigned to the Single Integrated Operational Plan (SIOP). So the United States was committed to the possibility SSBNs could launch SLBMs on behalf of SACEUR while retaining missiles dedicated to the SIOP, despite the risk of revealing SSBN locations. There were certainly contingencies in which both SACEUR’s plan and a SIOP option could be executed simultaneously but it is clear that was not always going to be the case. This reality has long been recognized by the U.S. submarine force, who have acted accordingly. As a 2008 report of the Naval Studies Board of the National Research Council noted, “the Navy long ago developed techniques to protect SSBNs after missile launch.”
Despite this long-standing feature of U.S. nuclear planning, the Russians have not attempted to deploy a long-range nuclear counterattack system. The Russians have not invested heavily in satellite early warning, much less an entire system to target submarines with nuclear warheads. It seems unlikely the deployment of low-yield SLBM warheads will change Russian calculus about the utility of investment in these systems, but if it does, the costs will be substantial relative to the costs to the United States. And even then, the consequence would only be that a low-yield SLBM might not be safely usable at some future point, after Russia (or in the future China) spends substantially on a countermeasure system.