The principle of the Cold War was Mutually Assured Destruction... but what if this idea is now obsolete?

In the 1950s the US planned to deliver hundreds of atomic weapons to Soviet targets via a bomber fleet. This was to prevent retaliation by obliterating all obvious sources of enemy command, launch, and industry.

With the development of Intercontinental Ballistic Missiles in the 1960s the reality changed, from the prospect of a first strike knock out to mutual annihilation.

Today the USA and Russia possess about 7,000 warheads each. The PRC has just under 300, and India and Pakistan have about 120 each. But are these numbers justified? We want to substantially cut the number of nuclear weapons held worldwide, and to do so we must prove that a much smaller number of warheads is still capable of being a viable deterrent.

Digital civilisation is highly complex and fragile, so instead of trying to wipe out the enemy, it makes more sense to attempt to inflict what we shall call "fatal wounds". These are a limited number of nuclear attacks upon populous and strategic cities which would cause so much chaos it would effectively cripple the enemy nation.

What is the minimum number of nuclear warheads necessary to knock out contemporary America or Russia or China? What should the targets be and why? Consider the question in the context of the USA wanting to knock out Russia and China, or Russia to take out the PRC and USA, or the PRC to get rid of Russia and the USA.

This being a reality-check question, it is possible to reject the premise of there being critical points of national failure (in terms of cities). Though this has to be specific to contemporary social and technological circumstance.

  • 2
    $\begingroup$ I fail to see how this is different from the MAD. Side A attacks and try to surgically wipe out "just the critical points" of side B, then side B counterattack on large scale, to which side A has also to answer... $\endgroup$ – L.Dutch Feb 13 '18 at 15:11
  • 7
    $\begingroup$ Are you basically asking, "Without access to the classified documents which contain the answer to this question, how can I defeat each of the nuclear superpowers?" $\endgroup$ – Cort Ammon Feb 13 '18 at 15:37
  • 1
    $\begingroup$ @CortAmmon Many of these documents, at least from the early cold war, have been declassified. But society and technology have moved on a lot, even from the end of that era. So today's official strategy is still very much let's wipe them from the map... instead of... let's create critical points of social failure given current technology and society. I think this is sufficiently distinct to warrant a question by asking how few nukes are required to cripple A B C. $\endgroup$ – inappropriateCode Feb 13 '18 at 16:19
  • 1
    $\begingroup$ What defines "knock out" in your question? When you're looking for a minimum application of force, you have to specify minimum damage as well. Consider the UK. It would be pretty easy to specify the number of nukes that could hurt the civilian networks, but UK's nuclear capabilities are known to be "fail deadly." If you cripple their entire nation, their nuclear submarines are still authorized to make nuclear strikes against you. When you're finding "minimum use of force," is it acceptable if that minimum permits them to take nuclear shots back at you? $\endgroup$ – Cort Ammon Feb 13 '18 at 18:27
  • 2
    $\begingroup$ Meta Question about on / off-topic of this question here $\endgroup$ – Mołot Feb 14 '18 at 17:16

One, detonated in the Van Allen Belt, at least in theory. It goes like this, the modern world is hyper-reliant on electricity and electrically powered machinery, so in theory the EMP from an enhanced nuclear weapon is potentially many times more destructive than the physical explosion and fireball. The idea is that the Van Allens will magnify the EMP effect of a nuclear detonation and spread the amplified EM Radiation around the world creating a complete blackout of all civilian and many military power systems and put the human race back into the industrial age overnight with little to no industrial age steampowered equipment around to actually run things with.

As an important note the group using said weapon could in fact shield themselves from much of the damage if they can crash their electrical grid just prior to detonation.

  • 1
    $\begingroup$ Interesting. Out of the box, and also suicidal. Good for a terrorist, not so much for a government $\endgroup$ – nzaman Feb 13 '18 at 15:51
  • $\begingroup$ @nzaman I can see why Ash answered my question in such a way, and have edited the question to be more specific. I hope this clarifies things. Sorry about that. $\endgroup$ – inappropriateCode Feb 13 '18 at 16:30
  • 9
    $\begingroup$ Do you have any citation on the Van Allen belts magnifying EMP? $\endgroup$ – ceejayoz Feb 13 '18 at 18:52
  • 4
    $\begingroup$ @dfri Nothing in that link backs up the idea that the Van Allen belts magnify the effect. $\endgroup$ – ceejayoz Feb 13 '18 at 20:38
  • 2
    $\begingroup$ If I'm reading that right, they're saying the belts would capture the radiation, not magnify the EMP. Would make the belts more intense, which would definitely present issues for satellites transiting them. $\endgroup$ – ceejayoz Feb 13 '18 at 21:38

It all depends on whether the belligerent parties are sane or not.

  • If the belligerent parties are sane, then a modern form of the Mutual Assured Destruction (MAD) doctrine applies. Consider for example the position of the United Kingdom or France, who maintain just enough nuclear warheads to be certain of their ability to hit Moscow. (Moscow is protected by a ring of anti-missile defences; so in order to be certain that one missile hits about a dozen or so need to be launched.) Russia is sure to win a nuclear exchange with the UK or France, but with the certain loss of Moscow. Since Russia is sane, they won't do it.

    In order for this strategy to work, the weaker party must maintain a capability of guaranteed retaliation, usually in the form of sufficient nuclear ballistic missile submarines so that at least one or, preferably, two of them are always at sea.

  • If at least one of the belligerent parties is insane, then no amount of nuclear warheads is enough, because a madman is, by definition, mad.

So the sad reality is that the U.S.A. cannot knock out Russia or China, nor can Russia knock out the U.S.A. Both Russia the the U.S.A. operate enough ballistic missile submarines to be certain that even in the case of a devastatingly massive nuclear attack they will be able to retaliate.

  • $\begingroup$ Cite for "in order to be certain that one missile hits about a dozen or so need to be launched" ? Because that's the answer to the question IMO: 1 + 12(?), times X, where X is the number of targets. A cite for "maintain just enough nuclear warheads to be certain of their ability to hit Moscow" would be interesting as well. $\endgroup$ – Mazura Feb 13 '18 at 20:25
  • 1
    $\begingroup$ @Mazura: Take for example France. At present their Strategic Oceanic Force has a total of 4 Triomphant-class ballistic missile submarines, providing the ocean-based component of their Force de dissuasion. Realistically, this allows them to have one or maximum two submarines on patrol at any time; each submarine has 16 M45 or M51 missiles. $\endgroup$ – AlexP Feb 13 '18 at 22:41
  • $\begingroup$ "France has deactivated all land-based nuclear missiles." - "Since the French military judged that a full-scale invasion of Western Europe by the Soviet Union and its Warsaw Pact Allies was unlikely to be stopped by conventional armaments, these short-range nuclear missiles were meant as a "final warning" (ultime avertissement in French) which would tell the aggressor that any further advances would trigger a nuclear armageddon upon its major cities and other important targets." – FdD . Close enough: one sub with 16 missiles. +1 $\endgroup$ – Mazura Feb 13 '18 at 22:58
  • $\begingroup$ "whether the belligerent parties are sane or not" If you play chicken game, then the sane thing is to pretend to be mad, sadly :( $\endgroup$ – Vashu Feb 13 '18 at 23:33

The concept that just because a nuclear device is detonated suddenly missile silos are all firing off like crazy everywhere is actually outdated, there are many options on the table that are all situationally dependent. The most likely cause for nuclear war is that two nuclear powers are engaged in conventional warfare and one begins losing and deploys a nuclear weapon against a military target. After this happens we have several options:

Option 1: Call their bluff. They aren't REALLY prepared to commit to a nuclear war, this is a scare tactic. Rally on the moral high ground that you will not use nukes and press the attack on the ground.

Option 2: tit-for-tat reciprocation. We signal that we are perfectly willing to enter into a nuclear war but would prefer not to by striking a single military target of equal value to the one we lost. The hope is that they realize the gravity of what they are escalating into and decline to escalate the situation further.

Option 3: Escalation. We destroy 2 targets of considerably higher value than the one they did. We are signalling that we aren't just ready to enter into a nuclear exchange, but we are willing to be the ones who escalate it.

Option 4: limited exchange. We retaliate against multiple targets avoiding civilian population centers and instead focusing on missile launch sites, bunkers, and military installations. We are hoping to destroy or disable enough of their ability to retaliate in one blow with minimal casualties because we believe their successful strike may have emboldened them into attempting to do the same to us.

Option 5: All out warfare. We fire everything we have at any target we can because we believe that now that our enemy has opened pandoras atomic box there is no going back.

A very major part of it is psychology, you don't always need to obliterate an enemy to defeat them. So the question "how many nukes to defeat a major superpower" is highly culturally and situationally dependent.

The USA has historically been a very, very poor nation to attack. The Lusitania, Pearl Harbor, and 9/11 have historically demonstrated that trying to psychologically sucker-punch the USA into an early defeat backfires spectacularly. The USA tends to actually GAIN morale from such events, repackaging a disastrous defeat or a horrific disaster into righteous fury and quickly mobilizes its culture for all out-war. The problem with the USA is that they grow tired of war equally quickly. The Tet Offensive in Vietnam or the Chinese offensive at chosen reservoir in Korea demonstrate how even a losing battle fought at the right time can shock and demoralize the war-weary American people into agreeing to a cease-fire or withdrawing early. A nuclear strike against the us right when the end of the war was in sight could very effectively break US morale just like the Tet offensive did in Vietnam.

Russia is a very different beast, they have always been stoic line-holders. The psychology of how they operate is reflected in a "retreat into winter, allow the enemy to exhaust themselves, counter attack when we've weathered the storm" mentality that is heavily ingrained to their culture. The Russians also rule more along oligarchical lines than the USA and components of the military are seen to a large part as expendable so long as they achieve some goal in preserving the momentum of the whole. Soviets tend to have a much more defensive mindset. You'd have to pound them pretty hard to break them. Look at Stalingrad, they lost almost half a million men fighting for a city that contained nothing of value except their leaders name. They were more than willing to sacrifice half a million men simply to create an expensive tar-trap for the nazis. You would have to launch a a decent sized series of strikes crippling their military to break the Russian stoicism.

The Chinese are a bit of a wild-card. The Majority of engagements they fight seem to be civil wars. The long march, and the hundred flowers purge shows that there are really two cultures in china. The inner party and everyone else. Everyone else are expendable in numbers the Russians would even find appalling. China's military is structured to match, despite all what you hear about their hundreds of millions of soldiers, there really two militaries in china. The first is the rank and file conscripts, the 2nd are the elite inner party controlled units. These inner party units are dedicated mostly towards preventing the non-party assets from deserting or rebelling, and that's during peace time. To defeat china you would need to launch a decapitating strike and hopefully without the inner-party's leaders their control over the elite units would collapse. Without guns at their backs the rank and file non-party military would surrender or desert.

These are all approximations and hypothetical meant more to demonstrate the differences of response. The reality is we've never tested nuclear tactics in battle so any talk of nuclear war is going to remain thankfully theoretical.

  • $\begingroup$ This is an interesting insight, but it would be improved by any references and elaboration on the defeat psychology of any of three powers mentioned. $\endgroup$ – inappropriateCode Feb 13 '18 at 20:24
  • 1
    $\begingroup$ Good idea, Ill edit and elaborate a bit on that. $\endgroup$ – TCAT117 Feb 13 '18 at 20:31
  • 1
    $\begingroup$ "Excepting ancient times [the Chinese] really havent fought anything but civil-wars." A couple others, too. World War II, for instance. $\endgroup$ – Ray Feb 13 '18 at 21:50
  • $\begingroup$ Correct, slipped my mind. I also pointed out they attacked the US in korea. I ought to edit it to say they "mostly" fight civil wars. Even ww2 was to a large extent various factions trying to force the others into being worn down by the japanese so they could win the upcoming civil war more easily. $\endgroup$ – TCAT117 Feb 13 '18 at 21:56

That depends on your definition of crippling damage.

Back in WWII, both sides thought about "target sets" for strategic bombing that would cripple the enemy economy. The Axis never really got to put it into practice, the Battle of Britain came closest but Germany didn't stay the course. The Allies got to try it.

  • They considered knocking the electricity grid out. Too many separate targets.
  • They went after aircraft production, including the famous/infamous Schweinfurt raid.
  • Going after transportation was partly effective. Going after oil was even more effective.

Lessons after the war showed that battle damage assessments of these strategic raids were overly optimistic. Germany was quite capable of substituting materials, of salvaging machine tools when buildings were damaged, etc.

Of course a megaton-range thermonuclear weapon is different from a couple hundred GP bombs. Still I believe that the lessons translate. Killing a city, any city, would not cripple a major nation. Killing any dozen cities won't cripple it.

Of course one nation which such damage would be at a major disadvantage fighting another nation which got off more lightly. If that is your definition of crippling, consider going after aircraft manufacturers and stockpiles of AAMs and AGMs. There are much fewer of those than there used to be, and a carrier battle group or deployed air wing would be inhibited in their operations if they know that the spares and missiles on board are it.

  • $\begingroup$ This is an informative way of looking at it, though the social and technological differences between now and then are staggering. Especially considering how many heavy industries necessary for war have been outsourced. Is there any more contemporary analysis to compliment your conclusion that there are not points of failure in terms of cities? $\endgroup$ – inappropriateCode Feb 13 '18 at 16:24
  • 2
    $\begingroup$ @inappropriateCode, I'm sure the most recent studies are still classified. You might note that the definition of crippling problems tends to slide. Consider the recent chip shortages -- calling them a famine is a tad alarmist ... en.wikipedia.org/wiki/Chip_famine $\endgroup$ – o.m. Feb 13 '18 at 16:59

On second thoughts, my first answer missed they key point.

As a credible deterrent, a country needs enough weapons so that it can inflict unacceptable damage even after an enemy surprise strike hurts their forces.

Unacceptable damage is not the same as crippling damage. The Brits called it the Moscow Criterion -- the ability to take out Moscow or Leningrad in exchange for the entire UK (this Wikipedia page gives a brief explanation in passing). The French had similar calculations, but I don't know if they had a snappy name.

  • $\begingroup$ The French used the phraseology that they only needed enough nuclear weapons to "tear the arm off" the Russians. $\endgroup$ – Thucydides Feb 13 '18 at 18:19
  • $\begingroup$ The link is pay walled. Googling Moscow Criterion sends me to a Tropes website... $\endgroup$ – Mazura Feb 13 '18 at 20:31
  • $\begingroup$ If link is behind paywall, please use quote formatting to put important sentence or paragraph into answer. Usually good practice anyway since links often die. $\endgroup$ – inappropriateCode Feb 13 '18 at 21:39
  • $\begingroup$ @inappropriateCode, the sentence just before the link does explain the Moscow Criterion. $\endgroup$ – o.m. Feb 13 '18 at 22:55
  • 1
    $\begingroup$ @Mazura, it is open when I click the google result, but not when I go there directly. Sorry. $\endgroup$ – o.m. Feb 13 '18 at 22:57

Nuke(s) * X = Nuclear Winter, X = 100

I think one way we can measure the minimum number of nukes required to be an effective deterrent, is as the minimum number of nukes required to cause a nuclear winter. This outcome is a sufficient deterrent and scales well vs other nations increasing technological might. The end of the world is the end of the world. Its game over no matter how strong your enemy is.

The "mess with us and we will kill everyone including ourselves" strategy is an effective one. Since no one wants to die, let alone have their entire family and everyone they ever knew die and or be exposed to a hellish wasteland life, this should be enough to make everyone involved seek alternative solutions to problems that may arise. These nukes don't even have to target enemy nations, they can be detonated on home soil and the resulting soot storm will be almost impossible to stop.

Current models place the number of nukes required to cause a nuclear winter at around 100 separate detonations.


Based on new work published in 2007 and 2008 by some of the authors of the original studies, several new hypotheses have been put forth, primarily the assessment that as few as 100 firestorms would result in a nuclear winter. However far from the hypothesis being "new", it drew the same conclusion as earlier 1980s models, which similarly regarded 100 or so city firestorms as a threat


Considering that most high tech equipment is sourced through a global logistics chain, we have the interesting situation that many nations don't actually make much of their own high tech "baseline" (there are always small, specialized facilities to ensure production of unique military goods, but these are almost hand crafted).

Going one step back, the raw materials for many types of technology is also sourced through a global logistics chain. You might be making smartphones for American customers in (insert nation x here), but where are the rare earth elements for the screens coming from?

This suggests that the primary targets are logistics hubs and ports of entry (since Russia is largely landlocked, attacking seaports is not an option in that particular case). You could go even further. Attacking naval choke points like the Strait of Malacca, the Straights of Hormuz or the Panama canal would essentially crater the entire global logistics chain. You might not even have to actually use nuclear weapons, just the credible threat that nuclear mine have been laid could close the lanes for shipping. In that case, the minimum number would be 5 (one at each end of the Panama Canal, one at each end of the Straights of Malacca and one to close the Straights of Hormuz).

enter image description here

Straight of Malacca

This global trade disruption strategy would affect China the most, since much of their raw materials and virtually all of their export trade is sent by sea. The United States would be affected as a Maritime trade power, but has the rather unique ability to be both largely self sufficient in raw materials (imported lithium is cheaper than US mined lithium due to different mining and environmental laws, for example), and has free access to both the Atlantic and Pacific oceans, so could work around some of the effects of closing major maritime choke points. Russia would be indirectly affected. Since they are a Continental rather than a Maritime power, the disruption of global maritime trade would indirectly affect their ability to access world markets, but their customers would also have far less money to purchase Russian goods. Since Russia has a very small GDP compared to that of China or the United States, this would have a considerable compound effect (they would crash harder and have a more difficult time recovering).

To directly attack Russia from a logistical viewpoint, Moscow is the rail hub of Russia, and there are a very few rail choke points in the Urals mountains which connect European Russia with the resources of Siberia through the Trans Siberian railway. Most oil, gas and telecommunications pipelines and conduits parallel the Trans Siberian Railway as well.

So these would be the minimum requirements to cripple today's superpowers.

  • $\begingroup$ Hong Kong - Sri Lanka 5300 km through strait of Malacca, 6400 through strait near Jakarta. Less than 20% difference. $\endgroup$ – Vashu Feb 13 '18 at 23:37
  • $\begingroup$ The issue is not length, but depth. Large, deep draft cargo vessels go through the Straights of Malacca because they won't run aground. $\endgroup$ – Thucydides Feb 14 '18 at 6:48
  • $\begingroup$ en.wikipedia.org/wiki/Sunda_Strait_Bridge "bridge should therefore allow for the passage of ... ships as large as the USS Enterprise" There must be a way to get large ships through the strait otherwise nobody would spend millions on modifying bridge for large ships. $\endgroup$ – Vashu Feb 14 '18 at 6:54

You might enjoy reading Nobel-prize-winning economist Thomas Schelling's book, 'The Strategy of Conflict', which explains (among other things) the theory that explains why the US and Russia have such large stockpiles of missiles.

  • The main thing stopping the Russians from nuking New York is the threat of an equally damaging counter-attack.

  • Hence the Russians' first target would be the retaliatory forces, to avoid the counter-attack (or at least, to reduce its damage).

  • Assume a counter-attack strong enough to deter the Russians needs 100 missiles, i.e. if the Russians can reduce America's stockpile to 99 missiles that would be victory for them (or as close as you can get to victory in a nuclear war)

  • Assume missiles have a 50% failure rate when fired at other missiles, which are after all in hardened underground bunkers.

  • Given the above, if America has 200 missiles, Russia need only knock out half to win, so they need only fire one missile per silo (200).

  • If America has 400 missiles, Russia needs to knock out 75%, which means firing two missiles per silo (800).

  • If America has 800 missiles, Russia needs to knock out 87.5%, which would mean firing three missiles per silo (2400)

  • In other words, the more missiles America has, the harder it is for Russia to get ahead, and it's not just linear; between 400 and 800, America added 400 missiles while Russia had to add 1,600.

  • Hence, adding more missiles to both sides makes it harder for either side to build up an advantage large enough for a surprise attack to make sense.

Needless to say, this is a pretty simplified theory (how do you fit MIRVs and submarines into it?), and it relies on certain assumptions you might find questionable - or at least, very reflective of the paranoia of the cold war where both sides publicly claimed they'd launch a first strike under the right circumstances - but this is the genuine thinking of a very influential cold war strategist. So this book would be a good read if you're interested in these things.

Needless to say, if you give your fictional world an Al-Quida-style attack with smuggled bombs, like in Tom Clancy's "The Sum of All Fears", where your attackers either avoid or don't care about retaliation, I'm sure even a single bomb could cause massive economic damage.


I've seen estimates as small as 50 for the number of nuclear missiles that China has. Wikipedia says 100 to 400, with around 250 being about the best guess. This suggests that they feel this is sufficient.

A single submarine should be enough to launch a devastating attack if every other submarine, bomber, and stationary launcher is destroyed. That's 24 missiles on an Ohio-class with 12 separately targetable warheads each, so 288 warheads.

The twenty largest combined statistical areas (urban centers, their suburbs, and related cities) in the United States have about 150 million in population. That suggests that 20 missiles would be enough to attack the US. But if anti-missile defenses are 50% accurate, that would mean that they'd need at least 40. But that understates things, as a 50% accuracy with two missiles per target means that five targets would get no missiles; five would get two missiles; and ten would get one missile each. That's only fifteen targets hit, which may not be sufficient.

Hitting the largest population centers does the most psychological damage. And remember, fallout will spread far beyond the initial blast. For example, nuclear testing in Nevada caused fallout in Vermont. Targets like Portland, San Francisco, Denver, Chicago, St. Louis, Dallas, Orlando, Washington/Baltimore, New York City, and Boston cover a lot of area with fallout.

Of course, it's possible that even one missile would be sufficient. After all, who wants to be known for allowing millions of citizens to be killed? There have been exactly two nuclear bombs (not even missiles) used in war. And a single modern missile could have done far more damage than the two combined. That has deterred everyone from using them since.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.