[mikeash]

What kind of attack could take out ICBMs but would be unable to take out bombers? You say that bombers on alert could launch in less time than it takes for an incoming weapon to destroy the base, but surely the same is true of ICBMs.

Adding bomber bases to the list of targets the enemy must destroy makes their life more difficult, but it seems like the difference would be small. There are, what, half a dozen or so bomber bases you'd have to destroy? One warhead each should do it. My understanding is that ICBM silos are sturdy enough to require a nearly direct hit to destroy them, so you'd need hundreds of warheads to destroy those.

The ability to launch bombers and then change (or cancel) their mission en route is super useful, but that seems separate from the alert role. If you want to put nuclear bombers in the air to put the fear of God into your adversary, you don't need them to be on a five-minute alert (or whatever the actual time will be).

[ChuckMcM]

Since you've read up on this your aware that at the height of the cold war both the Soviet Union and the US possessed enough strategic weapons to obliterate the other country. Further, the premptive use of such weapons would leave the other country unable to respond. That lead to a dangerously unstable system that more than once nearly triggered WWIII.

In an effort to return to sanity both countries agreed to bilateral reductions in strategic arms. The point of which was to reduce the number of warheads on strategic weapons (ICBMs) to make such a pre-emptive strike unlikely to succeed. The next step of moving back from the brink of annihilation was to limit the number of warheads in submarines, bombers and medium range missiles. By the end of the cold war both sides had reduced (and verified) that their respective stocks of nuclear warheads was at a level such that neither side was able to preemptively destroy the other, and entering into a conflict could be more controlled/restrained because there would be significant and incontrovertible evidence of first use while the victim would retain significant response capability. These were perhaps the most important treaties negotiated and executed in the 20th century. And have generally kept the chains on that horrific capability.

If New START is not renewed, and we see a build up in nuclear arms again, then it is possible that we see an adversary (or allied adversaries) who get to the point that they can take out all of the strategic capability of the US. If that is likely to occur, then the first step to combat it is to resurrect the bomber force requiring your adversary to expend more warheads to counter all more potential delivery systems. And the typical hawkish policy is to try to keep the number of delivery systems and warheads ahead of your worst cast threat.

In any event, it is the "wrong direction" we need to go if we wish to avoid killing everyone.

[mikeash]

So the idea of putting bombers on alert is to make it more difficult for the enemy to carry out a successful preemptive strike. But how? Bombers aren't harder to destroy, and they're clustered together, so the number of additional warheads needed is small. Is it just about requiring the enemy to use 1-2% more warheads in a preemptive strike, or is there something else to it?

[ChuckMcM]

To create a context, lets assign some numbers to the variables. No doubt the actual numbers would be classified.

Lets say it takes 8 warheads to reliably disable an ICBM silo and you have 80 silos, so that take 160 war heads for those silos. And lets say it takes 25 warheads to cover all the possible areas a mobile ICBM launcher might be in its various revetments and launch positions. So 185 war heads.

Now you add 15 SAC bases with bombers, each takes say 3 war heads to be sure you get it done, that is another 45 warheads. So now 230 warheads are needed. And you want to be sure and take out norad and the line of early warning radars, so call it 250 war heads total with delivery platforms.

If you limit each side to 125 strategic warheads then in the event of a first strike the adversary has to pick which places they can hit because they don't have enough warheads to take out everything.

So the idea of putting bombers on alert, is to increase the number of places you have to hit in order to eliminate a retaliation capability. That either helps reduce a threat of a preemptive strike (they can't be sure what they can kill) because it increases the number of warheads the adversary needs to insure success. Or it forces them to make more warheads in violation of the treaty which would be caught sooner by the verification tools in place.

Conversely, if your adversary does start increasing their warhead stock pile and you don't reactivate your SAC bases, then you open a window in time where a preemptive strike might be effective and end up with itchy trigger fingers because you think their MIGHT be a preemptive strike coming over the horizon and you don't respond fast enough you never will be able to respond. It is that situation, the so called "launch on warning" scenario with weapons that cannot be recalled, can turn a glitch into a war that destroys the planet.

And no the PRK is not a threat in this way as they don't have the ability to threaten all of the US and materially limit our response.

So the bottom line is that while you certainly could use strategic weapons to destroy SAC bomber bases, the major powers have artificially disabled their ability to do so (without leaving the ability to respond intact) through treaties to insure global stability.

[mikeash]

It makes sense with those numbers, but those aren’t anything like the real numbers. The real numbers for the US are more like 450 silos and 5 bomber bases. So you’re looking at about 1,350 warheads to take out the silos (assuming three each) and 15 to take out the bombers (again assuming three each).

You could expand the number of bases, but that’s tremendously expensive. It’s far cheaper to build a missile silo than a bomber base and a fleet of bombers to base there.

So in terms of the real numbers, you’re talking about adding around 1-2% to the number of warheads needed for a successful preemptive strike. Again, is that’s all there is to it, or is it something else?

[ChuckMcM]

I think I mentioned they were made up numbers, they were illustrative :-) not definitive. The actual numbers are probably in the treaty documents somewhere.

As for the relative cost of building a missile silo or reactivating a SAC base, I'd be interested in how you get to the notion that building a missile base would be less expensive. They were pretty complex facilities, and there are some great videos on youtube about them. An airbase is a bunch of asphalt spread out, a bunker to hold munitions, and a barracks to hold the pilots. I will be the first person to say I have not built (or even participated in building) either kind of facility but when I imagine how I might build one or the other, the fact that the SAC base is all above ground makes me believe that it would be cheaper to build. Could totally be wrong on that of course.

[mikeash]

I understood that the numbers were made up, but this is a case where they need to be accurate, at least in their proportion, for the result to make sense.

As for relative cost, that’s a big reason why ICBMs took on such a prominent role: once the technology was worked out, they were much cheaper than bombers.

It’s hard to find precise numbers, but imprecise ones should suffice. We can look at commercial airports to get an idea of what base construction would cost. An asphalt runway won’t cut it here: you need sturdy concrete runways to hold large planes like this. I believe a B-52 needs even more strength than airliners do, because they have fewer wheels. In any case, the cost of constructing a single new runway at a large commercial airport is many hundreds of millions of dollars, sometimes over a billion. And that’s just the runway, never mind all the other infrastructure you need. Bombers aren’t cheap either. It’s hard to say what a new B-52 would cost, since nobody has built one for a long time, but it’s peobably hundreds of millions each.

An ICBM plus silo is tens of millions. Various sources online list the Minutemen III at a $7 million cost per unit, but that must be nominal dollars decades back. They’d cost quite a bit more now, but tens of millions should be a reasonable estimate. The silo adds significant cost, but it’s ultimately a fairly small construction projects relative to a whole bomber base.

If you can reactivate old bases, that would help a lot, but they probably need substantial refurbishment if they’ve been idle for a long time, and you still need aircraft to base there. And there are only so many old bases to be had.

[cameldrv]

Accurate SLBMs, possibly on depressed trajectories. Russia didn't have these in numbers at the end of the cold war, now they do. They have even added GLONASS guidance to their ballistic missiles, which is worrying, because it means that they care about accuracy while GLONASS is still working, i.e. in a counterforce first strike.

Given the time it may take to validate and characterize a launch warning, have the message reach the President, have him make a decision, relay the message through STRATCOM, retarget the missiles (assuming we're still detargeted), and launch, you're talking at least 10-15 minutes from warning to launch. The Russians have missiles that might arrive faster than that. If bombers on the ground are maintained at a high enough level of alert, they can be launched at the very first warning as a precaution, before any irrevocable decision has been made.

[mikeash]

That makes some sense. Do the numbers work out, though? You could launch bombers on a less certain warning, but they'll take longer to get far enough away to be safe.