AAAS Center for Science, Technology and Security Policy organized a small series of meetings at which Ted Postol and I discussed the U.S. plans to put conventional warheads on some of the Trident submarine missiles, which are part of the Prompt Global Strike capability sought by U.S. Strategic Command. I thought that the discussion was very useful and will try to summarize some of my points here.

First, the good news. Congress did not approve the Pentagon request that would have allowed it to go ahead with the project and there are people in Congress who are skeptical of the idea. The bad news is that the concept is still very much alive, fairly popular in some quarters, and the DoD got money to study the concept further. I'm sure they will come back for more.

There are many problems with the "prompt global strike" concept (Ted Postol made some very good points about feasibility of the plan), but I will talk primarily about the risks associated with the plan that mixes nuclear and conventionally-armed missiles on the same submarines. Of course, we can hope that in normal circumstances no one would mistake a single Trident missile (why does it have to be a single missile, by the way?) that would be launched to "promptly strike a target of opportunity" for a beginning of a full-scale nuclear attack. Or maybe they would. The reality is that we don't really know that, just as we don't know in what circumstances this kind of launch may occur and how "normal" they might be. One thing is clear, though - deployment of conventionally-armed missiles alongside with nuclear ones can only increase the probability of misunderstanding or misinterpretation.

When asked about the risks, proponents of the Trident conventionalization present a sort of a multilayered defense in support of the idea. The arguments that are supposed to address the risk issue go roughly like that:

"The Russians won't see the missile"

"If they see it, they will be able to determine that the missile is not coming their way"

"We will notify them"

"Even if everything fails, the Russians will not do anything because they know that we are not after them."

Let's take a look at these arguments. First, "Russia won't see it". The idea here, as I understand it, is that since the Russian early-warning system does not have global coverage, Russia would not even know that a missile was launched, let alone mistake it for anything. Well, not necessarily. While the Russian early-warning system is quite far from its prime, it appears that it's not dead either. The Cosmos-2379 satellite, deployed in geosynchronous orbit, appears to have the capability to detect SLBM launches from North Atlantic (in my recent article I have a map that shows potential coverage provided by this satellite). Apparently, STARTCOM has been thinking about moving submarine patrol area to the Indian Ocean (out of sight of the Russian satellites and radars or just closer to potential targets?), but this may prove to be only a short-term solution. Russia has been paying more attention to its space-based early-warning system recently and it is quite likely that the plans to deploy the second-generation US-KMO satellites will materialize at some point. This won't happen this year or the next year, but if we are thinking longer term, it is certainly a possibility.

As for "they will determine that the missile is not coming their way" argument, I don't find it convincing. This certainly requires some additional analysis, but it is clear that a trajectory of a missile can be determined with limited accuracy, especially at the early stages of flight and especially from space. The accuracy can be improved, but it takes time. Besides, given that some "prompt global strike" missiles can aim at targets fairly close to Russian borders, it is not clear if that determination can be reliably made.

Now the notifications. The problem here is that the only notification mechanism that Russia and the United States have today is singularly unsuitable to the prompt-strike scenarios. The 1988 Ballistic Missile Launch Notification Agreement requires 24-hour notification of an upcoming launch, which doesn't fit into the 60-minute "prompt strike" timeline. Of course, that agreement, which was folded into the START I Treaty in 1991, could expire with the treaty in 2009, but this only means that we will have no notification arrangement at all. If it doesn't expire, it would have to be modified beyond recognition to allow for the notification required by the global strike.

Another arrangement, the Joint Data Exchange Center (JDEC), which could potentially provide the United States and Russia with the capability to notify each other on a very short notice, is showing no signs of life, despite General Cartwright's apparent belief to the contrary. But even if it were, it is not clear if JDEC is the right arrangement for this kind of job - putting too much faith in it may in fact increase the probability of an accident (I try to develop this argument in my "accidental launch" article).

As a bit of a good news, General Cartwright apparently discussed the matter with General Baluyevski, the Chief of Russian General Staff (at least this is what he tells Congress). But we are yet to see anything emerging from these discussions. Besides, there is a serious problem there - the kind of measures that would be necessary to find an adequate solution to the notification problem might be far beyond of what is possible today in the U.S.-Russian relations.

Which brings us to the last point - whether Russia would believe, of only for a brief moment, that the United States might attack it. Normally, the answer is no. After all, the cold war has been long over and there has been no shortage of declarations of partnership between Russia and the United States. However, we should not overestimate the ability of the militaries to change and to adjust their operational practices and plans to the new realities. The strategic weapon systems that they operate were build with cold-war missions in mind and it is only natural that they impose cold-war thinking on their operators.

Here are some examples. One of the fighter pilots who were scrambled into the air on September 11, 2001 was reported to testify that

"I reverted to the Russian threat - I'm thinking cruise missile threat from the sea. You know you look down and see the Pentagon burning and I thought the bastards snuck one by us."
If on September 10, 2001 someone would suggest that a U.S. pilot would assume that Russia might attack the United States, that person would have been laughed out of the room. But this is exactly what happened. Two more "coincidences" of that day - NORAD was scheduled to conduct an exercise, known as Vigilant Guardian, "which postulated a bomber attack from the former Soviet Union" (look for Note 116 in the 9-11 Commission Report), while Russian strategic bombers were indeed conducting an exercise that involved flights in the direction of the United States. As far as we know, NORAD never began the exercise that day and the Russian military grounded the bombers as soon as they learned about the events in the United States, but the number of coincidences is quite alarming.

Not that there are any signs that the military on both sides have changed their plans and no longer practice attacking each other. Just recently Russia conducted a large-scale exercise of its strategic bombers, in which they got close enough to the United States to be intercepted by NORAD fighter planes. The United States also routinely conduct exercises that involve a nuclear exchange with Russia. One of these, Vigilant Shield 07, described by Bill Arkin in his Early Warning blog just yesterday, has steps like these in its scenario (Ruebek is Russia):

* Ruebek Conducts Limited Strategic Attack on United States
• Wave 1 - 8 x Bear H Defense Suppression w/CALCM
• Wave 2 - Limited ICBM & SLBM Attack
- 2 x ICBM Launched (1 impacts CMOC [Cheyenne Mountain], 1 malfunctions)
- 2 x SLBM Launched Pierside (1 impacts SITE-R ["Raven Rock" bunker on the Maryland-Pennsylvania border], 1 malfunctions)
- 3 x Bear H from Dispersal Bases w/ALCM (Eielson AFB, CANR, Cold Lake)
* US Conducts Limited Retaliatory Attack on Ruebek

Bill Arkin rightly points out that these scenarios (see also an earlier one in which Russia was Slomonia) have very little to do with the reality. But they help keep alive the idea that a nuclear exchange (even if a "limited" one) between Russia and the United States is possible. Unfortunately, ideas like this are likely to be around for a while - the inertia of the system is enormous.

It is true that the United States and Russia have made significant progress in overcoming the inertia of the cold war mentality. But this process has slowed down in the last few years and even shows signs of being reversed. The United States has grown indifferent to Russia, believing that it's no longer a problem, while Russia is becoming increasingly suspicious (and sometimes paranoid) about U.S. intentions and policies. This is not a healthy situation, which is largely a result of the slow disappearance of the network of contacts and cooperative programs between Russia and the United States. Some programs are still holding on, but the cooperation field is shrinking in almost all areas. It is not hard to guess what will take its place if this filed is allowed to disappear completely - the "limited strategic attack scenarios".

The bottom line, I guess, is that it's not that Russia and the United States should check their every step to see if its fits into a cold war-type deterrence relationship. Certainly not. This would be a sure way to get locked in that kind of relationship for a long time. But I believe it is dangerous to simply dismiss the machinery of strategic nuclear forces that was built during the cold war as if it is not there. Which, unfortunately, is exactly what the "prompt global strike" proposal does.