Nuclear Security Literacy in a Post-Cold War Age

When the Cold War ended, much of the world seemed to breathe a momentary sigh of relief. The greatest existential threat in human history, that of an all-out nuclear holocaust, appeared to have practically vanished overnight. Soon it became clear that we were not quite out of the woods yet. State-versus-state tensions between the United States and Russia did not completely dissipate. New nuclear powers, like Pakistan and North Korea, emerged publicly, and suspicions about potential future nuclear powers abound. The threat of nuclear terrorism, something which had been discussed publicly since the late 1960s, took on a new relevance in the face of threats about loose fissile material and extremist terrorism.  And the global effects of even a regional nuclear exchange, such as one in South Asia, became more acutely appreciated. 

So the nuclear threat hasn’t gone away. But it has changed. A full, Cold War exchange of massive numbers of sophisticated nuclear weapons in the near future seems unlikely.  These other nuclear threats present situations that differ significantly in character from the Cold War scenario, for better and for worse. At the same time, with the apparent recession of the existential threat, public knowledge of nuclear weapons issues has declined precipitously. The generation of Americans born sooner than 30 years ago did not grow up in an environment where discusses of the effects of nuclear weapons on cities, civil defense, and nuclear fallout were part of a general education. And the bulk of the literature on nuclear weapons is still in reference to the classical Cold War threat.

It is for this reason that many educators have felt the need to bridge the gap of nuclear literacy between the Cold War and the post-Cold War. My own contribution to this has been the development of the NUKEMAP, a nuclear weapons effects simulator, one that uses the familiar interface of Google Maps.

The NUKEMAP was developed with the idea that it is difficult to appreciate the effects of nuclear weapons in the abstract. Distances, expressed as miles or kilometers, are hard for most people (including myself) to intuitively visualize. By using a dynamic map, the user can “experiment” with the effects of nuclear weapons upon an area of the world they know intimately — such as their hometown, or the nearest metropolis.

The use of maps to visualize nuclear weapons effects, of course, is not a new thing. As a visual practice it goes back to Hiroshima and Nagasaki, the results of which were almost immediately superimposed on maps of American cities (New York City is without question the most virtually-nuked city of all time, outranking even Washington, DC, and certainly Moscow in terms of common representation in this genre). Several other online nuclear simulators using maps existed at the time I made the NUKEMAP. But none of them were as flexible or easy to use as the NUKEMAP ended up being, which perhaps explains the success of the NUKEMAP, which has had over 5 million visitors since its launch.

The technical flexibility of the NUKEMAP allows one to easily contrast the Cold War with the post-Cold War threat environment. This is because it not only models the standard prompt effects of nuclear weapons — such as blast pressure, thermal radiation, and ionizing (nuclear) radiation — but because it allows complete flexibility in terms of burst height, models radioactive fallout, and can give rough casualty and fatality estimates by applying the modeled effects to an underlying global population density database. The nuclear effects codes, including the fallout and casualty estimates, are all based on declassified effects codes developed during the Cold War for civil defense purposes.

So, as an example, one could easily model a “typical” Cold War-style nuclear attack against the United States. Let us imagine that we are curious about the effects of a Soviet weapon from the Cuban Missile Crisis period, the R-12 (SS-4), targeted against a “hard” nuclear target, like the bunkers at Dyess Air Force Base near Abilene, Texas. On the NUKEMAP, we can actually select the R-12 from a list of preset nuclear weapons, which automatically plugs its yield (2.3 megatons) and its fission fraction (estimated at 50% fission) into the settings. I can then choose whether it is an airburst or a surface burst. In this case, since it is a “hard” target, the goal is to put as much overpressure onto the target itself as possible, so I choose the surface burst option. The result is visualized as such:

With the right options selected, we can see that if done with a modern population density database, the immediate fatalities of such an attack are estimated to “only” be around 14,000 people, with injuries around 44,000, judging by modern population levels. However we can also see that the fallout plume from such a burst would extend several hundred miles, and contain some tens of thousands of square miles of area — including much of the Dallas-Fort Worth metro area, depending on the wind direction and speed. Such is the consequence of a “hard” attack: by detonating at ground level, considerable local fallout will be achieved, contaminating significant areas downwind. This has the benefit of showing graphically that even a counterforce attack with such a weapon (whose size is in part a reflection of the lack of accuracy of the weapons) would contaminate significant amounts of civilians downwind.

What about a Cold War era strategic attack against a city? It is easy to change the location — in this case, to Atlanta, Georgia, which was in range of the R-12s being deployed on Cuba. In this instance, we might assume that if possible they would use an airburst attack, because it would impact a larger area with pressures more amenable to destroying a “soft” target like a city. The reasons for this change of area are because for lower blast pressures (e.g. around 5-10 pounds per square inch, which will destroy most civilian infrastructure, as opposed to the +300 pounds per square inch required to destroy hardened infrastructure) the ideal burst height is considerably higher, largely because the shockwave from the bomb reflects off of the ground and re-combines with itself, amplifying the maximum distance of destruction. (It is for this reason, and not because of fallout concerns, that the Hiroshima and Nagasaki bombs were airbursts.)

On the NUKEMAP, we can set “airburst” as an option and then choose what height we want for the detonation. For our purposes it is easiest to have it use the optimal blast height for maximizing the 5 pounds per square inch (psi) pressure area, though we can set it to an arbitrary pressure rating if we want to (these, and many other settings, are under “advanced options”). For a 2.3 megaton warhead, this is, according to the Cold War nuclear effects library that the NUKEMAP’s calculations are based on, just over 13,500 ft (around 2.5 miles). Detonating at this height, the 5 psi pressure range extends to an astounding 11 mile diameter, covering over a hundred square miles of area. The thermal radiation radius that would produce third-degree burns covers a whopping 370 square miles. It would kill some 330,000 people outright, and injure another 678,000. 

Even if we have chosen for fallout to be displayed, however, none will appear with such a setting. Why? Because for local fallout to be a concern, there has to be significant mixing of the nuclear fireball with dirt and debris, and at a height of 2.5 miles, this will not occur, even for such a high megaton weapon. In fact, any burst height above 0.75 miles will produce only negligible local fallout. Such is the tradeoff of Cold War weapons: aimed at cities, they could kill thousands to millions, but the contamination was relatively limited; aimed at hardened counterforce targets, they immediately kill far lower numbers of people, but their areas of contamination stretch for thousands of square miles.

What about the post-Cold War situation? In the case of nuclear terrorism, we can use the NUKEMAP to derive a different point. We might imagine that an improvised high-enriched uranium device could probably get within 10 kilotons of yield without too much technical sophistication. This is several orders of magnitude less explosive yield than our hypothetical Soviet warhead (which is not the maximum yield possible in the Cold War in the least; the US fielded weapons in the 25 megaton range, and the USSR developed weapons in the 50-100 megaton range). However, since a terrorist group would not likely have the ability to accurately put such a weapon at a useful airburst height, and likely would be interested in the psychological advantages of long-term contamination, this would be a surface shot against a city.

Modeled against mid-town Manhattan (that “classic” nuclear target), the NUKEMAP gives the following grim assessment. The weapon effects of a 10 kiloton bomb would themselves be relatively geographically limited, because of the yield. (They would probably be somewhat smaller even than the NUKEMAP’s codes give, because of the focusing effect of the heavy buildings, but at the moment the code cannot take this into account.) The fatalities, though, would be astounding, because of the high population density in the area: 240,000 dead and 344,000 wounded. Furthermore, the fallout would be significantly contaminating, because it is a surface burst. 

So we can see how the NUKEMAP lets us intuitively appreciate the difference in these situations. The terrorist threat involves lower-yield weapons, but surface bursts likely targeted at extremely “soft” targets. This means that their potential for casualties is far less than a strategic Cold War missile, but still more devastating than any other type of terrorist attack yet committed. Their contaminating power is considerably less than a multi-megaton Cold War exchange, but for an individual target could be devastating.

One can take issue with these simulation choices (for example, as noted, a Soviet attack against Washington would probably involve huge megatonnage and surface bursts). But that’s why the NUKEMAP exists, to allow easy manipulation and comparison of such outcomes. One need not conduct a long and length study to get back-of-the-envelope results, and in fact one can use the NUKEMAP to model quite complex scenarios (including multi-weapon attacks).

As alluded to before, the NUKEMAP has proven to be an immensely popular tool, attracting several million users. I also developed secondary version, known as NUKEMAP3D, was allowed a three-dimensional modeling of mushroom clouds, which has proven even more effective at communicating the size of small-yield nuclear detonations than the two-dimensional NUKEMAP. It is hard to visualize a 20,000 foot mushroom cloud without such an aid, because it is a scale of destruction that we do not interact with on a regular basis. Here is the maximum size of a 10 kiloton bomb on Boston (to spare New York for a change), as it would be viewed from an airplane coming into land at Logan Airport. 

For both of these tools, I have been monitoring the way in which they are discussed online, and the places where they are deployed. NUKEMAP appears to occupy a zone between the “fun” and the “scary,” which has traditionally had tremendous mass appeal (think horror movies and roller coasters). Because it is inherently interactive, it encourages experimentation and “play.” And judging from the way it has been used by news organizations and online debaters to visualize nuclear scenarios (ranging from North Korean nuclear attacks to re-creating the effects of nuclear accidents from the 1960s), it appears to have been an effective addition to the present discourse on nuclear arms, for both experts, media, and lay audiences alike.