Wednesday, April 09, 2008

The New Toon et Al. Study on Regional Nuclear Conflict and Ozone Depletion

From the new study published Monday in The Proceedings of the national Academy of Sciences:
We use a chemistry-climate model and new estimates of smoke produced by fires in contemporary cities to calculate the impact on stratospheric ozone of a regional nuclear war between developing nuclear states involving 100 Hiroshima-size bombs exploded in cities in the northern subtropics. We find column ozone losses in excess of 20% globally, 25–45% at midlatitudes, and 50–70% at northern high latitudes persisting for 5 years, with substantial losses continuing for 5 additional years. Column ozone amounts remain near or <220> even after three years, constituting an extratropical "ozone hole." The resulting increases in UV radiation could impact the biota significantly, including serious consequences for human health.
What's the catch? The authors are using the flawed 2007 Toon et al. study as the source of their figure for soot injection into the upper atmosphere. The authors assume that ALL targeted cities will produce mass fires with efficient pyroconvective pumping, that these fires will consume basically all available flammable material in the target cities, and that rainout will only remove 20% of the soot before it reaches the upper atmosphere. These are not valid assumptions, and real-world experience with atmospheric nuclear weapons detonations does not bear them out.

There is a fairly well-developed literature on fire effects from nuclear explosions- in particular, the work of Harold S. Brode, but also the work of civil defense researchers in both the United States and Soviet Union- that the authors of this study chose to conveniently ignore. There was actually considerable controversy regarding the extent of fire effects decades ago, which is described in a 2006 book by Lynn Eden, Whole World on Fire. Brode took the pessimistic view, arguing that mass fires would occur in most circumstances following nuclear explosions of yields in excess of a few hundred kilotons. Because modern strategic nuclear warheads all exceed this threshold, he postulated that fire effects would be far greater than had been predicted earlier. The civil defense researchers, meanwhile, used real-world experience to attempt to create a set of heuristics for understanding fire effects. They studied WWII bombing experience and were able to test real buildings in atmospheric nuclear tests. This memorable 1954 film is an example of their hands-on techniques:

The civil defense researchers concluded that firestorms were an unpredictable effect, responding strongly to the contingencies of weather, topography, fuel loading, and other factors. I'm personally a bit more sympathetic to their views than I am to Brode's, but Brode was no slouch- he is one of the foremost experts on nuclear weapons effects that the world has ever produced. But even Brode's relatively extreme views are incompatible with the bizarre assumptions made in the 2007 Toon et al. study, and consequently the many studies that now use its conclusions to model the effects of regional nuclear conflict. The extremely low 20% rainout value is particularly problematic- the authors have no real justification for it other than "because we used it in 1990, and it arguably happens this way in some forest fires." I have a hard time imagining that observers in the USSBS could possibly have missed noticing that the firestorms produced by Allied bombing had injected MOST OF THE MATERIAL IN THE TARGET CITIES into the upper atmosphere- which is the assumption that the study makes. (Examine section 6.1 to understand what I mean.) However, that doesn't keep them from telling reporters that their rainout value is one of their findings, rather than one of their assumptions:
"All the new models came up with the same results," Toon said, "and they gave us two surprises: One was the huge quantity of smoke that would be produced from even the limited nuclear war in our scenario, and the other was the conclusion that the smoke would remain dense above the stratosphere for as long as five years."
Indeed, neither Hiroshima or Nagasaki appears to fit the assumptions made in the study- rainout at Hiroshima seems to have been pretty pronounced (the famous black rain), and Nagasaki failed to develop into a full-blown firestorm due to the local topography. In any case, rainout after a nuclear explosion is NOT the same as rainout resulting from a forest fire, as is attested by this 1988 study and this 1979 study of rainout following the atomic bombings in Japan. In short, Toon et al. really didn't do their research- at all.


Stephen said...

Well, a couple of points. The first being the obvious: Whether or not the ozone is severely depleted by a nuclear war, an all out multi-gigaton exchange that destroys numerous cities around the world is going to be such a catastrophic event that it really does not matter whether or not there are fires or the ozone is badly effected. Undoubtedly there social and economic consequences as well as the pollution, both from radioactive fallout and from the incineration of the cities as well as the dust kicked up and the upheaval of civilization would be almost unimaginable.

You can sit around and debate the finer points of the severity, but no matter what there's the same conclusion: It would suck big time and it's best to avoid that kind of thing.

Secondly I think the prediction is a bit overly simplistic. Yes, there could be fire storms, although I don't think they would be that bad in the less dry weathered areas and in places where the topography does not favor it. However, there would be other effects on the atmosphere, including dust and ash being thrown up and also the creation of nitrous oxides and also - Ozone. Yeah, possibly quite a lot of it.

I suppose what effect that would have on the ozone layer would have a lot to do with the area of the detonations and also, of course, the altitude.

Sovietologist said...

Oh, I fully agree that the externalities resulting even from a very small nuclear war would be enormous. At the same time, I don't think it's excusable to shamelessly produce bad science in the name of the worthy cause of preventing nuclear war.

Viktor Weisskopf supposedly said of the original 1983 nuclear winter study that "the science is terrible, but maybe the psychology is good." However, Spencer Weart made a convincing argument in his 1987 book Nuclear Fear that the psychology was actually counterproductive. The thing is, ordinary people don't think nuclear war us survivable anyway. Studies like this just reinforce their preconception of nuclear conflict as an unsurvivable, unthinkable catastrophe, forestalling meaningful change.

The important question is really "will this study encourage the governments of India and Pakistan to become less reliant on nuclear weapons?" Unfortunately, it is likely to have the opposite effect. This is likely to convince the Indians and Pakistanis that their small nuclear arsenals offer a much larger deterrent than they actually do, rather than convincing them that nuclear weapons are NOT the solution to their security problems. Instead of inspiring real change, I suspect that it will only feed into the extremely dangerous nuclear dynamic on the Indian subcontinent. I would excuse all this if the study were based on the best science available; if it had a realistic effects model based on the best efforts of the many scientists who have studied nuclear weapons effects for the last three generations, I would offer no complaint. I cannot be opposed to scientific truth, no matter how dangerous. But the study gives every indication of willful sloppiness in the service of a well-meaning but highly misguided agenda. Given what is at stake, this is inexcusable. Nuclear war is far too important to lie about.

Stephen said...

Indeed. Bad science is never justified. Well.. at least not in general and times when there is not some direct stake in national security or something. Publishing false scientific claims might be okay in a situation of direct tactical deception of abilities.. but I digress.

I'll be the first to admit that I have absolutely no idea what the finer points of the enviornmental impacts of an all-out nuclear war would be. Global warming? Global cooling? Ozone depletion? Ozone increase? I don't think any scientist can really claim to know everything that would happen on all levels.

There are too many variables and the system is just too complicated. Look at Chernobyl, for example, generally the predictions made about the long-term effects have been totally off base and that was an isolated event and not a global event in multiple locations.

There's no point anyway. The consequences would be enormous. Nobody is going to dispute that a nuclear war would cause social, economic, health and enviornmental damage on a catastrophic scale. Exactly how that would manifest, I really would prefer never to find out.

We can spend all day debating:

You'll die from the fallout.

No you won't, not if you're not right near it. It'll be the mobs from devastated areas that'll eat all the food and you'll die.

No we'll have food on hand, but it'll be bad after a year or so because we won't grow any food with the sun blocked out.

No, the sun won't be blocked out, the crops will all be destroyed by the UV from the lack of Ozone

No, it won't be either, the nitrous oxides will make the earth heat up until it's uninhabitable.

No, the upset to the earth's magnetic field will cause the atmosphere to begin to be swept off by the solar winds.

No, it will make the magnetic field stronger.

Look, honestly, does it matter which of the effects is going to suck the worst?

Ashutosh said...

It will be an achievement even if Pakistan and India realize that they currently have a good minimal deterrent and stop building more weapons, concentrating on delivery systems and more importantly, civilian nuclear power.

Sovietologist said...

My rather pessimistic answer to this question is that the likelihood of nuclear war is greater than zero. So long as there are nuclear weapons, we need to make serious efforts to understand what the effects of their use will be. I admit that the science is extremely complicated and highly imperfect; but we have to try to quantify the likely outcomes of nuclear war in order to really understand the problem.

In the case of the US, we are currently in the novel situation of being in a world where the chances of a nuclear war which will not include us is increasing. Say that India and Pakistan have a nuclear exchange. What will be the impact on America? Obviously, nothing good, and we should do everything we can to prevent it from happening; but realistically, we cannot be certain that we'll succeed in forestalling such a war. We need to develop contingency plans for situations like this, and without solid science we cannot do it.

Stephen said...

Neither India nor Pakistan has an enormous nuclear arsenal. In fact, I believe Pakistan is limited to just fission weapons. Last I heard estimates were that they have less than 100 warheads (it may be as low as 60) and that they're not expected to dramatically increase that any time soon. Their just starting to employ boosted fission warheads but I don't know if any are deployed. So it seems like you're looking at <100 warheads at maybe 15 or 20 kilotons. Also they don't have the best delivery systems, so many might never make it to the target and actually explode as planned.

India has no true thermonuclear devices. They tried once in a test but it's believed the second stage failed because the yield was less than half what was predicted. India does not have much more than Pakistan. No thermonuclear devices are deployed - their most powerful weapons may be boosted but are believed to be under 50 kilotons. Most of the warheads are likely under 30 kilotons. They have enough weapons grade material for around 100-120 weapons if used efficiently, but it's estimated that they only have about 35 battle-ready assembled weapons.

Considering the number of atmospheric tests done in the 1950's without destroying the atmosphere or driving humans to near extinction or anything remotely like that, I can't imagine that there would be much of a global problem from an India-Pakistan exchange.

Sovietologist said...

What the recent studies like the one my post is about have been arguing is that a regional nuclear war would have global environmental implications- nuclear winter, ozone depletion, etc. I don't buy that, but it doesn't mean that there won't be consequences of another kind. For instance, it has been suggested that a regional nuclear war could disrupt the global economy enough to cause a worldwide depression. I also suspect that minor powers would scramble to acquire nuclear weapons, making the international climate more precarious.

nige said...

The TAPPS (Toon et al.) studies have been wrong from day 1. In 1983 they used flawed assumptions for everything, from the absorption coefficient for sunlight by soot, to ignoring scavenging and atmospheric turbulence, etc. They also exaggerated the burnability of the fuel loading.

When a building collapses, most of the combustible material is buried under tons of debris and dust and can't burn. You don't get firestorms anymore like you did in wood-frame buildings such as those in the old, medieval part of Hamburg or Dresden, or Hiroshima and Nagasaki.

In addition, they ignored the fact that for surface bursts (unlike the air bursts over japan) the EMP deposition region will overlap the ground surface and couple thousands of amps in microsecond surges into all the electrical conductors, which would branch out throughout the city before the crater had even formed, and would blow all the fuses/circuit breakers and cut off the electricity supply to buildings, reducing the fire risk.

For a surface burst, the fireball evevation angle is such that most buildings will be "shadowed" by other buildings, preventing ignition.

In their 1989 paper, the TAPPS team failed to retract their earlier errors and apologise for hyping poorly researched trash, and instead changed the targetting assumptions to oil refineries, in an attempt to maintain some climatic effects. It was still wrong! The smoke from mass oil refinery fires doesn't hang around freezing the ground for months. It gets blown around and dispersed by atmospheric winds, turbulence, and it gets washed out by rainfall.

Another popular myth is that the entire crater volume gets converted into dust which enters the stratosphere. Actually, 99% of the apparent crater volume is due to compression of the ground and material dumped around the crater to form the crater "lip" and the ejecta zone surrounding the lip. Only 1% of the cratered mass ends up in the atmosphere, and that forms the fallout, 70% of which is deposited within 24 hours.

On the topic of ozone depletion, please notice that the prompt gamma rays from a nuclear explosion ionize the air, creating ozone. This effect seriously modifies the early-time history of the thermal pulse output, and has been intensively studied in nuclear tests (although early studies were classified).

Hence, the production of ozone-destroying nitrogen oxides in the air shock wave at high overpressures must be balanced against the production of ozone by gamma rays.

For increasing burst altitude, the amount of ozone produced by a nuclear explosion becomes greater than the nitrogen oxide ozone depletion effect, because at high altitudes the air shock wave does not reach sufficient overpressure to produce nitrogen oxides (the equilibrium concentration of nitrogen oxides is a strong function of the pressure).

Hence, high altitude bursts - which have been threatened on the West by Russian leaders due to EMP effects - will actually increase the amount of ozone in the stratosphere!

In addition, the net ozone depletion by a low altitude burst is a lot less than 1970s and 1980s predictions (ignoring the production of ozone in nuclear explosions) suggested. Much of the nitrogen oxides combine with water vapour in the fireball and you end up with nitric acid, eventually gets washed out of the atmosphere by rain and doesn't affect ozone. See:

"Nitrogen oxides, nuclear weapon testing, Concorde and stratospheric ozone" P. Goldsmith, A. F. Tuck, J. S. Foot, E. L. Simmons & R. L. Newson, published in Nature, v. 244, issue 5418, pp. 545-551, 31 August 1973:


Below is an extract from a British Civil Defence magazine article written by George R. Stanbury, head of civil defence research on the British "Operation Hurricane" nuclear bomb test at Monte Bello, and before that an expert on the incendiary bombing of Britain in World War II.

'Restricted' classified U.K. Home Office Scientific Adviser's Branch journal Fission Fragments, W. F. Greenhalgh, Editor, London, Issue Number 3, August 1962, pages 22-26:

'The fire hazard from nuclear weapons

'by G. R. Stanbury, BSc, ARCS, F.Inst.P.

'We have often been accused of underestimating the fire situation from nuclear attack. We hope to show that there is good scientific justification for the assessments we have made, and we are unrepentant in spite of the television utterances of renowned academic scientists who know little about fire. ...

'Firstly ... the collapse of buildings would snuff out any incipient fires. Air cannot get into a pile of rubble, 80% of which is incombustible anyway. This is not just guess work; it is the result of a very complete study of some 1,600 flying bomb [V1 cruise missile] incidents in London supported by a wealth of experience gained generally in the last war.

'Secondly, there is a considerable degree of shielding of one building by another in general.

'Thirdly, even when the windows of a building can "see" the fireball, and something inside is ignited, it by no means follows that a continuing and destructive fire will develop.

'The effect of shielding in a built-up area was strikingly demonstrated by the firemen of Birmingham about 10 years ago with a 144:1 scale model of a sector of their city which they built themselves; when they put a powerful lamp in the appropriate position for an air burst they found that over 50% of the buildings were completely shielded. More recently a similar study was made in Liverpool over a much larger area, not with a model, but using the very detailed information provided by fire insurance maps. The result was similar.

'It is not so easy to assess the chance of a continuing fire. A window of two square metres would let in about 10^5 calories at the 5 cal/(cm)^2 range. The heat liberated by one magnesium incendiary bomb is 30 times this and even with the incendiary bomb the chance of a continuing fire developing in a small room is only 1 in 5; in a large room it is very much less.

'Thus even if thermal radiation does fall on easily inflammable material which ignites, the chance of a continuing fire developing is still quite small. In the Birmingham and Liverpool studies, where the most generous values of fire-starting chances were used, the fraction of buildings set on fire was rarely higher than 1 in 20.

'And this is the basis of the assertion [in Nuclear Weapons] that we do not think that fire storms are likely to be started in British cities by nuclear explosions, because in each of the five raids in which fire storms occurred (four on Germany - Hamburg, Darmstadt, Kassel, Wuppertal and a "possible" in Dresden, plus Hiroshima in Japan - it may be significant that all these towns had a period of hot dry weather before the raid) the initial fire density was much nearer 1 in 2. Take Hamburg for example:

'On the night of 27/28th July 1943, by some extraordinary chance, 190 tons of bombs were dropped into one square mile of Hamburg. This square mile contained 6,000 buildings, many of which were [multistorey wooden] medieval.

'A density of greater than 70 tons/sq. mile had not been achieved before even in some of the major fire raids, and was only exceeded on a few occasions subsequently. The effect of these bombs is best shown in the following diagram, each step of which is based on sound trials and operational experience of the weapons concerned.

'102 tons of high explosive bombs dropped -> 100 fires

'88 tons of incendiary bombs dropped, of which:

'48 tons of 4 pound magnesium bombs = 27,000 bombs -> 8,000 hit buildings -> 1,600 fires

'40 tons of 30 pound gel bombs = 3,000 bombs -> 900 hit buildings -> 800 fires

'Total = 2,500 fires

'Thus almost every other building [1 in 2 buildings] was set on fire during the raid itself, and when this happens it seems that nothing can prevent the fires from joining together, engulfing the whole area and producing a fire storm (over Hamburg the column of smoke, observed from aircraft, was 1.5 miles in diameter at its base and 13,000 feet high; eyewitnesses on the ground reported that trees were uprooted by the inrushing air).

'When the density was 70 tons/square mile or less the proportion of buildings fired during the raid was about 1 in 8 or less and under these circumstances, although extensive areas were burned out, the situation was controlled, escape routes were kept open and there was no fire storm.'

nige said...

My comment about the fact that high altitude nuclear explosions produce an excess of ozone (by gamma ray emission) without producing nitrogen oxides that destroy ozone (nitrogen oxide formation requires a very compressed shock wave, which can't occur in low density air at high altitude), needs the following reference:

U.S. Congress Commission to Assess the Threat to the United States from Electromagnetic Pulse (EMP) Attack, 2004. These EMP hearings discuss the politics, such as an outrageous threat allegedly made by the Soviet Ambassador to the U.S., Vladimir Lukin, who said to the Americans in Vienna in May 1999: 'we have the ultimate ability to bring you down [by EMP from high altitude nuclear detonations]'.

Sovietologist said...

What's interesting in the new nuclear winter studies is that the authors actually admit that the 1983 TTAPS paper was wrong. But they assert that this doesn't matter because of a new effect that they "discovered" (i.e., hypothesized)- solar lofting of soot into the upper atmosphere. But even generously allowing them that this effect is real, the figure they're using for soot created by firestorms is outrageous. Because all their new papers are really about the soot and its supposed effects, their bogus methodology for calculating the amount of soot initially deposited into the upper atmosphere doesn't give me confidence that any of this stuff would actually happen. The authors' insistence that the nuclear winter controversy ended the arms race also drives me bonkers.

Your examples illustrate the reason why I sympathize with the civil defense researchers' projections of fire effects from nuclear explosions. I had never heard of this British research (I study American and Soviet civil defense, generally)- very interesting.