The Big Bangs !
During my periodic browsing of your very informative website, I recently came across a fascinating almost "throwaway" line from Steve Hemphill concerning the alleged cause of the apparent global cooling which occurred according to the surface temperature record between 1940 and 1976. As this cooling runs counter to both commonsense expectation (at least if one assumes that increasing greenhouse gas concentrations should result in global warming) and most of the computer climate models, I have always regarded this mid-century cooling as one of the more convincing reasons for adopting a sceptical/contrarian attitude towards the claims of global warming proponents. In his email of 14 February, Steve Hemphill said:
"What evidence do you have that the 1930's were as warm as the 1990's? Everything I've seen points to a pretty constant temperature until W.W.II, then a little cooling until after the test ban treaty when we stopped throwing millions of tons of atomized dirt (no pun intended) into the stratosphere. Since then we've way surpassed the 1930's in warmth according to everything I've seen."
Although it is a bit oblique, what Steve seems to be asserting is that the apparent global cooling shown by the surface record between 1940 and 1976 may actually be caused by the cooling effect of "millions of tons of atomized dirt (thrown) into the stratosphere" by nuclear tests conducted during that period. My initial attitude towards this fairly startling claim was one of extreme scepticism. I had not previously read about such a theory (although I must confess that my reading on topics greenhouse is somewhat less than encyclopedic). Nevertheless, I found Steve's suggestion intriguing (if somewhat improbable), and so I set out to research the question. Much to my surprise, I found that my initial response flowed mostly from my own ignorance. I embarked on some web-based research on the extent and effects of global nuclear testing, and I found some very useful material.
The Coalition to Reduce Nuclear Dangers at http://www.clw.org/coalition/ctfallou.htm contains numerous graphs and tables setting out information about nuclear testing. See especially Bar Chart of Atmospheric and Underground Tests by Country (1945-98) at http://www.clw.org/coalition/chartworldtests.pdf and Known Nuclear Tests Worldwide, 1945-1996 at http://www.nrdc.org/nrdc/nrdcpro/nudb/datab15.html and The Nuclear Testing Tally at http://www.armscontrol.org/ACT/may98/ffmy98.htm While this material gives a lot of information, it does not tell us the size of any given test i.e. whether it was larger than 1 megaton or not. I was only initially able to find information of that degree of specificity about the US nuclear testing programme. An amazingly comprehensive set of data about the US programme are available at the US Department of Energy's website at http://www.nv.doe.gov/news&pubs/publications/historyreports/default.htm See especially http://www.nv.doe.gov/news&pubs/publications/historyreports/usnuctests.htm which contains links to huge amounts of data on every US test conducted between 1945 and 1992. Unfortunately, other countries do not seem to be quite so free with their data.
At the outset of my research, I was advised by John Daly that his understanding (from previous reading some years ago) was that a nuclear weapon needed to be at least 1 megaton in size to be powerful enough to propel material into the stratosphere. I was eventually able to verify this information from the FAQ section of the Federation of American Scientists website (see end of article for main link to this excellent site). The FAQ's can be found at http://www.fas.org/nuke/hew/Nwfaq/Nfaq0.html. The author says:
'If the explosion is a true air-burst (the fireball does not touch the ground), when the vaporized radioactive products cool enough to condense and solidify, they will do so to form microscopic particles. These particles are mostly lifted high into the atmosphere by the rising fireball, although significant amounts are deposited in the lower atmosphere by mixing that occurs due to convective circulation within the fireball. The larger the explosion, the higher and faster the fallout is lofted, and the smaller the proportion that is deposited in the lower atmosphere. For explosions with yields of 100 kt or less, the fireball does not rise abve the troposphere where precipitation occurs. All of this fallout will thus be brought to the ground by weather processes within months at most (usually much faster). In the megaton range, the fireball rises so high that it enters the stratosphere. The stratosphere is dry, and no weather processes exist there to bring fallout down quickly. Small fallout particles will descend over a period of months or years. Such long-delayed fallout has lost most of its hazard by the time it comes down, and will be distributed on a global scale. As yields increase above 100 kt, progressively more and more of the total fallout is injected into the stratosphere.'
Almost all the Russian tests of 1961-62 were airbursts, and many were above 1 megaton, so it is certain that vast amounts of fine stratospheric dust were generated, with consequent major effects on weather. In addition, the height of the tropopause varies from around 18 kilometres in equatorial regions to a mere 8 kilometres in arctic latitudes. More than half of the huge Russian test series of 1961-62 (see later) took place at its Novaya Zemlya testsite on the edge of the Arctic Circle, including the largest weapon ever exploded by any country, on 30 October 1961. Its power has been variously reported at between 50 and 63 megatons. Because of the reduced altitude of the tropopause in Arctic regions, it is likely that blasts much smaller than 1 megaton would have had a significant effect on the stratosphere (and that larger explosions would have had an absolutely enormous effect) . To explain this assertion, it has been reported that one of the earlier US atomic (i.e. non-thermonuclear) tests, codenamed "Able", in June 1946, which was a mere 21 kilotons, had a mushroom cloud some 30,000-40,000 feet high (9-12 kilometres). Although that is a long way short of the stratosphere in equatorial regions (where the test actually took place), it would have penetrated well into the stratosphere had it occurred in arctic latitudes.
Another basic fact we need to know, before we embark on a detailed examination of nuclear testing effects on global climate, is exactly what effects dust/debris in the stratosphere can be expected to have. `Fine' aerosol particles with sizes between about 0.1 and 1 micrometre can influence climate in the troposphere in two ways. Under clear sky they scatter and absorb solar radiation; some of the scattered sunlight goes back to space (the direct effect), thus resulting in cooling. The second way is that they may act as cloud condensation nuclei, enhancing reflectivity and life-time of clouds (indirect effect), thus leading to additional cooling. Given the lack of clouds in the stratosphere, the second effect is not really relevant to stratospheric 'fine' aerosol particles. Globally averaged, fine human-made tropospheric aerosols may currently cancel about 50 % of the warming effect of human-made greenhouse gases, according to the IPCC. So far, though, the uncertainty range is large, stretching from roughly 10 to 100 %. [IPCC 94, sections 3, 4.4, 4.7] [IPCC 95, sections 2.3, 2.4.2] [Schwartz]. Given such a very large assumed tropospheric effect for 'fine aerosol particles (and also remembering that the eruption of Mount Pinatubo in 1991 resulted in a cooling of around 0.5°C for at least 2 years, directly as a result of stratospheric debris), one can reasonably expect that there would have been a very significant cooling effect on global climate if the world's superpowers were so irresponsible as to test significant numbers of large nuclear weapons in the atmosphere. As we are about to see, that is exactly what they did, especially during the 1950's and early 60's. The results of my initial research into the extent of nuclear testing were as follows: ...
1. There were no nuclear tests anywhere in the world of anywhere near 1 megaton in size until 31 October 1952 when the US exploded its first thermo-nuclear device codenamed "Mike" at Enewetok in the Marshall Islands in the central Pacific. Nor were there any tests of any size anywhere near the Arctic circle before 1952. Thus the record of global cooling from 1940 to 1952 cannot by any stretch of the imagination be ascribed to nuclear testing. But from 1952 onwards it is quite a different story. "Mike" was a 10.4 megatons blast and was a surface test, thus propelling large amounts of dust into the stratosphere fairly near the equator (rather like Pinatubo in that respect). "Mike" is reported to have had a mushroom cloud some 32 kilometres high and 160 kilometres wide.
The tropopause in equatorial regions is at around 18 kilometres, and so "Mike" clearly must have had a very significant effect on stratospheric fine aerosol concentrations. By way of comparison, the Mount Pinatubo volcanic eruption in 1991, which is generally agreed to have reduced global temperatures by around 0.5°C for at least 2 years, had an eruption cloud of around 500 kilometres in diameter. "Mike" was followed by "Bravo", an even larger 15 megaton blast at Bikini Atoll (also in the Marshall Islands) on 28 February 1954. It was also a surface blast and must have had an even more significant effect on stratospheric dust levels (and thus on global temperatures).
It appears that the nature of stratospheric wind circulations is such that dust thrown into the stratosphere tends to circulate around the entire globe, rather than being confined to the latitudes where it was generated. It is reasonable to postulate that these two tests had a measurable cooling effect on global temperatures between (say) 1952 and 1956 (assuming that they continued to affect surface temperatures for approximately the same time period as Pinatubo appears to have done).
2. Between 1954 and 1962 the US conducted a total of 27 atmospheric tests of greater than 1 megaton size, and 2 of sub-megaton (i.e. almost a megaton) size. However, 14 of these were conducted at sea with the weapon fired from a barge moored in deep water. It appears reasonable to assume that little dust would have been generated by these tests, though lots of water vapour (however, see later for details of a particular US test in 1958 which appears to have been an exception to this proposition). I would be interested in expert views, but I suspect that super-heated radio-active water vapour would precipitate very quickly and therefore only have transient local effects on weather.
Another 5 of these above 1 megaton tests were conducted at high altitude (greater than 43 kilometres) with the weapon being delivered by rocket. There would appear to be no way that such tests could generate any stratospheric dust at all, and thus there could be little or no effect on global weather. Finally, nine (9) of these megaton-plus tests were conducted by airdrop (from aircraft). Those tests occurred during 1962 (in the runup to the US and USSR signing the Limited Test Ban Treaty on August 5, 1963). These 9 tests involved dropping thermo-nuclear devices over the ocean near Christmas Island (apparently, this is the Christmas Island in the south Pacific, not the one in the Indian Ocean) and Johnston Island in the Pacific near Hawaii. To the extent that these 9 tests could be characterised as "true airbursts" (see earlier), it is likely that they generated significant amounts of microscopic stratospheric debris, and therefore had significant climate effects.
3. As mentioned above, the United States and the former Soviet Union signed the Limited Test Ban Treaty which effectively banned testing of nuclear weapons in the atmosphere, the oceans, and space on August 5, 1963. After that time neither the US nor the USSR conducted any tests at all in the atmosphere: all were underground. Although there are documented instances of radiation leakage from some undergound test sites, there is no suggestion that any of them generated any airborne dust or debris at all.
I then discovered a fascinating book-length (487 pages) document on the U.S. Department of Energy WebSite, which provided a wealth of additional detail about the US nuclear testing programme, and also quite a bit of information about the Russian programme. The article is by one William E. Ogle, a senior member of the U.S. nuclear testing team from the early 1950s through to his death in 1984. The article goes by the most unromantic name DOE/NV 291. It is an account of the return to nuclear weapons testing by the United States after the test moratorium 1958-1961. It is said to be the only detailed account by an "insider" of the United States nuclear testing programme. I found it at http://www.nv.doe.gov/news&pubs/publications/historyreports/nvo291/default.htm .
Of course, Ogle's article does not expressly canvas the effect of the nuclear testing program on global temperatures, probably because this was not a matter of wide concern in 1983 when the article was written. Nevertheless, it contains considerable factual material enabling various significant inferences to be drawn. The information in the article filled in a lot of facts I had not previously known.
For example, the USSR exploded its first H bomb in 1953, just 12 months after the first U.S. test (code-named "Mike") on 31 October 1952. However, the Soviet test program did not really get into full swing until 1956, 1957 and 1958, when they conducted 9, 16 and 34 tests respectively (all Soviet tests up until 1961 were atmospheric). Thus, I think one can reasonably conclude that the combination of specific "dirty" U.S. tests (namely "Mike" in 1952, "Bravo" in 1954 and "Oak" in June 1958 -- see later) and the Soviet tests, especially in 1956-58 and 1961-62, may well have meant in aggregate that there was a measurable cooling effect on global surface temperatures for most of the period 1952-63. However, there was a complete moratorium on atmospheric testing by both the U.S. and USSR between November 1958 and September 1961.
The precise chronological extent of the cooling effect of course depends upon the persistence of the resulting atomic dust in the stratosphere, and the precise geographical extent of the effect depends on stratospheric circulation (a subject on which I understand a great deal of work is presently being done). For present purposes, I assume that there would be an ongoing cooling effect for at least 12-18 months after an explosive event (whether volcanic or nuclear) which propels significant amounts of material into the stratosphere. This accords with what appear to be the generally accepted parameters for the Pinatubo explosion. However, it is conceivable that extremely large amounts of stratospheric dust might persist for longer, with a correspondingly greater and longer cooling effect on temperature. It is also conceivable that larger blasts generate a higher proportion of fine aerosols, which may persist in the stratosphere for even longer than larger diameter particles. As we will see below, Soviet testing during 1961 and 1962 almost certainly propelled vast amounts of dust into the stratosphere (probably significantly more than the Pinatubo eruption). I extracted the following highlights (or rather lowlights) from Ogle's article:
1. Ogle reported in relation to the first U.S. thermo nuclear test (Mike in 1952) that "debris was carried up and dispersed over a much larger area than was thought possible" and "the large crater produced by Mike shot in Operation Ivy made it obvious that the Marshall Islands could not support a long series of high-yield shots fired in that manner, with the inevitable destruction of Marshallese homelands."
2. Ogle also reported Atomic Energy Commission Health and Safety Office representative Merrill Eisenbird as saying that they had been unable to find the main body of fallout from the "Mike" blast. Eisenbird is reported as saying: "Whether it all shot into the stratosphere or not is one of the mysteries of the nuclear age."
3. The next 1 megaton- plus blast was code-named "Bravo" in February 1954. Ogle reported that the test team had expected this blast to be around 6 megatons, but they were shocked when it turned out to be between 15 and 16 megatons! Clearly there was a lot they did not know about thermo nuclear explosions at this time. The Bravo blast produced fallout that extended to the islands of Rongelap and Utirik, where there were significant native populations, and to Rongerik, where a Nuclear Task Force weather station was sited. Fallout was also dumped on a Japanese fishing trawler.
4. At around the same time, radioactive rain was reported falling on Japan in the wake of a Soviet nuclear test in Siberia. As Ogle reported, "These incidents, plus an increasing study of the quantity of radioactive material in the atmosphere and its possible effects, began to produce a move on the part of other nations to exert pressure on the U.S. and Russia to stop testing." Public demands to stop testing came from sources as diverse as the Pope, Albert Schweitzer and President Nehru of India. Nevertheless, the two superpowers were able to resist this pressure for another four years.
5. However, the U.S. responded by stopping surface blasts and moving to blasts over relatively deep water, with the weapon carried on a floating barge. Effectively all the larger U.S. tests between 1954 and 1958 (when the three-year testing moratorium began) except high altitude tests were conducted by this means. Ogle says that the specific reason for the conversion to delivery by barge was to minimize fallout and atmospheric effects. One interesting facet, however, is that Ogle also reports that on several occasions the test barge was loaded with a large quantity of silica sand, apparently to test whether this assisted in reducing the geographical spread of fallout (why it should do so I do not understand). He does not report what happened to the silica sand, but it is reasonable to assume that some of it ended up in the stratosphere.
6. Although the U.S moved to barge-based testing on the open ocean from 1954 onwards, Ogle reported as follows concerning the "Oak" test in June 1958: "Oak was moved from comparatively deep mooring at Bikini Atoll to a position on the reef at Enewetok Atoll in order to increase even further the proportion of solids in the radioactive cloud...". Oak was an explosion of some 8.9 megatons. The use of the expression "even further" in this quote may throw some doubt on my earlier assumption that there was very little solid material in these barge-based tests which could be thrown into the stratosphere. However, I suspect it is just a reference to the loading of sand on some earlier barges in an attempt to limit the geographic spread of fallout. The amount of dust that would be atomised and thrown into the stratosphere by a large test on a shallow coral reef would certainly be many orders of magnitude higher than the amount of dust created by sand loaded onto a barge.
7. Russia's first test larger than 1 megaton was in November 1955. It exploded 11 thermonuclear devices of more than 1 megaton between then and the temporary (3 year) moratorium on testing which began in November 1958. All of them took place at Novaya Zemlya. The largest was 2.9 megatons, and there were also 7 tests with a yield of more than 100 kt conducted at Novaya Zemlya, which would also have had significant stratospheric effects for reasons discussed earlier.
8. Starting from early 1955 Russia started to make tentative overtures towards a test ban treaty, but the U.S. showed little interest at first. Then in November 1955 Russia was the first nation successfully to drop a thermonuclear weapon from an aircraft, and in August 1957 was the first to successfully deliver a nuclear weapon using an inter-continental ballistic missile. The U.S. then suddenly decided that it might be a good idea to begin discussing a test ban treaty! Russia announced a unilateral pause in its testing program from March 1958 (mostly because it had just finished an extensive planned testing program, and had no immediate plans to begin a new one anyway). The U.S. was reluctant at this point to agree to an immediate pause, because it was just about to begin its "Hardtack" testing program. Nevertheless, the 2 superpowers began tentative discussions on verification procedures that might apply to any test ban. Then, on 22 August 1958 (by which time the U.S. had almost finished its "Hardtack" testing program) President Eisenhower announced that the U.S. was ready to begin test ban negotiations on 31 October, and to suspend nuclear weapons tests on that date for one year (extendable) while the negotiations proceeded. Krushchev agreed to the negotiations but not at that stage to a moratorium. As Ogle reported, "In fact, Soviet testing, which had been in abeyance since March, resumed on September 20 with two very large explosions, and continued until November 3."
9. The last large U.S. test prior to the moratorium was code-named "Orange" and occurred on 12 August 1958. It was 3.8 megatons. As Ogle then reported: "No further tests then were performed by either nation until the Soviets burst forth with an astonishing 45 shots in 65 days beginning on September 1, 1961. Of these, 14 were above a megaton, and one yielded 63 megatons -- the largest bomb ever fired by any nation." The latter test by itself appears to have been at least comparable in size to the Pinatubo volcanic eruption.
10. Russia exploded 58 tests in the latter part of 1961, and 78 in 1962, all of which were atmospheric tests. The Russian test program in 1961 took place between 1 September and 4 November. All of the larger tests were at Novaya Zemlya and all were airdrops (and probably "airbursts" - see earlier). Thirteen (13) were over 1 megaton, including the huge 50-63 megaton blast mentioned before. In addition, 6 were over 100kt. The 1962 Russian program took place between 1 August and 25 December. All were airdrops. During that period there were 17 tests yielding more than 1 megaton, the largest being 24.2 megatons on Christmas Eve 1962 (the second largest weapon ever exploded in the atmosphere by any nation). In addition, there were 10 tests yielding more than 100kt. The Russian test program wound up with a 3.1 megaton blast on Christmas Day 1962. This was the last atmospheric test Russia ever conducted - a belated Christmas present to the planet !
11. The U.S. responded by hastily conducting 24 tests in 1962 (the Dominic programme). Nine of these were of greater than a megaton in size, the first of which occurred on 2 May 1962 . All of the larger ones were by air drop over the open ocean, mostly near Christmas Island (although a few were near Johnston Island near Hawaii -- the nuclear flashes were visible from the island of Oahu).
12. The US and USSR signed the Limited Test Ban Treaty on August 5, 1963. Russia conducted no atmospheric tests at all in 1963 and the U.S. conducted only 4 (none above a megaton). Neither nation has ever conducted any atmospheric testing since that time.
13. The UK first exploded a thermonuclear device in May 1957 at Montebello Island in the Indian Ocean. It subsequently stopped all atmospheric testing in 1958 (roughly the same time as the US-USSR temporary moratorium), and thereafter only tested underground. Thus Britain appears not to have made any significant contribution to global nuclear-induced cooling.
14. China first conducted a thermonuclear test in 1967 at its Lop Nor range in central China. It conducted some 12 tests in total between then and 1976, all of them atmospheric. I have been unable to ascertain how many of these were thermonuclear (and therefore almost certainly larger than 1 megaton), but it is certain that a fair percentage were that size or larger. China conducted a further 5 atmospheric tests between 1976 and 1980, but has only tested undergound since then. If we make the reasonable assumption that the global effect of the massive blitz of Russian nuclear tests in 1961-62 may have persisted for up to 4 years afterwards (consistent with the effect of Pinatubo), then the Chinese and French tests (see immediately below) took over virtually without a break and continued the global cooling effect right through until around 1976.
A certain amount of caution should be exercised, however, because it has also been estimated that the eruption of Mount Agung in Bali in 1963 contributed some 0.5°C to global cooling during the succeeding 2 years or so. If anything, this factor would help to "bridge the gap" from the end of atmospheric testing by the US and USSR at the end of 1962 and the beginning of testing by China and France in 1967-67, without the need to postulate that dust from the Russian tests persisted in the stratosphere for as long as 4-5 years in sufficent quantities to continue causing a measurable global cooling effect at the surface. One may note that global warming only begins to be apparent again in the surface temperature record from 1976 onwards.
15. France first tested a thermonuclear device in 1968. It conducted some 33 tests in total between then and 1976, all of them atmospheric. All its tests after that were underground. Again I am not sure exactly how many of these atmospheric tests were thermonuclear, but a fair few obviously were.
In summary, it now seems to me that there is a quite compelling case that nuclear testing had a significant cooling effect on global temperatures almost continuously from 1952 through until 1976, when global temperatures again began to rise. If one assumes that the size of this effect may have been as much as 0.5°C (the size of the Pinatubo effect), then this was more than enough to have masked any global warming effect from greenhouse gases that might otherwise have been evident. Of course, nuclear testing provides no explanation at all for the global cooling that occurred between 1940 and 1952. Nevertheless, the nuclear testing effect goes quite a long way towards explaining why the surface record showed cooling throughout the middle part of the twentieth century, despite large increases in atmospheric carbon dioxide. Although that certainly doesn't prove that the global warming theory is correct, it greatly reduces the force of one of the factors most often cited by contrarians as an argument against it.
Additional Information is available from:-
See: The High Energy Weapons Archive at http://www.fas.org/nuke/hew/ a website of the Federation of American Scientists. It is an information clearinghouse on nuclear weapons and arsenals, contains a nuclear weapons FAQ; a glossary of terms; background material on nuclear physics; histories and data on post-World War II nuclear weapons testing by the U.S., the former Soviet Union (Russia), England, France, and China; a complete list of U.S. nuclear weapons; profiles of nondeclared nuclear states Israel, India, Pakistan, and South Africa; and extensive collections of related articles and photographs. It also features a history of the development of atomic weapons, including coverage of the Manhattan Project, the Trinity test, and the Hiroshima and Nagasaki bombs. The site also contains an excellent list of weblinks to no less than 699 other nuclear weapons-related web resources. I have to confess I have hardly begun to delve into them yet.
Also see Project Whistlestop (website of the Truman Library), especially the documents section at: http://www.whistlestop.org/study_collections/bomb/large/bomb.htm . It contains a collection of papers, memos, telegrams, and other materials concerning the issues surrounding atomic weapons during the Truman administration, from the Harry S. Truman Presidential Library.
c/o Law Faculty
Northern Territory University
Darwin, Northern Territory
`Open Review' comments are invited
on this important new information
on the climatic effects of atmospheric nuclear testing.
Contributions should be emailed to
with `The Big Bangs' in the subject line. - JD
Return to Climate Change: Guest Papers page
Return to `Still Waiting For Greenhouse' main page
FastCounter by bCentral