This `Open Review' forum is open to expert and non-expert alike to submit comments. Comments should be addressed to daly@microtech.com.au with `The Big Bangs' in the subject line. Review comments should be relevant to the topic of nuclear testing and/or its climatic impacts, and not contain personal and/or ad hominem remarks.   - John L. Daly

Subject: Atomic tests and global cooling
Date: Thu, 2 Mar 2000 08:09:37 +0930
From: Aaron Coyan <ascoyan@zoomnet.net>
To: Ken Parish <kparish@legalnet.net.au>

Hello,

I read with great fascination your paper on this topic on John Daly's website this morning! I believe you present strong evidence that the atomic tests could have had some cooling effect on the global climate, but I feel compelled to point out an important fact.

You state that the cooling effect of atomic tests could be as great as that of Mt. Pinatubo. However, I believe such a comparison is flawed. One of the reasons the climatic cooling effect of Pinatubo was so great was the eruption's high content of sulphur dioxide (SO2). This led to the development of sulphate aerosols in the atmosphere, which are more effective at backscattering solar radiation than simple dust. Since atomic blasts send only dust and no (?) SO2 into the stratosphere, it seems that their cooling effects would have to be smaller, unless the quantity of the dust was enough to outweigh the lack of SO2..

The extensive Russian atomic tests in 1961 and 1962 preceded the extremely severe winter of 1962-63, which was severe throughout much of the northern hemisphere, and the very cold USA winter of 1963-64 (which was dominated by a frigid December 1963)! In the USA, those two winters were the coldest of the 1960s.

I have always believed that the cooler climate of the 1940s to the 1970s was a strictly natural occurrence. Your research has now swayed me to believe it is very possible that the dust from the atomic tests served to amplify this already naturally cooler period. It's probably no coincidence that the coolest portion of this period was the 1960s and early 70s.

Good work!

Aaron Coyan
Private Weather Consultant and researcher

P.S. I hold a master's degree in Geography (emphasis in meteorology and climatology).

Subject: atomic tests and global cooling
Date: Thu, 2 Mar 2000 08:54:30 +0930
From: Ken Parish <kparish@legalnet.net.au>
To: Aaron Coyan <ascoyan@zoomnet.net>

Dear Aaron,

You make a valid point that sulphate aerosols have a greater effect on backscattering solar radiation than simple dust. One of my purposes in having this article published was in the hope of eliciting points that I (as a non-scientist) may have missed, from debate participants more expert than myself. However, a couple of points occur to me:

1. The sheer massive volume of the dust created by the test programs of 1961-62 particularly may well have generated a temperature effect comparable to Pinatubo. Have another look at the number and size of the tests Russia conducted in a compressed time frame of just 65 days in 1961, and only slightly longer in 1962. Obviously the intensity of testing for the balance of the period 1952-62 (and 67-76) was considerably less, and therefore one would expect to see a lesser temperature effect during those times. No doubt it is true to say that the effect on surface temperature of (say) a 10 megaton nuclear blast would be significantly less than a volcanic eruption of similar size, because of the lack of sulphate aerosols in nuclear dust. But the effect is nevertheless measurable, and it doesn't need to be anywhere near 0.5°C in order to mask any underlying GHG-induced global warming trend (not, I hasten to add, that I am pushing the global warming barrow).

2. It appears to me from a fairly non-rigorous look at the station temperature graphs on John Daly's website that nuclear-induced cooling better explains the record for the 1950's and 60's than more conventional explanations (ENSO, solar activity effects, or even Dr. Vincent Gray's somewhat less conventional suggestion that the cooling derives from a mass movement of measurement stations to undeveloped airport sites etc). I would be most interested to hear what someone like Dr Theodor Landscheidt has to say about this.

3. On a more localised level, although the effects of stratospheric aerosols are global, commonsense would suggest that the temperature effects of a particular violent event (whether volcanic or nuclear) would be greater in the immediate vicinity. That is partly because there is tropospheric debris as well which, although it is in the atmosphere for a much shorter time, has a major local temperature effect while it lasts. The very apparent dips in the temperature graphs for 1961-64 seem significantly more marked for Russia/Siberia than for most other regions. This seems to me to lend at least circumstantial support to the proposition that there was a significant nuclear-induced cooling.

3. My non-expert understanding is that, while sulphate aerosols are more effective at backscattering solar radiation than simple dust, particle size is also a factor i.e. the smaller the particle the greater the backscattering effect. Might it not be the case that the concentrated power at the centre of a thermonuclear explosion generates particles whose average size is significantly finer than a volcanic eruption? I haven't yet been able to find material on this subject, and would be most interested in feedback from others who may have information. It occurs to me that there might be no definitive information on the subject, because there was no ready method of obtaining samples from the stratosphere back in the early 60's.

Regards     Ken Parish

Subject: The Big Bangs !
Date: Wed, 1 Mar 2000 16:29:31 -0700
From: Chick Keller <cfk@lanl.gov>
To: daly@vision.net.au

John,

Fascinating! What else have we missed?

Charles "Chick" F. Keller
Institute of Geophysics and Planetary Physics/University of California
Mail Stop MS C-305
Los Alamos National Laboratory
Los Alamos, New Mexico, 87545
cfk@lanl.gov
Phone: (505) 667-0920 FAX: (505) 665-3107
http://www.igpp.lanl.gov/climate.html

Subject: Big bangs
Date: Wed, 01 Mar 2000 15:48:53 -1000
From: George and Teresa Birchard <gfb@aloha.net>
To: daly@vision.net.au, kparish@legalnet.net.au

Ken,

What an interesting history lesson. Thanks for the contribution.

I have several bits of info. to add. Sulphate (as sulphuric acid aerosol) is a very long-lived and high albedo contributor to the global cooling effects of volcanic eruptions. The volcanic ash drops out much more quickly.

If it reaches the stratosphere, water vapor is removed slowly. There is a possibility of enhanced stratospheric ice cloud formation from nuclear testing done over the ocean. Dust is not necessary for stratospheric effects.

The sheer number of tests is impressive. Even if the effects of each test was shorter than volcanic sulphate, the repeated testing would have kept on pumping water and dust into the stratosphere. Global cooling could be expected.

The ice thickness measurements are not invalidated by this effect. How much thinner would the ice be today without the effects of nuclear testing? [Given that they showed thinning on one side of the Arctic and not the other, a number of other arguments against the ice thickness research do not appear to be valid. However, I agree with John that one study over a relatively short period of time is not proof of anthropogenic global warming.] It's just one of many complexities that needs to be considered.

Regards, George

Subject: Big bangs
Date: Fri, 3 Mar 2000 01:14:03 +0100
From: "Jorge Sereno" <sereno@zeelandnet.nl>
To: <daly@vision.net.au>

Dear Ken,

I've read things claiming this about 14 years ago. It were the scientists Kondratyev and Nikolsky that proposed it. I still have the book which deals with it. You can find the original article in "New scientist", 10 september 1981.

I however do not believe this is true at all. First of all, within the 1950-1976 period the 1957-1963 period stands out as the warmest. Of the 12 warmest years in this epoch all of the years between 1957-1963 are a member of those twelve years (ranked 3,4,5,6,8,9,12).

A combination of the GISS/CRU dataset reveals this:

1973 +0.21K (Moderate/strong El nino)
1953 +0.19K (El nino)
1958 +0.19K (moderate/strong El nino)
1957 +0.15K
1961 +0.15K (solar influences?, warm despite La Nina!!)
1959 +0.12K (solar?)
1952 +0.12K (moderate El Nino)
1963 +0.12K (solar?, warm despite La Nina!!)
1962 +0.11K (Solar?)
1969 +0.10K (Solar?,warm despite some volcanic influences)
1970 +0.09K (Solar?)
1960 +0.08K (solar?)
1951 +0.05K (despite weak La Nina)

The coldest years:

1964 -0.13K (Gunung Agung)
1976 -0.12K (Fuego eruption?, strong La Nina)
1956 -0.11K (Strong La Nina)
1950 -0.06K (Weak-Moderate La Nina)
1965 -0.06K (Weak La Nina, Gunung Agung)
1971 -0.06K (Moderate La Nina)
1955 -0.03K (weak La Nina)

All temperature anomalies based on the 1868-1998 mean. Nino3 indices are used (CRU) for SOI indices, Sato et al. (1993) for optical depth indices for volcanic aerosol influences. Hoyt and Schatten+Lockwood and Stamper models are used for solar influences.

Almost all warm anomalies occur in combination with strong EL ninos and/or possible strong solar influences (cycle 19 was a strong one and it peaked near 1960...). All cold anomalies occur in comination with a La Nina and sometimes with volcanic influences.

I further would like to add the fact that according to Sato et al. volcanic influences were not just confined to the Agung eruption and so not for two or three years afterwards. The 1963-1970 period had some other eruptions which had very clear effects on the optical depth (aerosol debries in the stratosphere). As did 1974 (Fuego). There is no need for any other factor to explain the anomalies in this period.

If we would subtract 0.5K in any year, than in this period any year would have been unexplainably cold. If you would stretch this 0.5K for the whole period, the 1960-1975 period would have brought us back to little ice age figures. This obviously has not happend. It isn't even the coldest period in this century.

To conclude: all anomalies can be easily explained with a combination of SOI, volcanic and sometimes solar factors. On a year to year bases, volcanic and strong SOI departures are dominant.

Nuclear blasts throw NOx in the atmosphere and no sulpuric acid. NOx are greenhouse gasses, they do not enhance the albedo of clouds nor the albedo of our atmosphere (as far as I know). If they had any sign because of nuclear blasts, it could very well be the opposite of what you claim.

Best regards,      Jorge

Subject: The big bangs
Date: Thu, 02 Mar 2000 20:30:12 -0700
From: Steve Hemphill <shemphill@bigfoot.com>
To: daly@vision.net.au

Excellent and well researched article. I was initially under the impression the primary stratospheric residuals were of vaporized solids (therefore not really "dust" but atomic particles), but George's comment on water vapor made me wonder if it could be significant as well. What would be the time for water vapor to "drop out" of the stratosphere? I'm also wondering if any sampling for deuterium or tritium has been done in the stratosphere as I think their concentrations in the stratosphere may be higher than the typical percentages elsewhere on Earth as a result of these tests, and the residual may be a good indication of how much is still up there.

This site is an index for many images. You can see the fireballs entering the stratosphere in many of them, and a great deal of moist tropical tropospheric air moving up with many of the blasts as well: http://www.fas.org/nuke/hew/Usa/Tests/index.html

Steve

Subject: Big Bangs
Date: Thu, 2 Mar 2000 21:27:27 -0500
From: "Douglas V Hoyt" <dhoyt1@erols.com>
To: "John Daly" <daly@vision.net.au>

Although the idea atomic testing may have changed atmospheric transmission and caused a cooling is interesting, I don't think the observations support. Atmospheric transmission measurements in Chile and California by the Smithsonian Astrophysical Observatory for 1923-1954 show constant transmission except for volcanic eruptions, with one exception that I will return to later. At Mauna Loa Observatory, we find the transmission is constant from 1957 to 1963 when Agung erupts. It returns to the 1957-63 mean after a couple of years. Davos Switzerland has constant transmission except for volcanic eruptions from 1909-1979. For 1951-1970, there are no transmission changes in Belgium. For 1950-1972, the transmission is unchanged in Ireland. These studies are mentioned in my book The Role of the Sun in Climate Change on page 210.

At the Smithsonian Astrophysical Observatory station in Chile, there was a decrease in atmospheric transmission in the early 1950's. Rob Roosen postulated that it was caused by atomic testing. Others, including myself, thought it was caused by the nearby large open pit copper mine. The issue was never resolved in the sense that a unanimous opinion was reached.

To summarize, I don't think the evidence supports any global trends in atmospheric transmission from atomic bombs or other causes. The only global changes in transmission appear to arise from explosive volcanic eruptions.

Doug Hoyt

Subject: Reply to Jorge Sereno
Date: Fri, 3 Mar 2000 18:04:10 +0930
From: Ken Parish <kparish@legalnet.net.au>
To: "John Daly" <daly@vision.net.au>

Jorge Sereno raises some interesting questions.

Does the surface temperature record for the period 1952-76 coincide with the pattern one would expect from nuclear testing during that period? Does it coincide with ENSO cycles over that time? My short answer to both questions is YES. See below for more detail.

However, if the surface temperature coincides so well with ENSO, why is there any need to resort to nuclear-induced cooling as an explanation? The simple answer is that, although the year to year peaks and troughs of surface temperature correlate quite closely with El Nino/La Nina events, ENSO patterns do not in themselves explain the fact that there was a net cooling trend of around -0.15 degrees C over the period 1945 to 1976.

Of course, one or more of the numerous other factors which influence global climate, most obviously variations in solar activity, may well play a role. My point is that the sheer number and size of nuclear tests, whose duration coincides remarkably closely with the period of the net cooling trend, may well provide at least part of the answer to why there was such a net cooling while atmospheric CO2 levels were rising.

My article does not claim that the effect was as large as 0.5°C throughout this 24 year period: it obviously wasn't that large for most of the time. I do, however, think it likely that the size of the nuclear-induced cooling could well have been as large as 0.5°C in the aftermath of the massive testing blitz of 1961-62.

For the overall period, however, there is no need to posit an effect as large as that: the net cooling is only 0.15°C. Earlier contributors have already pointed out that nuclear dust in general might not have as large or as long a stratospheric effect as volcanic residue, because of the lack of sulphate content. But surely Mr. Sereno does not suggest that such a large volume of fine particles in the stratosphere, continually being renewed by fresh blasts, would have had no effect at all!

As I have already postulated, the very small particle size of the nuclear dust might partially compensate for the lack of sulphate content and, as George Birchard has suggested, water vapour might also play a role by way of formation of stratospheric ice clouds (a prenomenon that only sometimes accompanies volcanic eruptions). I am in no sense denying that ENSO, volcanic and solar activity play important roles in global climate: manifestly they do. I simply suggest that the effect of nuclear testing may be an additional factor which has not until now been recognised, and which at least partly explains the overall cooling trend.

The easiest way to observe the comparative patterns between surface temperature and ENSO is by reference to the global mean surface temperature chart found at The Climatic Research Unit at the University of East Anglia website (the Gnomes of Norwich, as John Daly and others call them), which is based on data produced by Jones, P.D., 1994: Hemispheric surface air temperature variations: a reanalysis and an update to 1993. J. Climate 7, 1794-1802. http://www.cru.uea.ac.uk/cru/data/hemglob.gif , with Figure 1 from "Solar Activity Controls El Niño and La Niña" by Dr Theodor Landscheidt, found on John Dalys website at http://www.john-daly.com/sun-enso/sun-enso.htm . The latter chart compares ENSO cycles between 1950 and 2000 with sea surface temperature anomalies in the southern Pacific. As you will see, the surface record compares quite closely with both SST and ENSO cycles:

1. Despite a "double whammy" El Nino in 1952 and 1954, the corresponding surface warming is very modest, conceivably reflecting some effect from the intial US and Russian tests.

2. Evident and deep cooling between around 1955 and 1958 coincides both with a large La Nina peaking in 1956 and with the intense US and USSR nuclear testing over that same period.

3. Significant warming between around late 1957 and 1962 roughly coincides both with the temporary 3 year nuclear testing moratorium and with another "double whammy" El Nino in 1958 and 1960. The surface record does not coincide precisely with either ENSO or nuclear testing during 1961-62, in that temperatures remained warm despite both the resumption of testing and a modest La Nina in the same period. It should be remembered, however, that Russian testing did not resume until September 1961 and the huge 63 megaton blast did not occur until 31 October.

Moreover, American testing did not resume until May of the following year 1962. Clearly, however, the persistence of warming through 1962 and early 63 is not explained either by ENSO or testing. It suggests, as Mr. Sereno observes, that there is also some other countervailing factor at work, possibly enhanced solar activity.

4. The remainder of the 1960's, from the second half of 1963 right through to 1969, is a very clear cool period. It certainly isn't fully explained by ENSO: the temperature peaks are much smaller and the troughs much deeper than one would expect from the size of ENSO events. Indeed, there is almost no La Nina event to speak of through this period, and the relatively large El Nino event of 1966 barely makes a "blip" on the temperature chart. Of course, this could be entirely due to volcanic activity (including Agung), but one wonders why Mr Sereno feels compelled to dismiss nuclear testing as a possible contributing factor.

Comparative analysis such as the above neither proves nor disproves the role of nuclear testing in global climate during 1952-76, but I suggest that it certainly provides at the very least a possible explanation for the net cooling trend. Whether vulcanism can provide such an explanation I simply don't know. It might do so if one could demonstrate that there were more and larger eruptions during that period than either before or afterwards, but I am not aware that that is the case. Similarly, solar activity may provide an explanation, but others much more knowledgeable than I can examine that question.

The remarkable thing about nuclear testing is that its duration and frequency coincide so closely with the only period of net cooling in the twentieth century. Unlike volcanoes, there were no atmospheric nuclear tests before this period, and none afterwards. Of course, this could just be pure coincidence, but to dismiss the possibility of a climate effect without any real examination does not seem to accord with the scientific method.

Regards       Ken Parish

Subject: Big Bangs
Date: Fri, 3 Mar 2000 10:26:43 +0100
From: "Gerhard Grasruck" <Menschmaschine@gmx.de>
To: <daly@vision.net.au>

Hello,

you present here a very interesting theory, but I remain rather sceptical. The main problem, I think, is how much of the thermonuclear explosions did actually eject a significant amount of dust into the atmosphere.

But what is significant? I got the impression from the text, that you consider the amount of material from the H-bomb itself that is ejected into the stratosphere, as sufficient to have an impact on climate.

I may be wrong, but intuitively I would consider this as impossible. The radioactive dust into which the bomb gets converted may be significant regarding its health hazards for humans, but I simply can't imagine how 1 or 2 tons of material blown into the stratosphere can have any measurable impact on the earths solar albedo. There has to be additional material present. However, if we assume that, there aren't too many possibilities left. We may divide the thermonuclear tests into the following categories:

-ground explosions: in these cases, surely a vast amount of dust gets ejected into the stratosphere (How much would be interesting - are there any numbers available?). However, none of these were conducted after 1958.

-air explosions high enough to prevent the fireball from touching the ground: In these instances no dust will go into the stratosphere save of course a part of the material the bomb was made of.

-air explosions over sea, low enough to cause the fireball to touch the surface: No additional dust here, too, only water vapour

-air explosions over land, low enough to cause the fireball to touch the ground: This will cause a certain amount of dust from the ground to get into the stratosphere, though less than in a ground explosion.

-water explosions: Once again, even if there is some additional material from the carrying barge, I really can't imagine this to be significant.

While it seems indeed possible, that the dust from the ground tests up to 1958 did have some effect on temperatures, after that the theory is on extremely shaky ground. All thermonuclear tests conducted by the US were airdrops over water and therefore can't possibly have had any significant impact. Regarding the Soviet tests, it would be necessary to know if the bombs were ignited in a high enough altitude to prevent a contact with the ground. If this is so - and Ken says that it is probable - then these tests would have to be written off as a source for stratospheric dust as well.

The same goes for the Chinese thermonuclear tests starting in 1967; it would be necessary to know if any of the explosions took place close enough to the ground to decide if they could have propelled dust into the stratosphere. Regarding the French tests starting in 1968 we can be sure that this didn't happen, since all of them were carried out over water (The French had a nuclear test range in southern Algeria, however they had to abandon it in 1964 after they lost control of that country).

Whether thermonuclear tests have contributed to the decline in global temperatures from 1940 - 1976 (and if yes by how much) is an interesting question. But in contrast to you and John Daly I don't think that it has much influence on the 'greenhouse'-climate debate. Why? Well, even if one assumes that a hypothetical warming caused by the rising CO2-concentration would have been 'masked' by stratospheric dust, the rise in temperature would have been obvious as soon as the tests stopped and the dust was removed from the stratosphere. This however clearly did not happen, the temperature did, after the discontinuitation of the tests, not reach a level anywhere near that what the 'greenhouse'-climate theory demands.

mfG        Gerhard Grasruck

Subject: Douglas Hoyt contribution
Date: Fri, 3 Mar 2000 21:51:23 +0930
From: Ken Parish <kparish@legalnet.net.au>
To: <daly@vision.net.au>

Douglas Hoyt's contribution to the debate on the effect of nuclear debris on global climate is clearly more substantial than that of Jorge Sereno. Indeed, if Hoyt had cited evidence that significant numbers of observatories across the world detected no transmission changes over a substantial proportion of the 24 years of nuclear testing, then I might well be ready to concede that my hypothesis must be wrong. Even then, I would want to know why such a large injection of stratospheric material should have no detectable effect, and I would also need to know whether the standard error/limitations of accuracy of the instruments in question (presumably telescopes) enabled them reliably to detect transmission changes as small as those we are considering (perhaps on average only 20% of that of a large volcanic event).

But Hoyt does not cite a large number of observatories nor a long time period. His first two examples, Chile and California between 1923 and 1954, are almost entirely irrelevant. Testing of nuclear weapons of the requisite yield only began in 1952. Thus, probably only a single test ("Mike" in 1952), out of the hundreds in total, occurred in the period covered by the measurements from these observatories (the first Russian thermonuclear test in 1953 was only 400 kt, but may have had some effect because it took place at Novaya Zemlya where the tropopause is at a reduced altitude).

Mauna Loa between 1957 and 1963 is irrelevant for similar reasons: - this is largely the period of the temporary testing moratorium (although one might expect to see some effect by 1963).

Davos, Belgium and Ireland provide more difficult cases. They cover a long enough time span that one would expect to see some effect. However, it should be remembered that the surface record shown in the Jones graph (see my last email) is an adjusted average of a very large number of station records world wide. The fact that it shows a cooling at a particular moment does not mean that this will be equally detectable at every point on the globe. A significant part of the temperature effect of a nuclear blast is tropospheric and therefore by definition short-term and relatively localised. But the sheer volume and frequency of tests means that this fairly local effect may have a measurable cooling influence on the global mean.

Even with stratospheric effects, I have hypothesised that they may be more marked in the immediate area of the tests in question. Again that may reflect significantly in the global average figure without necessarily being detectable at a site many thousands of miles away. Whether it will be clearly detectable a long way away may depend upon the nature and direction of stratospheric flows (which I understand change seasonally). If the fallout has travelled right around the globe before reaching the site in question, it may be so diffused as to be almost undetectable.

It may be significant in this regard that at least one of the six observatories in question detected anomalies which at least some experts thought were caused by nuclear testing. The level of uncertainty surrounding this data seems no higher than with a lot of other areas potentially crucial to resolution of the question of whether there is a detectable anthropogenic warming trend (a very recent example is the disagreement on this website about whether the Arctic is presently warming or cooling, a question one would have thought would be more easily capable of definitive resolution than the causes of temperature trends from 25-50 years ago).

Regards    Ken Parish

Subject: Gerhard Grasruck contribution
Date: Sat, 4 Mar 2000 06:19:59 +0930
From: Ken Parish <kparish@legalnet.net.au>
To: <daly@vision.net.au>

On rereading the FAQ quote from my article, I see what Gerhard Grasruck was getting at. It appears that the FAQ author, in referring to "true airbursts" throwing large amounts of fallout high into the stratosphere, was referring to fallout gases rather than solids which, though they would be highly radioactive and possibly dangerous to humans (although the FAQ author suggests elsewhere that the half life of the most dangerous elements is so short that the fallout is relatively harmless by the time it re-enters the troposphere), could be expected to have little effect on backscattering solar radiation.

If that is correct (and I agree that it probably is), then Gerhard Grasruck is also correct that we can only quantify the likely climate effects of nuclear testing if we know both the altitude of each explosion and the extent to which the fireball intersected the ground (in the case of Russian tests) or sea (in the case of most US tests). I do not know either of these things. I suspect that such information is not publicy available on the Russian tests: I had considerable difficulty finding out as much as I did about their program. On the other hand, much information is available about US testing, because of the US Freedom of Information Act . I was rather hoping that my raising of this issue might prompt someone with more scientific understanding than myself to delve into questions such as this, because I think they are quite important (see below). One may reasonably conjecture that both the Russians and Americans exploded weapons of differing yields at a wide variety of altitudes: that would seem inherent in the notion of "testing". Some would have intersected the ground/ocean and some would not. We need to know.

The last point Gerhard Grasruck makes, that any nuclear testing effect is unimportant because GCMs so overstated global warming to date that any modest "masking" effect from nuclear testing cannot possibly explain away the discrepancies between the models and observed reality, is, with respect, misconceived.

Let me explain. - It is certainly true that earlier GCM's generated results which greatly overstated the expected warming to date (and during the mid century cooling). More recent models, however, have been brought more in line with oberved reality, to some extent by knowing more about what was and is actually going on, but also to some extent, one suspects, by fiddling with variables and unquantifiable unknowns and the modelled interrelationships between them and more precisely quantifiable factors. In other words, they fiddle with the models until they fit observed reality! However, as contrarians frequently point out, this is retrodiction not prediction. The GCM's will continue to have scant predictive value until many more of the presently insufficiently known factors (e.g. stratospheric circulation, stratosphere-troposhere exchange mechanisms, thermohaline flows, ocean-atmosphere coupling, positive and negative feedbacks) are fully understood and mapped. Many of these issues are presently being studied intensively with the assistance of generous government funding elicited with the aid of frequent greenhouse warming scare stories. The research may one day give the models some useful predictive value, but don't hold your breath.

If we assume (as I do) that the models in their current state of development are a joke, it is in that context that my hypothesis regarding the possible climate effects of nuclear testing assumes a degree of imortance (although it should not be overstated). By adding a possible additional anthropogenic factor to the range of things which influenced global climate during the mid century cooling period, we must consider the intriguing though probably unverifiable possibility that there might have been a "masked" underlying anthropogenic warming trend. Since we can't place any credence in the GCM's, trying to understand more clearly from observed reality what actually happened in the past and what is happening now assumes a much greater importance.

I should stress that vulcanism is logically in the same category as nuclear testing for the purpose of assessing whether there is an anthropogenic warming trend, even though it is not itself anthropogenic. Both vulcanism and nuclear testing have the capacity to "mask" underlying anthropogenic warming trends stemming from GHG's. If one can say that that there was greater vulcanism during the mid century warming than either before or since, then it is possible that it too masked any underlying anthropogenic warming trend. The importance of the nuclear testing factor is that we can say unequivocally that it occurred during the cooling period, but neither before or since. What remains unknown is the exact extent of the climate effect. Unless one can eliminate both vulcanism and nuclear testing as possible masking factors during the mid century cooling, it seems to me that one is compelled to be an agnostic on global warming rather than a committed atheist.

Regards    Ken Parish

Subject: The Big Bangs
Date: Fri, 03 Mar 2000 16:35:21 -0700
From: Steve Hemphill <shemphill@bigfoot.com>
To: daly@vision.net.au

One concept missing in this conversation so far is that ENSO is a reaction to increased temperatures. Therefore, if Earth is cooling from some mechanism, whether natural or anthropogenic, El Ninos should decrease. This is shown in correlation with both El Nino to temperature on a decadel scale and with the "snuffing out" of El Ninos between the big ones of 1983 and 1998 by Chichon and Pinatubo. Therefore we can't blame cooler temperatures on La Nina. There is a correlation because they are both responses to previous stimuli.

Also, I'm curious about the atmospheric transmissivity data. Common sense would indicate that transmissivity from below the stratosphere (refraction) would react differently than from above (reflection). What are the assumptions, procedures and wavelengths measured? Is there a web site with this information?

Steve

Subject: Big Bangs
Date: Fri, 3 Mar 2000 22:31:49 +0100
From: "Jorge Sereno" <sereno@zeelandnet.nl>
To: <daly@vision.net.au>

Dear Mr. Parish,

I have sound arguments to think you are wrong. I will try to do prove them within one week or even much shorter, as my research, which deals with reconstructing the climate back till 1868 and even 1693, is now finished. I can tell you that I have a correlation for the whole period (year to year basis) of 0,89 (1868-1998) and for the period 1958-1976 of 0,79. The years 1970-1976 show a reconstructed temperature that deviates from the observed temperature by less than 0,03K for every year. My models show the rise between 1910-1940, the fall from 1940-1976 and the rise afterwards. Thats why I could give you this answer, I have already done my research and (some) statistical analyses. i just looked at the numbers of my models (which incorporate SOI, Volcanic, Co2 and solar influences) and replied on your research.

You say:

"However, if the surface temperature coincides so well with ENSO, why is there any need to resort to nuclear-induced cooling as an explanation? The simple answer is that, although the year to year peaks and troughs of surface temperature correlate quite closely with El Nino/La Nina events, ENSO patterns do not in themselves explain the fact that there was a net cooling trend of around -0.15 degrees C over the period 1945 to 1976. "

No, but how about 0,08K, based on 6 year averages? According to my models. And about 0,08K for volcanic eruptions and about 0,1K for solar influences. This totals 0,26K. But the Co2 rose and in my model this leads to 0,12K rise. All 6 year averages, all for the period between 1940 and 1976. There you have 0,14K of cooling.

"My article does not claim that the effect was as large as 0.5°C throughout this 24 year period: it obviously wasn't that large for most of the time. I do, however, think it likely that the size of the nuclear-induced cooling could well have been as large as 0.5°C in the aftermath of the massive testing blitz of 1961-62."

1963 was on the warm side. 1964 was about 0,2K cooler. If you are right, Gunung Agung had no cooling effect or even a warming effect. While volcanoes erupt millions of tons of sulphur in the atmosphere, nuclear bombs do not. The effect of Gunung Agung is comparable with that of El Chicon and Pinatubo. So it is highly unlikely that nuclear bombs had any effect. Those years are no different than other year just after a major eruption.

"For the overall period, however, there is no need to posit an effect as large as that: the net cooling is only 0.15°C. Earlier contributors have already pointed out that nuclear dust in general might not have as large or as long a stratospheric effect as volcanic residue, because of the lack of sulphate content. But surely Mr. Sereno does not suggest that such a large volume of fine particles in the stratosphere, continually being renewed by fresh blasts, would have had no effect at all!"

The volcanic residue is only 2-3 years having an effect. Dust does not equal cooling! What if they injected enormous amounts of CO2? Would this cool our planet? Do cirrus clouds cool our planet. They don't. Worse: the opposite is true in the extra tropics. SO2 enhances the albedo of clouds, clearly visible from space. NOx are greenhouse gasses and it are those NOx's that are injected in our atmosphere (according to Nikolsky and Kondratyev, back in 1981).

"I am in no sense denying that ENSO, volcanic and solar activity play important roles in global climate: manifestly they do. I simply suggest that the effect of nuclear testing may be an additional factor which has not until now been recognised, and which at least partly explains the overall cooling trend. "

How much then? I think you cannot prove anything if it were 0,05K or even less. And even so: less than 0,05K doesn't seem very significant to me.

"4. The remainder of the 1960's, from the second half of 1963 right through to 1969, is a very clear cool period. It certainly isn't fully explained by ENSO: the temperature peaks are much smaller and the troughs much deeper than one would expect from the size of ENSO events. Indeed, there is almost no La Nina event to speak of through this period, and the relatively large El Nino event of 1966 barely makes a "blip" on the temperature chart. Of course, this could be entirely due to volcanic activity (including Agung), but one wonders why Mr Sereno feels compelled to dismiss nuclear testing as a possible contributing factor."

It is not a "very cool period". It is somewhat colder. Volcanic eruptions other than Gunung Agung are responsible (read Sato et al. paper of 1993). One in 1966 and one in 1968. That is why nuclear bombs are not needed.

"The remarkable thing about nuclear testing is that its duration and frequency coincide so closely with the only period of net cooling in the twentieth century. Unlike volcanoes, there were no atmospheric nuclear tests before this period, and none afterwards. Of course, this could just be pure coincidence, but to dismiss the possibility of a climate effect without any real examination does not seem to accord with the scientific method."

I have done my examination and can easily explain the period 1958-1976 without anything else than natural and antropogenic influences (other than nuclear blasts). I can prove statiscally that I am right. It does not necessarly mean that I really am right though.

From the period between 1922-1963 there were no major eruptions. All eruptions of the '60 were much larger than any of them in the period 1922-1963 (there were almost no eruptions at all!).

I would like to see some statistical analysis if possible. This could support your claim.

best regards,    Jorge Sereno

Subject: The Big Bangs
Date: Fri, 03 Mar 2000 16:06:10 -0800
From: Tim Ball <tball1@home.com>
To: daly@vision.net.au

The first of many times I took my undergraduate climate class to the Winnipeg Airport to have a briefing as commercial pilots flying to Vancouver I learned that the fight was about 2000m into the stratosphere for 90 percent of the flight. Subsequent investigation showed that most commercial flights outside the tropics in the northern hemisphere in winter were in the stratosphere. They are in the stratosphere even in summer in the high latitudes.

I have not calculated the total amount of gas added over the years, but suspect it's considerable. Water vapour is most important because it is the forgotten gas above and below the tropopause. It is possible that the increase in Polar Stratospheric and Noctilucent Clouds are explained by this injection. The development of high flying jets also coincides with the period of atmospheric nuclear testing so it will be difficult to sort out the white noise from the red. Of course, as another contributor noted, what else have we missed.

Tim Ball Ph.D

Subject: Further response to Jorge Sereno
Date: Sat, 4 Mar 2000 15:13:41 +0930
From: Ken Parish <kparish@legalnet.net.au>
To: Jorge Sereno c/o <daly@vision.net.au>

Here is a further extract from the FAQ section of the Federation of American Scientists website linked at the bottom of my article. It is written by Carey Sublette, although I do not know his qualifications. He confirms Jorge Sereno's contention regarding the high nitrate content of fallout. Also note the other bolded comment concerning soot.

"5.2.2 Effects on the Atmosphere and Climate

Although not as directly deadly as fallout, other environmental effects can be quite harmful.

5.2.2.1 Harm to the Ozone Layer The high temperatures of the nuclear fireball, followed by rapid expansion and cooling, cause large amounts of nitrogen oxides to form from the oxygen and nitrogen in the atmosphere (very similar to what happens in combustion engines). Each megaton of yield will produce some 5000 tons of nitrogen oxides. The rising fireball of a high kiloton or megaton range warhead will carry these nitric oxides well up into the stratosphere, where they can reach the ozone layer. A series of large atmospheric explosions could significantly deplete the ozone layer. The high yield tests in the fifties and sixties probably did cause significant depletion, but the ozone measurements made at the time were too limited to pick up the expected changes out of natural variations.

5.2.2.2 Nuclear Winter The famous TTAPS (Turco, Toon, Ackerman, Pollack, and Sagan) proposal regarding a potential "nuclear winter" is another possible occurrence. This effect is caused by the absorption of sunlight when large amounts of soot are injected into the atmosphere by the widespread burning of cities and petroleum stocks destroyed in a nuclear attack.

Similar events have been observed naturally when large volcanic eruptions have injected large amounts of dust into the atmosphere. The Tambora eruption of 1815 (the largest volcanic eruption in recent history) was followed by "the year without summer" in 1816, the coldest year in the last few centuries.

Soot is far more efficient in absorbing light than volcanic dust, and soot particles are small and hydrophobic and thus tend not to settle or wash out as easily.

Although the initial TTAPS study was met with significant skepticism and criticism, later and more sophisticated work by researchers around the world have confirmed it in all essential details. These studies predict that the amount of soot that would be produced by burning most of the major cities in the US and USSR would severly disrupt climate on a world-wide basis. The major effect would be a rapid and drastic reduction in global temperature, especially over land. All recent studies indicate that if large scale nucelar attack occur against urban or petrochemical targets, average temperature reductions of at least 10°C would occur lasting many months. This level of cooling far exceeds any that has been observed in recorded history, and is comparable to that of a full scale ice age. In areas downwind from attack sites, the cooling can reach 35°C. It is probable that no large scale temperature excursion of this size has occurred in 65 million years.

Smaller attacks would create reduced effects of course. But it has been pointed out that most of the world's food crops are subtropical plants that would have dramatic drops in productivity if an average temperature drop of even one degree were to occur for even a short time during the growing season. Since the world maintains a stored food supply equal to only a few months of consumption, a war during the Northern Hemisphere spring or summer could still cause deadly starvation around the globe from this effect alone even if it only produced a mild "nuclear autumn". "

As for Jorge Sereno's apparent claim that tiny nitrates would have a greenhouse enhancing effect, I am not sure of this. I know that nitrous oxide (N2O) is a greenhouse gas. I don't know whether it is nitrous oxide (N2O) that is produced in such large quantities by nuclear weapons, or other nitrate compounds (and what their effects might be). If it is nitrous oxide, this could well account for there being little net cooling effect i.e. the warming effect of the nitrous oxide could cancel out the cooling effect of nuclear dust and soot. Again, we need to know more.

I would be most interested to read Jorge Sereno's paper if he is intending to publish it (or at least an abstract of it) on this site. It sounds fascinating. The figures you give for each of the posited main contributors to global climate raise many issues. For example, it seems to be inherent in your figures for the years 1940-76 that, in the absence of significant volcanic activity (-0.08K) and assuming constant solar activity, there would have been a very slight warming (+0.1K (solar) - 0.12 (C02) = -.02K). Is that correct? However, you say that there were no significant volcanic events between 1922 and 1963. How then do we account for the cooling between 1945 and 1963? Presumably significantly decreased solar activity during that period. Do the figures show that? It appears that the result of your model over the very long term (when solar and volcanic variations could be expected to even themselves out) would be that anthropogenic CO2 would cause a global temperature rise of no more than about 0.1 K per hundred years (even assuming that your posited contribution for solar activity for the period 1940-76 was somewhat below the long-term mean - which is presumably the case). It seems that I won't need to sell my waterfront home for quite a few centuries yet!

Regards       Ken Parish

Subject: Big Bangs
Date: Sat, 04 Mar 2000 11:10:50 +0100
From: Jorge Sereno <sereno@zeelandnet.nl>
To: Steve Hemphill c/o daly@vision.net.au

Dear Steve,

Some scientists have called the El nino's of between 1990-1995 the "El Nino that refused to stop". There have been very strong El nino's in colder epochs, such the one in 1674 (Little Ice Age), the long one from 1790-1793 or even 1794. The one near 1800 etc. etc. The KNMI reported about El Nino and La Nina starts as follows: El Nino: much influence but not te be influenced.

However, there are scientists who say that El nino's will become stronger because of human induced global warming (based on ECHAM 4 Max Planck Institute models) while other models (based on the Hadley Centre models) report the exact opposite. So saying that a colder period will coincide with stronger La Nina vice versa is not proven at all. Pinatubo's eruption happens to prove the opposite of what you claim, from 1991-1995 the SOI is 95% of the time negative. After 1963 I see no sign at all, just a mix of comparably strong El nino's and La Nina, in 1966 a strong El nino occurs. In 1983, we saw the largest deviation of SOI ever recorded, just when the El Chicon eruption (1982) had its biggest effect. It was followed by a moderate La Nina in 84/85. El nino's can fade a way very quickly, look at the 1997/98 El nino, but many others follow the same behaviour. Dr. Theodor Landscheidt has correlated SOI with solar influences, so it could be that El nino La Nina are yet another way in which solar activity influences our climate. Dr. Landscheidts theory is not proven yet, however.

Regards,     Jorge Sereno

Subject: The Big Bangs
Date: Sat, 04 Mar 2000 22:37:18 +1100
From: "John L. Daly" <daly@vision.net.au
To: Douglas V Hoyt <dhoyt1@erols.com> CC: Ken Parish <kparish@legalnet.net.au>

Dear Doug

you wrote:

Although the idea atomic testing may have changed atmospheric transmission and caused a cooling is interesting, I don't think the observations support.

The whole thing with Ken Parish's `Big Bangs' has me quite intrigued. We know that numerous weapon tests penetrated up to the stratosphere, and we know there was a general cooling between 1940 and 1976 of about -0.25°C. Beyond that, it's all a big grey area. It's also peculiar that we had the biggest solar maximum on record in 1957, right in the middle of the post-war cooling, with no apparent surge in global temperature.

It is also the case that the scientific establishment during the early 1980s were supportive of the `Nuclear Winter' theory, including the NRC and the climate modellers. If they believed then that nuclear explosions could lead to a catastrophic cooling, it seems only reasonable to infer that frequent testing would have a similar result albeit on a much smaller scale.

This leads me to ask in respect of those atmospheric transmission observations you report, what exactly were they measuring? Was it transmission over a broad band of wavelengths, or was it applicable to visible light only? It occurred to me that a big test, where the fireball temperature runs to millions of degrees, might vapourise (or even `atomise') large amounts of material, creating not just debris, not just dust, but material in gaseous form. Is that possible or likely? If so, would transmission sensors detect exotic gases in the stratosphere, particularly if these gases were transparent to visible light but possibly absorbing of UV and near-Infra-red? If that were the case, visible sunlight would appear unaffected, but up to 55% of solar energy in the non-visible parts of the spectrum might be subject to some absorption much the way ozone absorbs UV-B now.

Taking another look at the Arctic stations, they all appear to suffer an anomalous cold spell in the late 1960s, but that was well after the test ban treaty of 1963. I had thought the Novaya Zemlya blasts would impact on nearby Arctic station temperatures, but I can't find a clear indication of that in various station records, neither the Russian ones, nor the Norwegian ones.

Maybe we should not regard the tests as volcano-like at all. Volcanoes blast debris and aerosols in one dramatic burst over a short time period of weeks at most. But these tests were like a `drip, drip' process, no single test matching a serious volcano, but collectively putting incremental amounts of material (dust, gas, whatever) into the stratosphere over a long period of years, creating what would amount to a continuous background loading in the stratosphere over the 25 years of testing by various powers. If that was how it happened, the extended slow cooling would be an expected outcome, a cooling which would end as abruptly as the tests themselves - from 1976 onward. And as we know, the cooling did end at 1976.

Your comments on these ideas would be welcome.

Regards    John Daly

--
John L. Daly
"Still Waiting For Greenhouse"
http://www.john-daly.com/daly

Subject: Big Bangs
Date: Sat, 4 Mar 2000 07:46:06 -0500
From: "Douglas V Hoyt" <dhoyt1@erols.com>
To: John L. Daly <daly@vision.net.au>

John,

The transmission is measured using pyrheliometers that record all the radiation from 0.3 to 4.0 microns. The transmission is monitored by the air mass ratioing technique invented by Howard Ellis. This technique is insensitive to any calibration changes and insensitive to changes in amounts of ozone and water vapor. It is sensitive to changes in aerosol loading, so the results are probably quite robust. The method can also detect small changes in aerosol amounts.

Doug

Subject: The Big Bangs
Date: Sat, 04 Mar 2000 17:12:55 -0400
From: "Dr. Theodor Landscheidt" <theodor.landscheidt@ns.sympatico.ca>
To: John Daly <daly@vision.net.au>

Dear John:

If scientists do not look through because their data are overly complex they often continue on a coarser scale to get a better perspective. The plot below is of this kind. I explained it in detail in the discussion of my paper "Solar activity controls El Niño and La Niña" in my letter of 5 April 1999 to Andris Auliciems.

The smoothed aa index (solid curve) is a proxy of the intensity of the solar wind near earth. It is an indicator of the Svensmark effect on temperature via variations in cloud cover. The combination of land air and marine temperature anomalies (dashed curve) lags near-earth variations in the solar wind by 4 to 8 years, thus indicating a causal connection. Allowing for this lag, an outstanding aa peak around 1990, not shown here, could explain the high temperature observed in 1998. The ensuing steep decline in the aa index, again not shown here, indicates decreasing global temperature during the next ten years. The close correlation ( r = 0.96; P = 10 to the -7) explains 92% of the variance in the temperature.

The temperature curve follows the undulations of the aa curve. Corresponding extrema are marked by equal numbers 1 to 6. There is only one exception between 1940 and 1952 with rising aa data and falling temperatures. If this were the period of big nuclear bangs, this would provide an explanation. Yet the big bangs begin just when the smoothed temperature curve begins to rise and tune in to the leading aa course. So there is no strong indication of a big bang effect, as far as I can see. Solar forcing seems to be much stronger.

By the way, I would be happy if you or someone else could explain what happened after 1940. Did mother earth shudder at the atrocities of World War II?

Kind regards,      Theodor

Subject: Big Bangs
Date: Sun, 5 Mar 2000 01:10:22 +0100
From: "Jorge Sereno" <sereno@zeelandnet.nl>
To: John L. Daly <daly@vision.net.au>

Dear John,

You said to Doug Hoyt:

"The whole thing with Ken Parish's `Big Bangs' has me quite intrigued. We know that numerous weapon tests penetrated up to the stratosphere, and we know there was a general cooling between 1940 and 1976 of about -0.25°C. Beyond that, it's all a big grey area. It's also peculiar that we had the biggest solar maximum on record in 1957, right in the middle of the post-war cooling, with no apparent surge in global temperature."

Well, you correlate the highest number of sunspots with solar influences on global temperatures. Is this valid?

Well: if we take a look at the Schatten & Hoyt model (1993), it isn't. The peak was near 1938. In 1958, the activity was already total solar irradiance was already considerably lower. The same is true for Lassen & Friis Christensens model. But Schatten & Hoyt used some of the same proxies (solar cycle length) as Lassen &Friis Christensen did. Then we have Lean & Foukal and Lockwood & Stampers model. Here, the peak in 1958 is larger than in 1940, but smaller than in 1980 and 1990. Lockwood and Stamper use the AA- or Ap-index as a proxy for solar irradiance variations.

So if we only consider the period Ken Parish is referring to, the answer isn't clear. I don't know which of the two type of models is right. So I use Schatten & Hoyts model and that of Lockwood & Stamper. Combined, we still see a fall after 1950. In this period (and only here), those models are so different that not both can be true. The best thing now is to use a combination of those two types, I think.

Dr. Landscheidt has pointed out more than once that sunspots are not the sign we should look for if we want to know (or reconstruct) solar influences on our climate. This supports the general view that while cycle 19 was a large one, it has not been as large as sunspots seem to indicate and quite possibly lower than the preceding decade (end of 1930's and 40's).

Jorge Sereno

Subject: The Big Bangs
Date: Sun, 05 Mar 2000 13:08:02 +1100
From: "John L. Daly" <daly@vision.net.au
To: Jorge Sereno <sereno@zeelandnet.nl>

Jorge Sereno wrote:

Well, you correlate the highest number of sunspots with solar influences on global temperatures. Is this valid?

Actually you are quite right about the sunspots being a less than adequate indicator of solar influence on climate. The Svensmark effect and Theodor Landscheidt's work is quite clear on that. I only mention 1957 in the context that it was the biggest solar maximum as measured from sunspot counts only. I would have expected to see some impact from 1957 on global temperature even though it was not the most significant solar event when judged by other solar irradiance models.

Regards      John Daly

Subject: El Nino is a response
Date: Sat, 04 Mar 2000 19:46:34 -0700
From: Steve Hemphill <steve@hemphill.net>
To: daly@vision.net.au

Above, Jorge reports a very high correlation with a combination of SOI, Volcanic, Co2 and solar influences. However, Dr. Landscheidt has correlated SOI with solar output. When one includes a response with causes one can easily find a high correlation, as the response masks the unknowns. Each volcanic eruption has different constituents, exactly as soils in different places have different constituents. The same would be true of thermonuclear surface bursts. In my opinion the ability to "make up the difference" with SOI only strengthens the argument that SOI is, in fact, a response.

Also, Doug Hoyt mentions that pyrheliometers are insensitive to water vapor. What is the effect and longevity of water vapor in the stratosphere? How about ice crystals in the stratosphere? It seems ice crystals have some unique reflective and refractive properties. How would these affect the stratosphere?

Steve

Subject: The Ghost of Winters Past
Date: Sun, 05 Mar 2000 15:18:35 +1100
From: "John L. Daly" <daly@vision.net.au
To: Ken Parish <kparish@legalnet.net.au>
CC: Doug Hoyt <dhoyt1@erols.com>

Dear Ken

I have followed up on Doug Hoyt's information and the observational evidence now seems clear.

The nuclear tests, numerous and powerful though they were, did not have an impact on temperatures (even at stations close to the test sites), and had no noticeable effect on the transparency of the atmosphere. Let me summarise what I have found so far -

On temperatures, we do find a general climatic cooling during the 1940-1976 period. As you pointed out in your paper, nuclear tests did not really come into the picture until the late 1940s (the 1945 bursts we agree being too small), meaning that the cooling was already underway prior to the main testing period. The worst of the tests were finished by 1963 (test ban treaty) and although the Chinese and French continued for some time afterward, the scale of their tests came no-where near the US-USSR series.

Attached is a plot of 1959-1965 mean monthly temperatures for Vardo (northern Norway), not so far distant from the Novaya Zemlya test site (where the worst of the tests took place). As the graph shows, there is no noticeable effect on the seasonal trend of temperature.

I have looked up some of the Arctic graphs from other stations with no suspicious coolings at any of them. The only Arctic cooling which can be considered anomalous is in the late 1960s, 5 years after the US-Soviet tests finished.

As to other stations I have looked at, I cannot see anything out of the ordinary, just the usual ebbs and flows of natural variability.

Next we come to Doug Hoyt's information. I looked into this also and found the pyrheliometers he refers to have been monitoring the transparency of the atmosphere at several sites around the world from before WW2 and while the volcanic eruptions of Mt Agung, El Chichon and Pinatubo show up clearly, the tests do not show up at all. If they are present, they are well hidden in the `noise'.

These instruments measure radiation from the sun on a broadband basis, covering both visible and near infra-red wavelengths. Their only `blind spot' appears to be in the UV-B and UV-C regions, which represent less than 1% of total solar radiation (Doug might want to correct me there). In other words, aerosols, dust and other particles would show up on these instruments, but have failed to do so in the case of the tests. See the attached plot from Mauna Loa to see what I mean.

In other words, unless the tests released a cocktail of invisible gases which only absorbed UV-B and UV-C, there seems little basis to conclude there was any climatic impact from the tests at all. I always say that observational evidence must take precedence over theory.

But this now higlights an entirely new issue -

Your hypothesis that the tests would cool the climate was a perfectly reasonable one on two counts -

1) It accords with `common sense'. Your catalogue of the tests would suggest to any reasonable person that a climatic effect would ensue. Indeed, I felt the same way on reading your paper.

2) Your hypothesis was fully consistent with the `Nuclear Winter' theory promulgated by the modellers and the climate community in the early 1980s. If a nuclear war could produce a `Nuclear Winter', is it not reasonable to assume that intense testing over a short period of time (esp. the early 1960s) would produce not exactly a `winter' but certainly a measureable cooling.

There was no shortage of scientific endorsement for the `Nuclear Winter' scenario as the following scientific references will show, all of them presumably peer-reviewed -

Paul J. Crutzen and John W. Birks, "The Atmosphere After a Nuclear War: Twilight at Noon", Ambio, Vol 11, No 2-3, p 114, 1982. Reprinted in Jeannie Peterson, Ed, The Aftermath: The Human and Ecological Consequences of Nuclear War, New York, Pantheon, 1983.

R.P. Turco, O.B. Toon, T.P. Ackerman, J.B. Pollack, Carl Sagan, "Nuclear Winter: Global Consequences of Multiple Nuclear Explosions", Science, V. 222, No; 4630, December 23, 1983.

Paul Ehrlich et al., "Long-Term Biological Consequences of Nuclear War," Science, V. 222, No. 4630, December 23, 1983.

Michael MacCracken, "Nuclear War: Preliminary estimates of the climatic effects of a nuclear exchange", paper presented at Third International Conference on Nuclear War, Erice, Sicily, August 19-23, 1983.

Curt Covey, Stephen H. Schneider, Starley L. Thompson, "Global Atmospheric Effects of Massive Injections from a Nuclear War: Results from General Circulation Model Simulations," Nature, Vol 308, No 5954, March 1, 1984.

S.L. Thompson, V.V. Aleksandrov, G.L. Stenchikov, S.H. Schneider, C. Covey and R.M. Chervin, "Global Climatic Consequences of Nuclear War: Simulations with Three Dimensional Models," Ambio, Vol 13, No 4, p 236, 1984.

So what do we have here? As the observational evidence clearly indicated, the nuclear bursts had no measureable effect on atmospheric transparency. This evidence was readily available to the above authors before promoting the `Nuclear Winter' scenario via their papers and all the media publicity that went with it.

`Nuclear Winter' filled the media headlines until the demise of the Cold War (with the rise of Mikhail Gorbachev in the USSR). By then, the public focus of the models had shifted from `Nuclear Winter' to `Global Warming'. This facilitated a rapid expansion of modelling activity, the continued financial support for which was underpinned by public fears arising from both scenarios.

But how soundly based was the science of `Nuclear Winter'? It was endorsed by the climate modelling community and I understand even by the NRC. And yet it seems to me incomprehensible that all this scientific effort did not give sufficient weight to the atmospheric transparency data collected during the testing period. In other words the modellers presented an alarming scenario which was contradicted by instrumental evidence readily available to them at the time.

It does raises questions as to whether the modellers are approaching Greenhouse in a similar way - discounting observed evidence (eg. lack of polar warming etc.) and just continuing to experiment with their models regardless to what the real world climate is doing, a case of making theory take precedence over observed evidence. The `Nuclear Winter' is not something the modellers should be proud of today.

As a postscript to this, Richard Turco of the Univ. of California (one of the modellers listed above) stated to "New Scientist" (12-Jan-91) that if Saddam Hussein blew up the Kuwait oil wells, there would be an adverse climatic reaction similar to that of a Nuclear Winter, albeit less severe. When all the Kuwaiti oil wells were blown up shortly afterward, injecting masses of black smoke continuously into the atmosphere for many months, his predictions failed to materialise. There was no such global or regional effect, just a local one.

Regards     John Daly

Subject: Big Bangs
Date: Sun, 5 Mar 2000 08:26:13 +0100
From: "Jorge Sereno" <sereno@zeelandnet.nl>
To: John Daly <daly@vision.net.au>

Dear John,

Actually, there is possibly some influence. However, when does it occur? What is the time lag of global temperatures on solar influences? Doug Hoyt thinks may be 3-4 years (personal communication) and Judith Lean thinks 5-10 years. For the "Ken Parish" period, a best fit occurs if we say 2 years. Actually, for the whole period after 1950, 2 years is the best fit. For the period 1868-2000 the correlatation is still very good (R=0,875), but a six year timelag produces a better fit (R=0,90). This all based on yearly averages and with my models, which combine SOI, Optical depth and greenhouse gas forcings, as I mentioned before.

As I mentioned in my first repsonse to Ken, all years between 1957 and 1963 are among the 11 warmest in the period 1950-1976 and most of them, within the top 6 (CRU/GISS combined dataset).

Best regards,    Jorge

<off topic, but is in response to some points by Theodor Landscheidt>

Subject: Big Bangs
Date: Sun, 5 Mar 2000 13:09:22 +0100
From: "Jorge Sereno" <sereno@zeelandnet.nl>
To: Theodor Landscheidt   c/o <daly@vision.net.au>

Dear Theodor,

While your graph looks o.k. for the last 140 years, an amplitude of 0,7K for this period only caused by solar changes does not appear to be valid according to me. If we elaborate this back with two solar models (Hoyt/Schatten and lean and Foukal) for the period 1700-2000 and compare it with proxy studies like those of Briffa et al and Mann et al, the correlation diminishes significantly. We get unexplainable lows just after 1800 (of more than 0,4K!) and unexplainable highs near 1770-1790 (about the same size, but +0,4K). Judith Lean concludes that 0,51K or somewhat less is the maximum influence for the 1600-2000 period. I think this is closer to the truth. This leads to about 0,3K for the period 1868-2000.

The larger peak 1960 in your graph is a result of overestimating solar influences. Also, the temperature rise for the period 1900-1960 is much larger in your graph as it is in reality.

If we add volcanic influences in your model (which I think we should), the rise starting from 1980 will be lower. Reducing solar influences near 1960 leads to a much less steep rise from 1975-2000. (if we calculate solar influences linearly). Then, the correlation found by you will diminish there. We need another factor, which will be (in my opinion) antropogenic factors. Not as large as the IPCC and almost all greenhouse models I am familiar with (with SO2 in it) want us to believe, but it isn't negligeble nevertheless. I come to 0,8-1,0K for a doubling near 2070, so the current influence is about 0,3K (since 1900). My models (and yes, I will send them to you whenever WP Office 2000 is letting me import Quattro Pro graphs in WordPerfect 9.0) have no problem with the peak near 1960 nor with the one we are experiencing right now.

Am I right? Well, I get very good correlations, just like you. This nevertheless does not prove my point. I hope you and others will be the judge in the near future.

best regards,     Jorge

Subject: My final contribution (probably)
Date: Sun, 5 Mar 2000 17:16:24 +0930
From: Ken Parish <kparish@legalnet.net.au>
To: "John Daly" <daly@vision.net.au>

Dear John,

I haven't yet had time to chase up the Kondratyev and Nikolsky article referred to by Jorge Sereno. However, I will deal with the most recent contributions anyway.

First, I think I am just about ready to concede that it appears nuclear testing may not have had a significant climate effect after all. However, there is one scenario that may still be open. It is as follows. Immediate fairly localised (mostly tropospheric) effects cancel each other out i.e the heat generated by the blast is countered by a short term cooling effect from troposheric dust etc. There may still, however, be some generalised stratospheric cooling effects from ice clouds (if not dust), which would not be picked up by Douglas Hoyt's pyrheliometers. This possibility is not entirely far-fetched. We know that virtually all the US tests after 1954 were on or over the ocean, so they they would by definition throw mostly water into the stratosphere (as well as nitrates -see below and my previous contribution).

The Russian tests all took place at Novaya Zemlya which, as we know, is covered by snow and ice. The reading I have done over the last day or two indicates that what both the Americans and Russians were working on in the 50's and early 60's (after their first couple of tests) was how to use the awesome destructive power of thermonuclear weapons so as to maximise military casualties, while avoiding catastrohic fallout causing horrific long-term environmental damage and lots of "collateral damage" (read civilian deaths). In general, the way they tried to do this was by experimenting with airbursts at varying altitudes and using varying strength weapons. The "right" combination means that troops are killed by the shockwave over a reasonably precisely calculated area, but not much tropospheric fallout is created, because the blast fireball does not intersect the ground. They probably weren't quite so worried about stratospheric effects, because the fallout is much less dangerous by the time it drops (the most dangerous isotopes having very short half lives), and they probably didn't even consider temperature effects from stratospheric ice.

If this scenario is right, it is not all that surprising that we don't find any trace in Doug Hoyt's instruments, because they don't detect water vapour. Nevertheless, one would expect to pick up some localised effect in Russian temperature stations and, as you say, we don't. Therefore, one must conclude that there is most probably no net effect. This could be because ice clouds do not in fact usually form, or because of the countervailing effect of nitrous oxide (assuming that is what is injected into the stratosphere at the rate of 5,000 tons per megaton of yield (see my last contribution), or because any solid particles projected into the stratosphere are larger than 1 micron (larger dust particles do not have a cooling effect, and can even have a warming effect). In relation to your observations about the 'nuclear winter scenario, here is a further extract from the nuclear FAQ article linked in my original article:

5.2.2.2 Nuclear Winter

The famous TTAPS (Turco, Toon, Ackerman, Pollack, and Sagan) proposal regarding a potential "nuclear winter" is another possible occurrence. This effect is caused by the absorption of sunlight when large amounts of soot are injected into the atmosphere by the widespread burning of cities and petroleum stocks destroyed in a nuclear attack.

Similar events have been observed naturally when large volcanic eruptions have injected large amounts of dust into the atmosphere. The Tambora eruption of 1815 (the largest volcanic eruption in recent history) was followed by "the year without summer" in 1816, the coldest year in the last few centuries.

Soot is far more efficient in absorbing light than volcanic dust, and soot particles are small and hydrophobic and thus tend not to settle or wash out as easily.

Although the initial TTAPS study was met with significant skepticism and criticism, later and more sophisticated work by researchers around the world have confirmed it in all essential details. These studies predict that the amount of soot that would be produced by burning most of the major cities in the US and USSR would severly disrupt climate on a world-wide basis. The major effect would be a rapid and drastic reduction in global temperature, especially over land. All recent studies indicate that if large scale nuclear attack occur against urban or petrochemical targets, average temperature reductions of at least 10 degrees C would occur lasting many months. This level of cooling far exceeds any that has been observed in recorded history, and is comparable to that of a full scale ice age. In areas downwind from attack sites, the cooling can reach 35 degrees C. It is probable that no large scale temperature excursion of this size has occurred in 65 million years.

With due respect to Richard Turco who erroneously predicted it, I would not have expected any long-term cooling effects from Saddam Hussain blowing up all the Kuwaiti oil wells. The "nuclear winter" scenario depends on soot from oil fires (generated by the bombing of large cities) being propelled into the stratosphere by the force of a large thermonuclear weapon. The soot (as opposed to ordinary dust) is said to be much more persistent and a much more effective backscatterer of solar radiation. Conventional explosions (even large ones) do not penetrate the tropopause and therefore cannot have long-term climate effects. Thus the fact that Saddam's dastardly schemes did not bring about a pseudo-nuclear winter (or even a mild autumn) does not invalidate the "nuclear winter" scenario. However, this has little or nothing to do with the nuclear tests of the 50's and 60's, because they didn't occur over major cities with oil refineries, wells etc., and so there was no soot!

Finally, it seems to me that a number of interesting things have emerged from this discussion. First, it again demonstrates the value of "open review". Not only are wrong propositions (as mine seemingly is) quickly detected and corrected (which hopefully also occurs with peer review, although with some of the published pro-warming material you have to wonder), but there is also a real communal learning experience. It was apparent from this discussion that quite a few very knowledgeable climate scientists had never really analysed whether nuclear testing had any effect on climate, or why, despite the fact that it is clearly a very plausible possibility whose timing coincides closely with the observed mid-century cooling. It is still not clearly apparent (to me at least) why the testing did not have a measurable effect (assuming, which I agree is most likely, that is the case), and one hopes that one of your learned contributors may yet give us the full picture.

Secondly, the debate allows us to muse further on the inadequacy of climate modelling as a predictive tool. I have further considered Jorge Sereno's claims about his hopefully soon to be published research. It seems that he has managed to achieve a close correlation with real-life temperatures by combining the effects of CO2, solar irradiance and volcanic eruptions. I note that something quite similar, although not including the effects of volcanic eruptions, was done some years ago by Baliunas et al. See the article "Estimation of the solar fraction and Svensmark factor" by Peter Dietze found at the Daly website http://www.john-daly.com/fraction/fraction.htm , particularly the graph of the Baliunas comparative analysis found in Dietze's answer to Dr. Warren B. White in the open review section at the end of the paper.

Baliunas et al also managed to achieve a close correlation with observed reality, although no closer (as Dietze observes) than the correlation between solar forcing alone and the observed temperature i.e. you don't need any CO2 forcing effect at all to explain climate change (at least if you assume a slight warming bias in the surface record over the last few years, as recently discussed by John Daly and Vincent Gray, to explain the progressive small divergence between the surface record and solar irradiance over the last few years). Including CO2 forcing is an act of religious faith rather than science.

No doubt Jorge Sereno has also achieved a good correlation with the inclusion of the extra factor of volcanic eruptions (which unquestionably do have an effect, and which clearly have increased markedly since 1963). But the way in which all these modelling efforts achieve such a correlation is by making particular assumptions about the assumed temperature forcing effects of CO2, solar irradiance (and volcanic eruptions as well, in Sereno's case). Presumably, since Baliunas et al achieved a good correlation without considering volcanic effects, Sereno can only have achieved a similarly good correlation by increasing the assumed forcing effect of either CO2 or solar irradiance (or both).

One could equally easily accommodate claimed effects of nuclear testing, and adjust the graph to fit observed reality by juggling with the respective assumed forcing values. It is a very interesting exercise, but at the end of the day doesn't prove anything much. If solar irradiance alone adequately explains the whole of the 20th century pattern of climate change (and it does - see the Baliunas graph), why should we postulate significant climate effects for increasing CO2 levels, given that there are perfectly good and plausible reasons (discussed elsewhere on this website) for concluding that even quite large CO2 increases will not have any significant temperature effects? At end of the day, at least in our current state of knowledge, it is a question of religious faith rather than science. I will end my contributions to this debate with a quote (found elsewhere on this website):

"The improver of natural knowledge absolutely refuses to acknowledge authority, as such. For him, scepticism is the highest of duties; blind faith the one unpardonable sin." - Thomas H. Huxley

Regards     Ken Parish