Response of Shaviv and Veizer (with our comments)


Drs. Shaviv and Veizer have now responded to the statement of the 14 scientists on their GSA Today paper. Their responses are reproduced below, with our comments added.

Response by Dr. Nir Shaviv (italic) with our comments:

> I carefully read the letter written by the cosigners and was quite bewildered by it. First, it is strange that they decided to hold a scientific debate by press releases. While I am sure that this debate is of interest to the general public, the nature of the medium allows them to bring forth accusations without actually presenting any supporting evidence. This is hardly a scientific approach, made even worse with general claims, such as "methodically very doubtful". To counter such a claim, we would have to show that all the steps we have taken were done carefully. Obviously I cannot condense about 75 journal pages of detailed analysis, published in 6 articles, and reviewed by 10 referees. I will, however, show how specific concerns are invalid.

We agree that a proper scientific discussion needs to take place in the scientific journals, and to our knowledge one such journal article has been submitted and a second one is in preparation. As you know, such articles take many months to appear in print. We took the decision to make a public statement after careful consideration, because strong political claims were being made based on your publication.
We would like to point out that you published a media release yourself on 12 August, titled "Global warming not a man-made phenomenon". This starts as follows: "Global warming will not be helped much by efforts to reduce carbon dioxide emission into the atmosphere, say two scientists who have studied the matter". You are cited as saying: "The operative significance of our research is that a significant reduction of the release of greenhouse gases will not significantly lower the global temperature, since only about a third of the warming over the past century should be attributed to man."
Even if your scientific analysis were completely correct, your paper would have merely provided one intriguing piece of evidence pitted against many other studies that come to a different conclusion, and it would have been irresponsible to publish such a far-reaching statement in a press release, especially since your paper does not study 20th Century warming. After making such a strong claim about the "operative significance" of your work in the media, you can hardly be surprised that this will be scrutinised and criticised also in the public arena.

> For example, the statement that "The reconstruction of the cosmic radiation is based only on 42 meteorite finds, which are interpreted differently by other experts" demonstrates the above. First, the reconstruction is based on 50 not 42 meteorites. This is a minor point, but it demonstrates that they read only the first analysis published in a short letter and not the 30 page subsequent paper detailing the cosmic ray flux reconstruction. Second, they failed to mention that the same periodicity and phase is obtained in a totally independent way using astronomical observations on galactic spiral arm dynamics. One can thus throw away the meteoritic data altogether, yet still reach the same conclusion. This makes the cosmic ray flux reconstruction quite robust since it is based on two independent methods. Third, if there are "completely different interpretations by other experts", why have they not been published in the scientific media (or non-scientific for that matter). Apparently, no other consistent explanation of the data actually exists.

We wished to point out that the meteorite data on which the cosmic ray flux curve is based are relatively few and uncertain, especially considering that it is an indirect inference based on apparent age clustering of the meteorites that is used to construct this curve. Other specialists on meteorite research working on cosmic-ray effects interpret the clustering of CRE ages as the result of break-up processes on parent asteroids, Mars and Moon, where large numbers of meteorites were ejected. The models of galactic spiral arm dynamics are also rather uncertain and do not directly provide data on cosmic ray flux. For the lay-person newspaper reader, it is important to point out that your data may be sufficient to put forward a speculative new hypothesis about cosmic ray variations (which we welcome), but they are not sufficient to overrule conclusions based on the much more abundant and accurate climate data from the more recent past, let alone to call for policy revisions. We expect you will agree on this point.

> As another example, writing that "The two curves show a statistic connection only because the time scale of the cosmic data was stretched arbitrarily in such a way, until an agreement arises", is simply wrong. The cosmic ray flux reconstruction gives a periodicity of 143 +/-10 Myr. Since the reconstructed tropical temperature shows a periodicity of 135 +/-9 Myr, the two signals are consistent with each other, without a needless "arbitrary stretching". Interestingly, also, the phases of the signals are consistent to within the measurement error.

Here you state that the periods and phase found in the two data sets are similar. We do not disagree with that statement. We do disagree with the much further reaching claim in your paper, namely that the two data sets show a statistically significant correlation which can explain 66% of the temperature variance. Our calculations show that your original data sets show no significant correlation; this only arises after you artificially enhanced the correlation by stretching the time scale (transforming the blue curve into the red curve in the upper panel of your Fig. 2). We note that you do not dispute this.

> And a last example, writing that "Even if it were material, this cycle (with three degrees of warming over 70 million years) would cause a warming of around a millionth degree in 20 years" simply is not relevant. What we strived to demonstrate is that the cosmic rays appear to affect climate on geological time scales as a result of the changing galactic "geography". If the result is correct, then the implication to global warming is not through the slowly changing flux outside the solar system, but arises from modulations by the solar wind, which is known to affect the cosmic ray flux reaching Earth. In particular, the increased solar activity over the past century should have translated into a quick rise in global temperature, explaining more than half of the observed global warming.

We understand the mechanism that you propose. Our statement cited above explicitely refers not to your paper but to a newspaper article by Edgar Gärtner. This article states that your work explains why global temperature has been rising over the past 20 years even though solar activity has not; it claims that this is due to diminishing cosmic ray flux as the Earth is leaving the Saggitarius-Carina spiral arm. We think that we all agree this is nonsense; we cited this as an example for the exaggerated media claims that were made in conjunction with your paper, and which required a response (especially given that this particular claim was even debated in the German parliament, the Bundestag).
We would like to ask you in which scientific publication we can verify your claim made above, that increased solar activity should explain more than half of the observed 20th Century warming.

> I could continue, but I think I have made my point. An avid reader should not take for granted a single word that I or the cosigners write in non-scientific media. Instead, he/she should take a look at the scientific articles and counter-articles and judge with a critical mind. One should always look also at the error quoted in these scientific articles. It may prove illuminating. The cosigners write that the doubling of CO2 in IPCC models would result in a "1.5 to 4.5 deg Celsius" rise in temperature. In our paper, we find that the upper limit (at 90% confidence) for the doubling CO2 is about 1.5 degs (at least on the geological time scales). Our results are thus inconsistent only with the upper range of the IPCC claims.

You refer here to "IPCC models" and "IPCC claims". We would like to point out that the IPCC does not run models, but that it publishes reports that are reviews of the published scientific literature. It does that after an extremely thorough and open process of discussion, drafting, peer review and revision of the individual chapters, involving hundreds of scientists. This process exists to provide policy makers and the public with a balanced overview of the state of knowledge, and to avoid that individual and exaggerated claims, such as those in your press release, unduly influence the public or policy.

Your estimate of the effect of doubling CO2 is in fact the main point of critique in our statement, since the method by which you arrived at it is flawed. You arrive at this by a regression analysis of temperature and CO2 time series. Climatologists have long known that this is not feasible, which is why they have not applied this simple analysis to the existing other, more reliable time series of temperature and CO2, such as those from Antarctic ice cores. One reason is that other forcing factors vary in a statistically not independent manner with temperature and CO2, so that the basic precondition for a meaningful regression is not fulfilled.
Imagine performing your regression analysis to the Vostok data, to determine the relative roles of insolation variations (due to orbital cycles) and CO2 variations on glacial cycles. The result would yield almost no role of orbital cycles, yet an effect of CO2 that is far too large. This is because other factors (e.g., the growth of ice sheets with their albedo, as well as methane variations) covary with CO2 and cause a large part of the temperature response, but these factors are irrelevant to the question of CO2 doubling on a time scale of decades. Only when these factors are carefully considered, as in the article of Lorius et al. (Nature 1990), can an estimate of climate sensitivity be derived from paleodata. Lorius et al. concluded from the Vostok data that the climate sensitivity to CO2 doubling is 3-4 ºC, in line with independent estimates based on the known radiative forcing of CO2 and the physical understanding of feedbacks as coded in climate models.
Your analysis applies to the hundred-million-year time scale and to a climate that for most of the time has far higher CO2 values than at present (between twice and ten times present levels). The position of continents also differs from the present. As is the case for the Vostok data, processes that operate on such time scales and at such high CO2 values will differ from those of interest for present climate. There are reasons to believe that the climate response could be much weaker than that to a CO2 doubling from current values. For example, due to the lack of ice, the ice-albedo feedback, one of the strongest amplifiers of climate change in a CO2 doubling scenario, would be much weaker for higher CO2 values.
We note that you did not respond to this fundamental issue, although this is the central point of our statement.

Response by Prof. Veizer (italic) with our comments:

> In a press release of October 24 (, a group of scientists denounced our publication in GSA Today (July 1) as "wissenschaftlich nich haltbar" and based "auf aeusserst fragwuerdigen Methoden" that should "in keiner Weise in Frage stellen" the "fundierte Wissen" about anthropogenic climate impact.
It concerns me that the debate stooped to this level, but it is symptomatic of the general atmosphere surrounding the climate change issues. A recent personal attack by the PIK cosigner on other scientists in Die Zeit is another example. I strive to refrain from the divisive public discussion of political ramifications (Kyoto) and would not conceive of attacking the scientific integrity of the IPCC-supporting scientists, despite the fact that these models too have a plethora of weak points (clouds, biology...) and yield predictions that are inconsistent with reality (balloons and satellites show no tropospheric temperature rise, the Antarctic and Greenland are mostly cooling...).

We would like to point out that we made no personal attacks. We only put forward strictly scientific arguments, which we believe are correct also after considering your response. We deemed it necessary to present these scientific arguments after exaggerated and irresponsible political claims were repeatedly being made with reference to your work. We would have much preferred it if you had made it clear yourself in public that your paper should not be used to draw political conclusions about the Kyoto protocol or other climate policy measures.
We would also like to point out again that the IPCC is a body which summarises published journal articles; its conclusions are based on all published work including yours and including all work in paleoclimatology and data analysis, not specifically on modeling work.

> Note also that in the Antarctic ice cores, cited by the signatories as the prime confirmation of the greenhouse theory, the CO2 commences to rise only centuries after the temperature has risen. Nonetheless, in the GSA article we still treated this cause/effect issue as an open question, striving not to belittle the research that attributes climate change to greenhouse gases. Regrettably, such is often the treatment of ideas exploring alternative scenarios.

We did not cite the Antarctic ice cores as the prime confirmation of greenhouse theory - we cited these ice cores as an illustration of why the kind of regression analysis performed in your paper does not yield correct results. We are not aware of any research that would suggest that greenhouse gases are the prime cause of the glacial cycles (or of shorter-term, millennial climate variations) seen in these ice cores. There is a wide consensus amongst climatologists that the glacial cycles are primarily caused by insolation variations due to orbital variations, which are amplified by various feedbacks including the build-up of continental ice sheets. It is hence to be expected that CO2 follows these temperature variations with a time lag that depends on the response time of the carbon cycle. None of this yields direct information on the feedback effect of these CO2 variations on climate.

> Arguing that our research is being "misused" by others is hardly a justification for personal attack on our scientific integrity. After all, is not the research of the signatories utilised in exactly the same way, albeit for an alternative view?

Far from attacking anyone on a personal level, we only criticise a few specific statements in one publication, and with good reason. Any scientist who makes such high-profile statements and operates as a public figure in the media can be expected to accept reasoned scientific criticism of their statements, not just in scientific journals but also in public discourse. We are subject to that as well.
Whenever our own research is misused for ill-founded or exaggerated claims in the public, in one sense or the opposite, we protest and try to rectify the matter. We believe it is very important that the public perception of science reflects in the most accurate and balanced way possible the actual state of knowledge, discussion and uncertainty that exists within the scientific community. That is why we deemed a public statement necessary in this case.

> My coauthor, Dr Shaviv, has responded to the points that concern the astrophysical issues. As for the geological part, it involved years of data gathering by many researchers, financed mostly by the Leibniz prize of the DFG, and the curve is based on about 4500 hard won measurements. Moreover, ours is the only comprehensive Phanerozoic database presently available, yielding a result in good agreement with the actual climate patterns deduced from sediments (see In contrast, the CO2 model (IPCC, page 40, figure 10e,f) is completely at odds with actual observations. Yet, it is the thousands of measurements corroborated by observations that are "fragwuerdig", while the theoretical construct that is at odds with reality merits inclusion in the IPCC summary chapter.

In our statement we merely point out that CO2 reconstructions so far back in time are still highly uncertain. You state this yourself in your paper, and you include three very different CO2 reconstructions in your Fig. 1. Hence we think we can agree on this point.

> The insinuation that our curves were shifted around until a fit was obtained is just that, unfounded insinuation.

As discussed above, we refer to the shift in curves from your blue to your red cosmic ray curve, which you describe as "fine tuning to best fit the low latitude temperature anomaly" in the caption to your Fig. 2. Without this "fine tuning", which shifts the oldest peak by about 20 million years, there is no significant correlation.

> Dr Shaviv and ourselves published our research and curves independently in top refereed journals, totally ignorant of each other's work. As for the rest, I invite the readers to peruse our publication and decide for themselves what we have actually claimed, and what is being imputed to us by the cosigners.
As scientists, we are not infallible and may eventually be proven wrong, but this should be done by factual science and not by denunciation of our scientific integrity in press releases.

As scientists, we are indeed not infallible. Hence, a thorough and open process of discussion and review of all relevant results needs to be performed; only the conclusions that stand up to such debate and scrutiny should be used to inform public policy. Such a process is facilitated by the IPCC. We believe that an individual publication, especially if it is at odds with the work of many other scientists, should not have been launched with a press release that suggests immediate policy implications. Nevertheless, we refrained from criticising such personal behaviour. Our statement focussed entirely on scientific arguments, since what ultimately counts is the science.

The original statement, signed by 14 scientists from Switzerland and Germany, can be found at