Guest commentary from Ben Santer
Part 2 of a series discussing the recent Guardian articles
A recent story by Fred Pearce in the February 9th online edition of the Guardian (“Victory for openness as IPCC climate scientist opens up lab doors”) covers some of the more publicized aspects of the last 14 years of my scientific career. I am glad that Mr. Pearce’s account illuminates some of the non-scientific difficulties I have faced. However, his account also repeats unfounded allegations that I engaged in dubious professional conduct. In a number of instances, Mr Pearce provides links to these allegations, but does not provide a balanced account of the rebuttals to them. Nor does he give links to locations where these rebuttals can be found. I am taking this opportunity to correct Mr. Pearce’s omissions, to reply to the key allegations, and to supply links to more detailed responses.
Another concern relates to Mr. Pearce’s discussion of the “openness” issue mentioned in the title and sub-title of his story. A naïve reader of Mr. Pearce’s article might infer from the sub-title (“Ben Santer had a change of heart about data transparency…”) that my scientific research was not conducted in an open and transparent manner until I experienced “a change of heart”.
This inference would be completely incorrect. As I discuss below, my research into the nature and causes of climate change has always been performed in an open, transparent, and collegial manner. Virtually all of the scientific papers I have published over the course of my career involve multi-institutional teams of scientists with expertise in climate modeling, the development of observational datasets, and climate model evaluation. The model and observational data used in my research is not proprietary – it is freely available to researchers anywhere in the world.
The 1995 IPCC Report: The “scientific cleansing” allegation
Mr. Pearce begins by repeating some of the allegations of misconduct that arose after publication (in 1996) of the Second Assessment Report (SAR) of the Intergovernmental Panel on Climate Change (IPCC). These allegations targeted Chapter 8 of the SAR, which dealt with the “Detection of Climate Change, and Attribution of Causes”. The IPCC SAR reached the historic finding that “The balance of evidence suggests a discernible human influence on global climate”. Information presented in Chapter 8 provided substantial support for this finding.
I served as the Convening Lead Author (CLA) of Chapter 8. There were three principal criticisms of my conduct as CLA. All three allegations are baseless. They have been refuted on many occasions, and in many different fora. All three allegations make an appearance in Mr. Pearce’s story, but there are no links to the detailed responses to these claims.
The first allegation was that I had engaged in “scientific cleansing”. This allegation originated with the Global Climate Coalition (GCC) – a group of businesses “opposing immediate action to reduce greenhouse gas emissions”.
In May 1996, a document entitled “The IPCC: Institutionalized ‘Scientific Cleansing’?” was widely circulated to the press and politicians. In this document, the Global Climate Coalition claimed that after a key Plenary Meeting of the IPCC in Madrid in November 1995, all scientific uncertainties had been purged from Chapter 8. The GCC’s “scientific cleansing” allegation was soon repeated in an article in Energy Daily (May 22, 1996) and in an editorial in the Washington Times (May 24, 1996). It was also prominently featured in the World Climate Report, a publication edited by Professor Patrick J. Michaels (June 10, 1996).
This “scientific cleansing” claim is categorically untrue. There was no “scientific cleansing”. Roughly 20% of the published version of Chapter 8 specifically addressed uncertainties in scientific studies of the causes of climate change. In discussing the “scientific cleansing” issue, Mr. Pearce claims that many of the caveats in Chapter 8 “did not make it to the summary for policy-makers”. This is incorrect.
The Summary for Policymakers (SPM) of the IPCC SAR is four-and-a-half pages long. Roughly one page of the SPM discusses results from Chapter 8. The final paragraph of that page deals specifically with uncertainties, and notes that:
“Our ability to quantify the human influence on global climate is currently limited because the expected signal is still emerging from the noise of natural variability, and because there are uncertainties in key factors. These include the magnitude and patterns of long term natural variability and the time-evolving pattern of forcing by, and response to, changes in concentrations of greenhouse gases and aerosols, and land surface changes”.
Contrary to Mr. Pearce’s assertion, important caveats did “make it to the summary for policy-makers”. And the “discernible human influence” conclusion of both Chapter 8 and the Summary for Policymakers has been substantiated by many subsequent national and international assessments of climate science.
There were several reasons why Chapter 8 was a target for unfounded “scientific cleansing” allegations. First, the Global Climate Coalitions’s “scientific cleansing” charges were released to the media in May 1996. At that time, Cambridge University Press had not yet published the IPCC Second Assessment Report in the United States. Because of this delay in the Report’s U.S. publication, many U.S. commentators on the “scientific cleansing” claims had not even read Chapter 8 – they only had access to the GCC’s skewed account of the changes made to Chapter 8. Had the Second Assessment Report been readily available in the U.S. in May 1996, it would have been easy for interested parties to verify that Chapter 8 incorporated a fair and balanced discussion of scientific uncertainties.
Second, the “pre-Madrid” version of Chapter 8 was the only chapter in the IPCC Working Group I Second Assessment Report to have both an “Executive Summary” and a “Concluding Summary”. As discussed in the next section, this anomaly was partly due to the fact that the Lead Author team for Chapter 8 was not finalized until April 1994 – months after all other chapters had started work. Because of this delay in getting out of the starting blocks, the Chapter 8 Lead Author team was more concerned with completing the initial drafts of our chapter than with the question of whether all chapters in the Working Group I Report had exactly the same structure.
The reply of the Chapter 8 Lead Authors to the Energy Daily story of May 22, 1996 pointed out this ‘two summary’ redundancy, and noted that:
“After receiving much criticism of this redundancy in October and November 1995, the Convening Lead Author of Chapter 8 decided to remove the concluding summary. About half of the information in the concluding summary was integrated with material in Section 8.6. It did not disappear completely, as the Global Climate Coalition has implied. The lengthy Executive Summary of Chapter 8 addresses the issue of uncertainties in great detail – as does the underlying Chapter itself.”
The removal of the concluding summary made it simple for the Global Climate Coalition to advance their unjustified “scientific cleansing” allegations. They could claim ‘This statement has been deleted’, without mentioning that the scientific issue addressed in the deleted statement was covered elsewhere in the chapter.
This was my first close encounter of the absurd kind.
The 1995 IPCC Report: The “political tampering/corruption of peer-review” allegation
The second allegation is that I was responsible for “political tampering”. I like to call this “the tail wags the dog” allegation. The “tail” here is the summary of the Chapter 8 results in the IPCC Summary for Policymakers, and the “dog” is the detailed underlying text of Chapter 8.
In November 1995, 177 government delegates from 96 countries spent three days in Madrid. Their job was to “approve” each word of the four-and-a-half page Summary for Policymakers of the IPCC’s Working Group I Report. This was the report that dealt with the physical science of climate change. The delegates also had the task of “accepting” the 11 underlying science chapters on which the Summary for Policymakers was based. “Acceptance” of the 11 chapters did not require government approval of each word in each chapter.
This was not a meeting of politicians only. A number of the government delegates were climate scientists. Twenty-eight of the Lead Authors of the IPCC Working Group I Report – myself included – were also prominent participants in Madrid. We were there to ensure that the politics did not get ahead of the science, and that the tail did not wag the dog.
Non-governmental organizations – such as the Global Climate Coalition – were also active participants in the Madrid meeting. NGOs had no say in the formal process of approving the Summary for Policymakers. They were, however, allowed to make comments on the SPM and the underlying 11 science chapters during the first day of the Plenary Meeting (November 27, 1996). The Global Climate Coalition dominated the initial plenary discussions.
Most of the plenary discussions at Madrid focused on the portrayal of Chapter 8’s findings in the Summary for Policymakers. Discussions were often difficult and contentious. We wrestled with the exact wording of the “balance of evidence” statement mentioned above. The delegations from Saudi Arabia and Kuwait argued for a very weak statement, or for no statement at all. Delegates from many other countries countered that there was strong scientific evidence of pronounced a human effect on climate, and that the bottom-line statement from Chapter 8 should reflect this.
Given the intense interest in Chapter 8, Sir John Houghton (one of the two Co-Chairs of IPCC Working Group I) established an ad hoc group on November 27, 1996. I was a member of this group. Our charge was to review those parts of the draft Summary for Policymakers that dealt with climate change detection and attribution issues. The group was placed under the Chairmanship of Dr. Martin Manning of New Zealand, and included delegates from the U.S., the U.K., Canada, Kenya, the Netherlands, and New Zealand. Sir John Houghton also invited delegates from Saudi Arabia and Kuwait to participate in this ad hoc group. Unfortunately, they did not accept this invitation.
The ad hoc group considered more than just the portions of the Summary for Policymakers that were relevant to Chapter 8. The Dutch delegation asked for a detailed discussion of Chapter 8 itself, and of the full scientific evidence contained in it. This discussion took place on November 28, 1996.
On November 29, 1996, I reported back to the Plenary on the deliberations of the ad hoc group. The Saudi Arabian and Kuwaiti delegations – who had not attended any of the discussions of the ad hoc group, and had no first-hand knowledge of what had been discussed by the group – continued to express serious reservations about the scientific basis for the detection and attribution statements in the Summary for Policymakers.
On the final evening of the Madrid Plenary Meeting, debate focused on finding the right word to describe the human effect on global climate. There was broad agreement among the government delegates that – based on the scientific evidence presented in Chapter 8 – some form of qualifying word was necessary. Was the human influence “measurable”? Could it be best described as “appreciable”, “detectable”, or “substantial”? Each of these suggested words had proponents and opponents. How would each word translate into different languages? Would the meaning be the same as in English?
After hours of often rancorous debate, Bert Bolin (who was then the Chairman of the IPCC) finally found the elusive solution. Professor Bolin suggested that the human effect on climate should be described as “discernible”.
Mr. Pearce – who was not present at the Madrid Plenary Meeting – argues that the discussion of human effects on climate in the IPCC Summary for Policymakers “went beyond what was said in the chapter from which the summary was supposedly drawn”. In other words, he suggests that the tail wagged the dog. This is not true. The “pre-Madrid” bottom-line statement from Chapter 8 was “Taken together, these results point towards a human influence on climate”. As I’ve noted above, the final statement agreed upon in Madrid was “The balance of evidence suggests a discernible human influence on global climate”.
Is “suggests” stronger than “points towards”? I doubt it. Is “The balance of evidence” a more confident phrase than “Taken together”? I don’t think so.
The primary difference between the pre- and post-Madrid statements is that the latter includes the word “discernible”. In my American Heritage College Dictionary, “discernible” is defined as “perceptible, as by vision or the intellect”. In Merriam-Webster’s Online Dictionary, one of the three meanings of the verb “discern” is “to recognize or identify as separate and distinct”. Was the use of “discernible” justified?
The answer is clearly “yes”. Chapter 8 of the IPCC’s Second Assessment Report relied heavily on the evidence from a number of different “fingerprint” studies. This type of research uses rigorous statistical methods to compare observed patterns of climate change with results from climate model simulations. The basic concept of fingerprinting is that each different influence on climate – such as purely natural changes in the Sun’s energy output, or human-caused changes in atmospheric levels of greenhouse gases – has a unique signature in climate records. This uniqueness becomes more apparent if one looks beyond changes averaged over the entire globe, and instead exploits the much greater information content available in complex, time-varying patterns of climate change.
Fingerprinting has proved to be an invaluable tool for untangling the complex cause-and-effect relationships in the climate system. The IPCC’s Second Assessment Report in 1995 was able to draw on fingerprint studies from a half-dozen different research groups. Each of these groups had independently shown that they could indeed perceive a fingerprint of human influence in observed temperature records. The signal was beginning to rise out of the noise, and was (using Merriam-Webster’s definition of “discern”) “separate and distinct” from purely natural variations in climate.
Based on these fingerprint results, and based on the other scientific evidence available to us in November 1995, use of the word “discernible” was entirely justified. Its use is certainly justified based on the scientific information available to us in 2010. The “discernible human influence” phrase was approved by all of the 177 delegates from 96 countries present at the Plenary Meeting – even by the Saudi and Kuwaiti delegations. None of the 28 IPCC Lead Authors in attendance at Madrid balked at this phrase, or questioned our finding that “the balance of evidence suggests a discernible human influence on global climate”. The latter statement was cautious and responsible, and entirely consistent with the state of the science. The much more difficult job of trying to quantify the size of human influences on climate would be left to subsequent IPCC assessments.
Mr. Pearce’s remarks suggest that there is some substance to the “political tampering” allegation – that I was somehow coerced to change Chapter 8 in order to “reflect the wording of the political summary”. This is untrue. There was no political distortion of the science. If Mr. Pearce had been present at the Madrid Plenary Meeting, he would have seen how vigorously (and successfully) scientists resisted efforts on the part of a small number of delegates to skew and spin some of the information in the Summary for Policymakers.
The key point here is that the SPM was not a “political summary” – it was an accurate reflection of the science. Had it been otherwise, I would not have agreed to put my name on the Report.
A reader of Mr. Pearce’s article might also gain the mistaken impression that the changes to Chapter 8 were only made in response to comments made by government delegates during the Madrid Plenary Meeting. That is not true. As I’ve mentioned above, changes were also made to address government comments made during the meeting of the ad hoc group formed to discuss Chapter 8.
Furthermore, when I first arrived in Madrid on November 26, 1995, I was handed a stack of government and NGO comments on Chapter 8 that I had not seen previously. I had the responsibility of responding to these comments.
One reason for the delay in receiving comments was that the IPCC had encountered difficulties in finding a Convening Lead Author (CLA) for Chapter 8. To my knowledge, the CLA job had been turned down by at least two other scientists before I received the job offer. The unfortunate consequence of this delay was that, at the time of the Madrid Plenary Meeting, Chapter 8 was less mature and polished than other chapters of the IPCC Working Group I Report. Hence the belated review comments.
The bottom line in this story is that the post-Madrid revisions to Chapter 8 were made for scientific, not political reasons. They were made by me, not by IPCC officials. The changes were in full accord with IPCC rules and procedures (pdf). Mr. Pearce repeats accusations by Fred Seitz that the changes to Chapter 8 were illegal and unauthorized, and that I was guilty of “corruption of the peer-review process”. These allegations are false, as the IPCC has clearly pointed out.
The 1995 IPCC Report: The “research irregularities” allegation
The third major front in the attack on Chapter 8 focused on my personal research. It was a two-pronged attack. First, Professor S. Fred Singer claimed that the IPCC’s “discernible human influence” conclusion was entirely based on two of my own (multi-authored) research papers. Next, Professor Patrick Michaels argued that one of these two papers was seriously flawed, and that irregularities had occurred in the paper’s publication process. Both charges were untrue.
On July 25, 1996, I addressed the first of these allegations in an email to the Lead Authors of the 1995 IPCC Report:
“Chapter 8 references more than 130 scientific papers – not just two. Its bottom-line conclusion that “the balance of evidence suggests a discernible human influence on global climate” is not solely based on the two Santer et al. papers that Singer alludes to. This conclusion derives from many other published studies on the comparison of modelled and observed patterns of temperature change – for example, papers by Karoly et al. (1994), Mitchell et al. (1995), Hegerl et al. (1995), Karl et al. (1995), Hasselmann et al. (1995), Hansen et al. (1995) and Ramaswamy et al. (1996). It is supported by many studies of global-mean temperature changes, by our physical understanding of the climate system, by our knowledge of human-induced changes in the chemical composition of the atmosphere, by information from paleoclimatic studies, and by a wide range of supporting information (sea-level rise, retreat of glaciers, etc.). To allege, as Singer does, that “Chapter 8 is mainly based on two research papers” is just plain wrong”.
In the second prong of the attack, Professor Michaels claimed that a paper my colleagues and I had published in Nature in 1996 had been selective in its use of observational data, and that our finding of a human fingerprint in atmospheric temperature data was not valid if a longer observational record was used. Further, he argued that Nature had been “toyed with” (presumably by me), and coerced into publishing the 1996 Santer et al. Nature paper one week prior to a key United Nations meeting in Geneva.
My colleagues and I immediately addressed the scientific criticism of our Nature paper by Michaels and his colleague Chip Knappenberger. We demonstrated that this criticism was simply wrong. Use of a longer record of atmospheric temperature change strengthened rather than weakened the evidence for a human fingerprint. We published this work in Nature in December 1996. Unfortunately, Mr. Pearce does not provide a link to this publication.
Since 1996, studies by a number of scientists around the world have substantiated the findings of our 1996 Nature paper. Such work has consistently shown clear evidence of a human fingerprint in atmospheric temperature records.
Disappointingly, Professor Michaels persists in repeating his criticism of our paper, without mentioning our published rebuttal or the large body of subsequently published evidence refuting his claims. Michaels’ charge that Nature had been “toyed with” was complete nonsense. As described below, however, this was not the last time I would be falsely accused of having the extraordinary power to force scientific journals to do my bidding.
A Climatology Conspiracy? More “peer-review abuse” accusations
Mr. Pearce also investigates a more recent issue. He implies that I abused the normal peer-review system, and exerted pressure on the editor of the International Journal of Climatology to delay publication of the print version of a paper by Professor David Douglass and colleagues. This is not true.
The Douglass et al. paper was published in December 2007 in the online edition of the International Journal of Climatology. The “et al.” included the same Professor S. Fred Singer who had previously accused me of “scientific cleansing”. It also included Professor John Christy, the primary developer of a satellite-based temperature record which suggests that there has been minimal warming of Earth’s lower atmosphere since 1979. Three alternate versions of the satellite temperature record, produced by different teams of researchers using the same raw satellite measurements, all indicate substantially more warming of the Earth’s atmosphere.
The focus of the Douglass et al. paper was on post-1979 temperature changes in the tropics. The authors devised what they called a “robust statistical test” to compare computer model results with observations. The test was seriously flawed (see Appendix A in Open Letter to the Climate Science Community: Response to A “Climatology Conspiracy?”). When it was applied to the model and observational temperature datasets, the test showed (quite incorrectly) that the model results were significantly different from observations.
As I have noted elsewhere, the Douglass et al. paper immediately attracted considerable media and political attention. One of the paper’s authors claimed that it represented an “inconvenient truth”, and proved that “Nature, not humans, rules the climate”. These statements were absurd. No single study can overturn the very large body of scientific evidence supporting “discernible human influence” findings. Nor does any individual study provide the sole underpinning for the conclusion that human activities are influencing global climate.
Given the extraordinary claims that were being made on the basis of this incorrect paper, my colleagues and I decided that a response was necessary. Although the errors in Douglass et al. were easy to identify, it required a substantial amount of new and original work to repeat the statistical analysis properly.
Our work went far beyond what Douglass et al. had done. We looked at the sensitivity of model-versus-data comparisons to the choice of statistical test, to the test assumptions, to the number of years of record used in the tests, and to errors in the computer model estimates of year-to-year temperature variability. We also examined how the statistical test devised by Douglass et al. performed under controlled conditions, using random data with known statistical properties. From their paper, there is no evidence that Douglass et al. considered any of these important issues before making their highly-publicized claims.
Our analysis clearly showed that tropical temperature changes in observations and climate model simulations were not fundamentally inconsistent – contrary to the claim of Douglass and colleagues. Our research was published on October 10, 2008, in the online edition of the International Journal of Climatology. On November 15, 2008, the Douglass et al. and Santer et al. papers appeared in the same print version of the International Journal of Climatology.
In December 2009, shortly after the public release of the stolen emails from the University of East Anglia’s Climatic Research Unit, Professors David Douglass and John Christy accused me of leading a conspiracy to delay publication of the print version of the Douglass et al. paper. This accusation was based on a selective analysis of the stolen emails. It is false.
In Mr. Pearce’s account of this issue, he states that “There is no doubt the (sic) Santer and his colleagues sought to use the power they held to the utmost…” So what are the facts of this matter? What is the “power” Fred Pearce is referring to?
- Fact 1: The only “power” that I had was the power to choose which scientific journal to submit our paper to. I chose the International Journal of Climatology. I did this because the International Journal of Climatology had published (in their online edition) the seriously flawed Douglass et al. paper. I wanted to give the journal the opportunity to set the scientific record straight.
- Fact 2: I had never previously submitted a paper to the International Journal of Climatology. I had never met the editor of the journal (Professor Glenn McGregor). I did not have any correspondence or professional interaction with the editor prior to 2008.
- Fact 3: Prior to submitting our paper, I wrote an email to Dr. Tim Osborn on January 10, 2008. Tim Osborn was on the editorial board of the International Journal of Climatology. I told Dr. Osborn that, before deciding whether we would submit our paper to the International Journal of Climatology, I wanted to have some assurance that our paper would “be regarded as an independent contribution, not as a comment on Douglass et al.” This request was entirely reasonable in view of the substantial amount of new work that we had done. I have described this new work above.
- Fact 4: I did not want to submit our paper to the International Journal of Climatology if there was a possibility that our submission would be regarded as a mere “comment” on Douglass et al. Under this scenario, Douglass et al. would have received the last word. Given the extraordinary claims they had made, I thought it unlikely that their “last word” would have acknowledged the serious statistical error in their original paper. As subsequent events showed, I was right to be concerned – they have not admitted any error in their work.
- Fact 5: As I clearly stated in my email of January 10 to Dr. Tim Osborn, if the International Journal of Climatology agreed to classify our paper as an independent contribution, “Douglass et al. should have the opportunity to respond to our contribution, and we should be given the chance to reply. Any response and reply should be published side-by-side…”
- Fact 6: The decision to hold back the print version of the Douglass et al. paper was not mine. It was the editor’s decision. I had no “power” over the publishing decisions of the International Journal of Climatology.
This whole episode should be filed under the category “No good deed goes unpunished”. My colleagues and I were simply trying to set the scientific record straight. There was no conspiracy to subvert the peer-review process. Unfortunately, conspiracy theories are easy to disseminate. Many are willing to accept these theories at face value. The distribution of facts on complex scientific issues is a slower, more difficult process.
Climate Auditing – Close Encounters with Mr. Steven McIntyre
Ten days after the online publication of our International Journal of Climatology paper, Mr. Steven McIntyre, who runs the “ClimateAudit” blog, requested all of the climate model data we had used in our research. I replied that Mr. McIntyre was welcome to “audit” our calculations, and that all of the primary model data we had employed were archived at Lawrence Livermore National Laboratory and freely available to any researcher. Over 3,400 scientists around the world currently analyze climate model output from this open database.
My response was insufficient for Mr. McIntyre. He submitted two Freedom of Information Act (FOIA) requests for climate model data – not for the freely available raw data, but for the results from intermediate calculations I had performed with the raw data. One FOIA request also asked for two years of my email correspondence related to these climate model data sets.
I had performed these intermediate calculations in order derive weighted-average temperature changes for different layers of the atmosphere. This is standard practice. It is necessary since model temperature data are available at specific heights in the atmosphere, whereas satellite temperature measurements represent an average over a deep layer of the atmosphere. The weighted averages calculated from the climate model data can be directly compared with actual satellite data. The method used for making such intermediate calculations is not a secret. It is published in several different scientific journals.
Unlike Mr. McIntyre, David Douglass and his colleagues (in their International Journal of Climatology paper) had used the freely available raw model data. With these raw datasets, Douglass et al. made intermediate calculations similar to the calculations we had performed. The results of their intermediate calculations were similar to our own intermediate results. The differences between what Douglass and colleagues had done and what my colleagues and I had done was not in the intermediate calculations – it was in the statistical tests each group had used to compare climate models with observations.
The punch-line of this story is that Mr. McIntyre’s Freedom of Information Act requests were completely unnecessary. In my opinion, they were frivolous. Mr. McIntyre already had access to all of the information necessary to check our calculations and our findings.
When I invited Mr. McIntyre to “audit” our entire study, including the intermediate calculations, and told him that all the data necessary to perform such an “audit” were freely available, he expressed moral outrage on his blog. I began to receive threatening emails. Complaints about my “stonewalling” behavior were sent to my superiors at Lawrence Livermore National Laboratory and at the U.S. Department of Energy.
A little over a month after receiving Mr. McIntyre’s Freedom of Information Act requests, I decided to release all of the intermediate calculations I had performed for our International Journal of Climatology paper. I made these datasets available to the entire scientific community. I did this because I wanted to continue with my scientific research. I did not want to spend all of my available time and energy responding to harassment incited by Mr. McIntyre’s blog.
Mr. Pearce does not mention that Mr. McIntyre had no need to file Freedom of Information Act requests, since Mr. McIntyre already had access to all of the raw climate model data we had used in our study (and to the methods we had used for performing intermediate calculations). Nor does Mr. Pearce mention the curious asymmetry in Mr. McIntyre’s “auditing”. To my knowledge, Mr. McIntyre – who purports to have considerable statistical expertise – has failed to “audit” the Douglass et al. paper, which contained serious statistical errors.
As the “Climategate” emails clearly show, there is a pattern of behavior here. My encounter with Mr. McIntyre’s use of FOIA requests for “audit” purposes is not an isolated event. In my opinion, Mr. McIntyre’s FOIA requests serve the purpose of initiating fishing expeditions, and are not being used for true scientific discovery.
Mr. McIntyre’s own words do not present a picture of a man engaged in purely dispassionate and objective scientific inquiry:
“But if Santer wants to try this kind of stunt, as I’ve said above, I’ve submitted FOI requests and we’ll see what they turn up. We’ll see what the journal policies require. I’ll also see what DOE and PCDMI administrators have to say. We’ll see if any of Santer’s buddies are obligated to produce the data. We’ll see if Santer ever sent any of the data to his buddies”
(Steven McIntyre; posting on his ClimateAudit blog; Nov. 21, 2008).
My research is subject to rigorous scrutiny. Mr. McIntyre’s blogging is not. He can issue FOIA requests at will. He is the master of his domain – the supreme, unchallenged ruler of the “ClimateAudit” universe. He is not a climate scientist, but he has the power to single-handedly destroy the reputations of exceptional men and women who have devoted their entire careers to the pursuit of climate science. Mr. McIntyre’s unchecked, extraordinary power is the real story of “Climategate”. I hope that someone has the courage to tell this story.
Benjamin D. Santer
John D. and Catherine T. MacArthur Fellow
San Ramon, California
February 22, 2010*
*These remarks reflect the personal opinions of Benjamin D. Santer. They do not reflect the official views of Lawrence Livermore National Laboratory or the U.S. Department of Energy. In preparing this document, I would like to acknowledge the assistance of Tom Wigley, Myles Allen, Kristin Aydt, Graham Cogley, Peter Gleckler, Leo Haimberger, Gabi Hegerl, John Lanzante, Mike MacCracken, Gavin Schmidt, Steve Sherwood, Susan Solomon, Karl Taylor, Simon Tett, and Peter Thorne.
Nice analogy with the rubber bands and weights!
Rod –
I don’t know about the emission and absorption from collision in the sense you describe; I would defer to others, but it does make sense to me that an acceleration of a dipole in some direction, no change in rotation, could emit some radiation. That radiation would be weak at any distance though…
(at infinite distance the effect goes to zero for the waves, hence the probability of photon emission should be low, but… photons once emitted shouldn’t depend on distance or lack-thereof for their existence (I don’t think quantum entanglement applies here???), but maybe the process that emits the photons should be regarded as occuring over a large volume in that case?,
…so it might be a minor source of radiation. … Unless the wavelength is comparable to the dipole size ?, which doesn’t apply to molecules for most solar and terrestrial radiation.
BUT what I do know is (and I suspect this is typically the dominant mechansim for continuum absorption and emission, since everything I’ve read only refers to this):
Spectral lines can be broadenned by three distinct mechanims:
Quantum uncertainty (not a large effect).
Doppler shifting (the individual atoms, molecules, crystal lattice cells, etc, are generally moving, with some component of that motion associated with nonzero internal energy – motions fit a statistical distribution with an average of zero velocity (if the macroscopic-scale motion is subtracted). Each particle or part of a particle (of the same type and condition) can have the same energy transition within it’s own frame of reference, but the effect for a single reference frame is a range of energy transitions.
Pressure/collisional broadenning – I’m a bit fuzzy on how this works, but I think it has to do with energy gained or lost from a collision combining with the energy transition.
Collisions can also produce new spectral lines in the same way that they are produced when matter in general is brought into a condensed form (liquid, solid, very dense plasma) – When an atom (or ion or …) has a neighboring atom (or ion or …), the electrons in one start to feel the effects of charges in the other when they are brought closer together (dipole fields decay as an inverse cube). This alters the energy levels. The energy levels are altered in such a way that the total number of states originally in the individual particles is the same but they take different values when combined (and some of those values are at lower energy, and electrons falling into those levels releases chemical energy). (Okay, you might wonder how this can be said, given that the number of states is actually infinite (I’m not just refering to the ground states and the excited states reached by your typical garden-variety photon). But there is a relationship between finite subsets of states between isolated and combined situations, such that the total number can be said to be the same.)
(I think momentary clusterings of water molecules give rise to additional absorption in the 8 to 12 micron range when the concentration is high enough via the last mechanism, which is not generally classified as a ‘broadenning’ mechanism.)
It deserves some EMPHASIS that:
The entire CO2 absorption spectrum is not the result of line broadenning from a single line near 15 microns. There are many, many individual lines of varying strength. Line broadenning fills in the gaps between the lines, to an extent varying with altitude.
(As it is with many gases in general)
Hank, you’re (still?) mixing up the two different convention categories of quantization in #996. The vibrating molecules with bound electrons in a crystalline solid or whole body gaseous molecular translation energies or free electrons are quantized vastly ala meter-kg pendulums. To the naked eye, or to the most precision measuring devices, and given a large enough population, it looks like, acts like, and smells like continuous radiation. But the electron internal energy levels or molecular vibration and rotation energy levels do not have vastly many possibilities and generate line spectra.
Hank (997), you left out some detail, but I agree with your summary. My continuous blackbody radiation generation or quanta of pendulums, etc. contradicts it.
Nor do I have a problem with you second example. Except for one detail in your first experiment — 12 balls hanging from 12 different rubber bands gives you 12 different oscillations, not a huge variety. In the second experiment, don’t just use 12 balls — use 10^30 balls, and have each one impinged by insolation at some specific frequency but all impinged after a while by all frequencies in the continuous spectrum of insolation (or downwelling IR, too). Now you will get a vast range of energy levels among the balls in a continuous spectrum for all practical purposes.
Rod, nope. You’re not reading the references on the difference between solid and gas behavior. Those electrons — in a solid — aren’t nicely arranged in those nifty little orbits, they’re — you can look this up. Read the section in the link I’ve given you twice explaining how light is reflected from a surface. It’s just above the Einstein paragraph. Where are the electrons? You can read it for yourself.
I’m giving you poetry, just paraphrasing what’s in the references.
You need the mathematics — or if you can’t follow the math, you need to decide whether you can rely on the people who do understand it.
If you can’t follow the math, and can’t believe what the people doing the math tell you, then what can you do?
Try it with poetry and logic — the physicists will correct the failings in mine as they’re trying to correct the failings in yours, this is all I can offer you. After this, I’m done until your next annual return to it anyhow.
Take liquid or solid iron, or carbon, or oxygen, or any other element.
It doesn’t need to be crystalline, just liquid or solid.
As long as it’s in solid or liquid form —
Where are the electrons?
Take an isolated atom of iron, or carbon, in vapor form.
Where are the electrons?
Take two atoms of oxygen as a diatomic molecule in gaseous form.
Where are the electrons?
“whole body gaseous molecular translation energies”
Which has nothing to do with temperature, Rod.
This is called “moving”.
“But the electron internal energy levels or molecular vibration and rotation energy levels do not have vastly many possibilities and generate line spectra.”
What do you mean “internal energy levels”??? Technobabble.
And the vastly many possibilities generate DIFFERENT LINE SPECTRA. A molecule in an excited state will relax but if it undergoes collision with another object during the time of emission, that collision will be inelastic and the energy of the emitted photon will be the energy of the transition state of that excited molecule PLUS some random component of the other molecule.
[edit]
Patrick, #1001, there’s also the interference of the electron WITH ITSELF! It’s called the Lamb Shift.
Pretty small.
PS the quantum effect depends on the time over which an emission takes place (mathematically, a short pulse of wave motion has a wider array of frequencies included to make that shape: fourier transform). This is why the second is defined by the number of wavelengths of a METASTABLE decay in Caesium: metastable means that it takes a long time to emit the photon, which means a very nearly monochromatic source, which means that you have a very good definition of where the wavelength is.
“Give me a break!. Please explain solar radiation and insolation. A little doppler action on the H atoms line spectrum perhaps??!!?”
Nope, that’s the effect of thermodynamic equilibrium.
Read up on stellar physics. It takes a full size book to turn someone from ignorant to the physics of stellar atmospheres into someone who knows what’s going on. And even that requires that the student want to learn.
very true, sidd (999); it’s just not the focus of the discussion.
David Warkentin, it seems you took a mid turn that I didn’t follow. I was discussing the difference between radiation from changes in plain and simple internal molecular energy levels (electron orbits, vibration and rotation) and that from temperature ala Planck. The discussion of energy bands, valence and conduction, which gets mixed up with ionizations and free or “semi-free” electrons, is not really relevant and certainly not helpful, at least to the specific discussion. I misinterpreted your minor shift to the latter, which, for example of course, is integral to semiconductor physics!
ps to my #1004: OOPS! Typo: should say doesn’t contradict, not just contradict…
CFU (1006), Huh? Whole body gaseous molecular translation energies has everything to do with temperature. It defines temperature.
Mr. Rod B writes on the 15th of March 2010 at 1:55 PM:
“the difference between radiation from changes in plain and simple internal molecular energy levels (electron orbits, vibration and rotation) and that from temperature ala Planck.”
The populations of the internal states are coupled to the temperature of the environment. If you choose you can define separate electronic, spin and other ‘internal’ temperatures, but in the vast majority of cases these will all be the same as the external temperature It is quite difficult to decouple these at NTP, even in the lab.
CFU (1008), Let me see. The H and He electrons change orbits probably induced by the heat. Those changing electrons emit a line spectral photon. That photon does a U-ey or something and performs some LTE magic and causes something (what?) to radiate at all frequencies between the spectral lines. Hummm. Can you explain that in plain physics or astrophysics? One thing in particular: how is it the H still has bound electrons? Please tell me the large book you get this info from so I can be sure to avoid it ;-)
“CFU (1008), Let me see. The H and He electrons change orbits probably induced by the heat”
Uh, orbits of WHAT?
“Those changing electrons emit a line spectral photon”
Nope, they emit a continuous or at least only bounded spectral photon.
“That photon does a U-ey or something and performs some LTE magic and causes something (what?)”
What’s a U-ey?
Nope, the photon doesn’t cause something. The photon gets absorbed and emitted many, many, MANY times.
If you don’t think this is possible, please explain the Sun.
“1012
Rod B says:
15 March 2010 at 2:13 PM
CFU (1006), Huh? Whole body gaseous molecular translation energies has everything to do with temperature.”
Nope, whole body molecular translation energies are movement.
Internal random motions within that CofM if there are enough bodies for statistical averages to be reliable estimates is expressed as temperature.
[edit]
PS I note that Rod has still not explained what “internal energy levels” means.
Forget my earlier point about dipoles; that same logic would seem to imply that even oscillating and rotating dipoles don’t emit any or much radiation, which is obviously not true; there is a distinction between the near field and far field effects; the near field is a little like the field of a static dipole in that it decays more rapidly with distance; my understanding is it has no time average Poynting vector. The magnetic field is caused by the current and the changing electric field from the dipole and the changing electric field caused by the changing magnetic field, that last part being the mechanism for the freely-propagating wave… something like that.
The reason why a an accelerating translation of a dipole would not involve emission (even though it could be modelled as a series of pulsed dipoles) must be something else. Of course, molecular collisions involve particles rebounding off each other’s fields, so their is transfer of momentum and energy via the fields there, but a single particle’s translational energy depends on frame of reference and so would the sign of energy gain for the same acceleration … … and a bunch of other stuff I probably don’t know.
Rod, please note there are line-broadenning mechanisms (‘natural/quantum’, doppler, pressure/collisional) and two identical electron states of isolated atoms/molecules will combine into a pair of states that will not be identical (and the difference varies with how close they are, and relative orientations where relevant, so crystal lattice vibrations could broaden those lines too, I’d expect (though in a quantized way…). And in an amorphous solid, or liquid, … etc.)
Patrick, you also have the bandpass electronic states in (semi-) conductor solids and the spectra of the energies of the photoelectric effect. The photon has to be ABOVE a certain energy to eject an electron but not a SPECIFIC energy or energies: the ejected electron is not quantised.
Rod, It sounds as if you are VERY confused. You are asking how the hydrogen atom has bound electrons? It’s a proton (positive) and an electron. You know, the Bohr atom. Nonetheless, I think it would be instructive for you to review this system–it’s easy to forget the details. In particular, remember that only the electron orbits that correspond to a standing wave (complete # of electron wavelengths) will be allowed. In these energy states, the electron does not radiate (it would be unstable otherwise). It is only in transitions between allowed energy states that you get radiation.
Likewise, neutral atoms and molecules only radiate in transitions between states corresponding to allowed energies–giving rise to spectral lines. In solids, the periodicity of the atoms perturbs the spectral lines and gives rise to allowed energy bands. These are also quantized, but the electrons/holes can have a continuum of energies within these bands. (Think of solid-state lasers–line emission.)
Again, Rod, I highly recommend the treatment of blackbody radiation by Landau and Lifshitz in their Stat. Mech. text. It is probably the clearest exposition of the phenomenon. Whatever reference you are using seems to be confusing you.
CFU, your #1015 post has no coherence. Ray said, “…there are no sources of radiation between the spectral lines.” I simply asked where the radiation of insolation between the sun’s spectral lines came from — solar radiation at TOA being almost a perfect continuous blackbody type. You said I just do not understand stellar physics. I asked you to then explain it to me, in other than HPFM.
re #1016: If you have trouble with 1/2mv^2 = 3/2kT go take it up with Boltzmann. I’m sure he can explain it better than I.
re #1017: you really want me to define internal energy? Boggles the mind, but O.K. Of interest to climate and similar stuff, the internal energies are molecular rotation, atom bond harmonics (vibration) and electron atomic orbits. When an electron’s internal energy exceeds ionization potential, then it is no longer technically considered as “internal”. There are other internal molecular energies in the chemical and nuclear/nucleus realm but are of no interest to climate studies.
Ray, my question to CFU was asking about the effect of the preponderance of H in the sun being ionized — a minority of electrons bound to the H nucleus (proton).
You still have not credibly explained the continuous spectrum emanating from the sun (other than one inference — how it sounded to me — that it doesn’t exist…??), or the cosmic background radiation for that matter, or the continuous spectrum emanating from the surface of the earth (the crystalline solid band gaps, conduction and valence bands, etc doesn’t cover it) for that matter.
“Ray, my question to CFU was asking about the effect of the preponderance of H in the sun being ionized — a minority of electrons bound to the H nucleus (proton).”
What is the temperature of the plasma phase of the sun (where the proton is naked)?
“You still have not credibly explained the continuous spectrum emanating from the sun (other than one inference — how it sounded to me — that it doesn’t exist…??)”
The sun deep down is a plasma.
The sun has a high pressure high temperature atmosphere.
The sun is optically thick at ALL FREQUENCIES, even those where the cross section of interception with a photon is minisculem, therefore the photons are thermalised.
If you don’t think this is a solved issue in stellar physics, then please explain how astrophysics has managed to miss this for all these years.
“or the cosmic background radiation for that matter”
Do you even know what this is?
HINT: It’s not photons.
“or the continuous spectrum emanating from the surface of the earth (the crystalline solid band gaps, conduction and valence bands, etc doesn’t cover it)”
Yes it does. The proof of this is done in quantum mechanics by showing the H atom binding energies arrived at by solution of the Schroedinger Equation and then showing the introduction of a bandpass energy level when two H2 atoms are considered and then extending this to the conduction band of (semi-) conductors.
If semiconductors could only absorb quantised energies, we would not have semiconductors that produce, for example, CPUs.
This is not basic physics for schoolkids, but they’re old hat for the speci alisms that rely on these phenomena.
What makes you think you’ve discovered how they all got it wrong?
Rod, The Sun is a very complicated magnetohydrodynamic structure where light takes YEARS to travel from the center to the photosphere. During this time, and because there are plenty of magnetohydrodynamic waves of various frequencies, it has plenty of time to come to equilibrium. Even so, the solar spectrum is not a perfect black body:
http://en.wikipedia.org/wiki/File:Solar_irradiance_spectrum_1992.gif
http://en.wikipedia.org/wiki/File:Solar_Spectrum.png
Likewise, the cosmic microwave background had hundreds of thousands of years to come to equilibrium before the Universe stopped being opaque to photons. Given long enough time, even very improbable events happen. However, there are inhomogeneities even in the CMB.
Rod, not being something of read about, but I’d guess a dense plasma is a bit like a liquid metal or amorphous solid in it’s electronic structure. The energy bands have fuzzy edges (think of amorphous Si) because of the deviations from regular repeating lattice cells. But they still exist.
For some electrons in some materials, the electrons are not bound to individual atoms, but they are bound to the material as a whole (if they left, they’d leave a net positive charge that would tend to pull them back). This can be true to varying degrees, as I understand it. In the limit of ‘free electrons’ within the material, in at least some cases if not more, an approximation can be made to treat the electron wave function as a plane wave (in an atom the wave functions for stable states (not in the process of emitting or absorbing energy) have the shapes of spherical harmonics – the s, p, d, f, … orbitals (because the energy well has spherical symmetry – except for perturbations from other electrons, which shift the energies from what they would otherwise be). Neighboring atoms can distort those states, shifting energy and shape, as occurs for the orbitals that form chemical bonds in particular (various sp hybrid orbitals occur, for example; p orbitals can form pi bonds or sigma bonds, if I remember correctly; the difference is orientation; when two such states combine, they form two new states).
[PS in the following, I used ‘wave number’ when ‘wave vector’ would have been more accurate.]
In the case of an electron approximated by a plane wave, the wave is still affected by energy wells (positively charged ions) (** PS one point I haven’t completely understood is that the plane wave’s wavelength should tend to shorten and lengthen going through a succession of energy wells; perhaps the plane wave approximation applies when those variations are small) and various redirections of the plane wave will produce an interference pattern; if the pattern is a standing wave then that is a state that can be occupied for a nonzero time (when the energy wells are not shifting). I’d imagine the electrons in dense plasma would occupy states within bands as in a liquid metal.
(PS in a crystalline material, an energy band is a series of individual states that vary with energy over electron wavenumber (proportional to momentum); by reasons I don’t fully understand, all wavenumbers are equivalent to wavenumbers within the first Brillioun zone, so that multiple energy levels exist at a given wavenumber, each being part of a different energy band; this is important for electrons to be able to aborb or emit a photon in full (as opposed to scattering or electron-phonon-photon (?) interactions and other stuff I probably don’t know)) because a photon carries very little momentum relative to it’s energy, and so to a first approximation, can change an electron’s energy without changing it’s wavenumber; thus photons are emitted or absorbed when electrons move from one energy band to another, not (at least not in a simple way) when they move among the states within a band. (Direct-band gap semiconductors have the bottom of the conduction band and the top of the valence band (or at least one bottom and one top if multiple minima and maxima have the same value) at the same wavenumber. Indirect-band gap semiconductors can’t absorb photons with band-gap energy in the most simple and likely way because the gap between states at any single wavenumber is larger; absorption of the lower energy photons occurs via a more complex mechanism that is less likely to occur, thus requiring a greater thickness of material to absorb the fraction of radiation (as is the case for c-Si.) The way the energy within a band varies as a function of wavenumber (proportional to momentum) determines the group velocity and also how electrons react to electric and magnetic fields (I think – right?); it is necessary for at least one band to be neither completely full nor completely empty in order for an electric current (via electrons) to exist, because the average motion of electrons for all states within a band is zero.
Something similar occurs for amorphous materials; there isn’t long-range order but there is a characteristic texture (typical atomic spacings and angles between bonds). The energy bands will be fuzzy.
Note that different energy bands can overlap in their ranges of energies; sometimes they converge to the same energy at certain wave vectors.
If my understanding is correct, electron-phonon interactions allow electrons to shift wavenumber and energy within a band. (phonons are quanta of crystal lattice vibrations (or the analogous phenomena within an amorphous material); they can be either thermal or acoustic; the thermal phonons are a manifestation of internal energy).
At a given temperature, their is an equilibrium distribution of electrons over energy levels, as their is for translational energy of molecules in a gas (and for any other form of energy), though the distributions are not the same; The density of occupied ‘states’ for molecular translational energy declines exponentially with increasing energy; the average energy is the energy at which the number of molecules per unit energy is 1/e times that at zero energy; this e-folding energy scale increases with increasing temperature. (the modal energy is zero; the modal speed is not because … math).
(Their is a tendency toward equipartitation (or otherwise, equilibrium distribution) of energy among states (because of a relationship between the probability of energy going from one form to another and the probability of energy going the reverse route, and the probability that energy can be found in one form obviously affects the probability that energy can be lost from or put into that form; thermodynamic equilibrium occurs when such transitions happen in forward and reverse at equal rates), but because of quantization of vibrational and rotational states, the energy may be more concentrated in translational form at lower temperatures (higher translational energies at higher temperatures can be transferred to rotational and vibrational states – PS one of those comes before the other; I think it’s rotational states that come into play first; vibrational states dont’ contribute much to heat capacity until even higher temperatures; but I could have that mixed up.) (hence smaller heat capacity at lower temperatures, at least for a gas; more complex stuff going on in condensed matter?).
(So, translational energy can vary continuously and can be zero; for gases, significant amounts of internal energy go into rotational and then vibrational states (or the other way around) as temperature increases; then comes electronic states, etc. In condensed matter, smaller electronic energy transitions become possible (in general; for metals in particular; based on band structure I’d guess not for insulators and large band-gap semiconductors (?? a band first has to have unoccupied states before electrons can make small transitions within a band – right??), so that significant amounts of internal energy can go into electronic states even at low temperatures.))
For electrons, the distribution over energy level is not a simple exponential because of the difference in the distribution of the allowable states. Instead, … I think the fraction of states that is filled varies from near 1 at lower energies to near 0 at higher energies, with the most rapid variation occuring where the fraction is 1/2 – that being the fermi level. At zero temperature, the electrons can settle down to a complete ground state, wherein the occupancy rate is 1 up to the fermi level and zero everwhere above. Thermal energy can perturb the electrons so that some are above the fermi level, leaving some states below the fermi level unoccupied. If an occupied state at higher energy and an unoccupied state at lower energy occur, then there might be an allowable energy transition wherein the electron loses energy going from one state to another; such a transition can occur in reverse if energy is around of a form that can be transfered to the electron for that transition. The probability of the upward shift for the electron depends on the abundance of energy; the probability of the reverse depends on state occupancy and thus the energy distribution among electrons. For a given temperature, there is an equilibrium distribution wherein energy transitions are occuring at the forward and reverse directions at the same rate; if there is more energy that electrons could gain, then at equilibrium, there must be more exited electrons and unoccupied lower-energy states so that electrons could lose energy at the same rate that electrons gain energy. The energy scale over which occupancy decreases from 1 to 0 across the fermi level increases with increasing temperature (I infer this is because it is more likely to find higher-energy phonons and photons when internal energy is more abundant, so transitions involving larger amounts of energy are more likely, AND ALSO, a greater abundance of energy available to electrons may increase the probability that an exited electron is further exited before it loses energy, and increase the probability that a lower energy hole (unoccupied state) can be filled by an electron exited from even lower energy before an exited electron falls back to recombine with the hole (fill a vacancy) – and of course, this increases the rate at which electrons can make bigger transitions from higher to lower energy states, supplying more higher energy photons (and phonons, I’d guess, with the scale of momentum transitions behaving similarly)).
Note what this implies for semiconductors and insulators. A large band gap centered at the fermi level reduces the probability of finding electrons in a conduction band or holes (unoccupied states) in a valence band, when at LTE. This tends to reduce electrical conductivity for the material. Also, exited electrons above the fermi level, even if they don’t recombine with holes, will tend to lose energy until they are near a minimum energy within the conduction band, while holes likewise tend to migrate toward maxima in the valence band.
The probability of any energy transition also depends on other things, so that, for example, photons may take longer to be aborbed and exited electrons may take longer to emit a photon, but the effect for both is such that it doesn’t alter the equilibrium distribution. The reason for that (which applies in general, not just to electronic states) goes to a level of physics that is beyond me at this point (or maybe it’s not; let’s see … well I’ll let somebody else do that part).
Anyway, even in a plasma, there are possible energy transitions for electrons, as one could expect by way of liquid metals (If molten Fe glows incandescently, why shouldn’t a dense plasma).
I suggest it’s time to make an apology to Ben Santer and return to the subject.
Rod does this same routine every year, hooking a new bunch of people.
He always grinds in the same direction.
Can we return to the subject of the thread?
“(in an atom the wave functions for stable states (not in the process of emitting or absorbing energy)”
Bad use of ‘stable’ there, since it has another meaning in this context, but what I meant was that these are the states that can be occupied for nonzero times (solutions to the time-independent Schrodinger equation, which is derived from a more general time-dependent Schrodinger equation. It is actually a misunderstanding that electrons cannot have intermediate energies between those discrete quantized states; they can, but they have to be in the process of gaining or losing energy, which actually takes a small but nonzero time).
(PS because of the variation of density of states over energy, variations in electron energy distribution can actually require a shift in the fermi level itself.)
—
PS as pointed out by others, even if the photospheric plasma doesn’t have the density of quantized states sufficient for a thin layer to emit and absorb photons with complete absorption of each photon, having a thick enough layer could make up for that; of course, if the photons can traverse distances over which temperature varies significantly, then the spectrum will be a bit distorted from a perfect blackbody.
However, … and I don’t know how important this is or how unimportant the alternative explanation is, but there is also inelastic photon scattering.
In the Earth’s atmosphere and at the surface, most scattering is elastic; it doesn’t involve net energy entering or leaving non-photons (but it tends to increase the entropy of the photons, effectively lowering the brightness temperature in some directions and raising it in others until the radiation is (if the index of refraction is isotropic, or otherwise if properly scaled by the index of refraction) isotropic (this being an aspect of LTE for the photons)), if the process occurs in isolation of other processes over a large enough expanse for the given opacity). However, there is something called Raman scattering. There is also Compton scattering of photons off electrons, which occurs in the Corona of the sun. These are forms of inelastic scattering. If inelastic scattering were modelled as absorption coupled to emission, it would not be a process at LTE – the emission and absorption cross sections couldn’t be assumed to be equal. However, as with all other energy exchanges, if this process occured in an isothermal expanse, presumably an isolated system would tend toward thermodynamic equilibrium whereby photons in inelastic collisions would gain energy at the same rate as they lose energy in inelastic collisions, and the radiation intensity would fit the blackbody value for the temperature. So, not knowing for myself what makes the bigger contribution – emission or inelastic scattering – those are two possible contributions to (approximately, maybe not at extremes of the spectrum, etc.) blackbody radiation from the photosphere.
(Sorry Hank, just wanted to finish that.)
Clarification: An electron with energy greater than vacuum zero could escape material (there may be some complexities as to how that actually happens; surface states are different because the crystal lattice ceases to repeat on one side, etc.). However, while within a solid or liquid, or, I would presume, a dense plasma, the electron wave propagation is still affected by the energy wells and there are still energy bands of electronic states.
Correction: Compton scattering is elastic.
Compton scattering:
http://hyperphysics.phy-astr.gsu.edu/hbase/quantum/comptint.html#c1
Raman Scattering involves photon scattering with a gain or loss of energy from/to vibration or rotation; it depends on polarizability of molecules:
http://hyperphysics.phy-astr.gsu.edu/hbase/atmos/raman.html
Raman scattering is inelastic in the sense that energy is lost to or gained from forms within a single particle involved; however, that energy can be transfered back in other collisions, unlike the inelastic collisions of macroscopic objects.
More on energy distribution:
Equipartition of energy:
http://hyperphysics.phy-astr.gsu.edu/hbase/kinetic/eqpar.html#c1
Maxwell-Boltzmann Distribution:
http://hyperphysics.phy-astr.gsu.edu/hbase/quantum/disfcn.html#c2
Other Distributions:
http://hyperphysics.phy-astr.gsu.edu/hbase/quantum/disfcn.html#c1
http://hyperphysics.phy-astr.gsu.edu/hbase/quantum/statcn.html#c1
Fermi-Dirac Distribution:
http://hyperphysics.phy-astr.gsu.edu/hbase/quantum/disfd.html#c1
Bose-Einstein Distribution:
http://hyperphysics.phy-astr.gsu.edu/hbase/quantum/disbe.html#c1
Ray (1026), I just don’t see how “coming to thermal equilibrium” explains the genesis of blackbody radiation. I know it takes many years for a photon (actually a much regenerated photon) to traverse from the sun’s core to the photosphere. I also know the sun is not a perfect blackbody. That’s an irrelevant non sequitur. It’s damn close to a blackbody spectrum, enough to clearly be generated according to Planck function. It’s almost infinitely closer to blackbody that the very few narrow He and H spectra lines.
Nor am I sure I get the relevance to the cosmic background radiation having hundreds of thousands of post Bang years to come to equilibrium. (I hope it has nothing with CFU’s claim that the cosmic background radiation is not photonic.)
Just a teeny probably insignificant clarification to Patrick 027: the plasma of the sun is predominately ionized with free, not bound electrons. Though the relatively few bound electrons have an effect, and I don’t think this changes the points in your treatise.
Hank might be right that we have milked this topic in this thread enough. Though I think discussion of this core science is terribly important — at least to me.
Can anyone explain how CO2 makes it into the upper atmosphere? Is this measured directly? At what level does energy finally escape back into space and what are the CO2 levels in that area?
What papers are out there for reference?
thanks!
“Nor am I sure I get the relevance to the cosmic background radiation having hundreds of thousands of post Bang years to come to equilibrium.”
The microwave background radiation has not been made microwave by thousands of years of post big bang years. It had an optically thick atmosphere before disassociation ended after the big bang when the universe was opaque. (just after the big bang, the universe was a plasma with dissasociated electrons).
That was at a temperature of tens of thousands of degrees and the expansion of the universe since then has stretched the wavelengths to microwave wavelengths such that it now resembles a 3 kelvin black body radiation.
This wavelength shift is known as “Red Shift”.
When countless numbers of journalists are going to meet on this conference:
http://www.gijc2010.ch/en/program
One of the scheduled items are:
The lobby that tried to kill Copenhagen: Investigating climate change
Thursday 22.04 – 10:45 Plenary Conference Room 2
Mark Schapiro (Center for Investigative Reporting, USA), Kate Willson (International Consortium of Investigative Journalists, USA), Brigitte Alfter (Journalism Fund, Belgium), Murali Krishnan (National Affairs Editor, India)
How certain groups opposing mesures against climate change attempt to influence discussions and disrupt the negotiation process. A first-class transnational investigation.
What this symphosium should had on their agenda, is how to investigate allegations opposed to scientific papers. Are the stories catch out of thin air? Yes, mostly.
Why can’t journalist do what journalist use to do – back in the good old watergate times – doing some real investigation. It is sooo easy nowadays. You simply have to use http://www.skepticalscience.com/argument.php as your starting point, and all the gory links to real research papers are available to pin-point these AGW deniers as smoking-industry look-alike lobbyist.
Twitter: kjell_arne
> Adam
> how CO2 makes it into the upper atmosphere
Diffusion. It’s a well mixed gas in the atmosphere.
> is this measured directly
Not by taste or touch or smell, but by various instruments, calibrated.
Try the video at this link for a demonstration of how CO2 blocks infrared slowly as the amount of it is increased between a warm face and a camera.
http://www.teachersdomain.org/resource/phy03.sci.phys.matter.co2/
> at what level … what papers
Try the “Start Here” link at the top left corner of the page, and the first link under “Science” in the right sidebar. That will take care of many of the basic questions.
Adam (1034) — CO2 is a well mixed gas in the atmosphere, with a mixing time of about 2 years, primarily due to the slow mixing of all gases across the equator (really, ITCZ). Radiant energy could, in principle, escape from any level, but mostly from high up; I can’t do better than to point you to
http://rabett.blogspot.com/2010/03/simplest-explanation.html
and then suggest you start with some textbooks if more thorough treatments are required. Unfortunately, atmospheric physics appears to be rather difficult (I find it so). I suggest some patience, if you can, and wait until Professor Ray Pierrehumbert’s book is published by Cambridge University Press (in October?).
Re Rod B –
“Just a teeny probably insignificant clarification to Patrick 027:”
Yes, noted; see my 1031 comment.
Rod B. says, “I just don’t see how “coming to thermal equilibrium” explains the genesis of blackbody radiation.”
Yes, that is EXACTLY the problem! Because that, ultimately, is all there is to a blackbody radiation distribution–the fact that it is THE equilibrium distribution for a photon gas at a temperature T. It does not matter HOW it gets there. What matters is that once it reaches equilibrium, it will only fluctuate slightly about that equilibrium.
As I said, though, photons do not interact with other photons (to first order), so they have to APPROACH equilibrium by interacting with matter in their vicinity. But they can only interact with matter if their energy corresponds to an allowed transition of the matter, right? So, real matter can only put the photon gas into a distribution that approximates a blackbody. If the photons have many years to come to equilibrium in a complicated plasma like the Sun, they will be very close to a blackbody distribution. They will even fill in some of the gaps with very improbable interactions. On timescales of days, weeks or even months, the approach to blackbody is not nearly so close.
Remember, to get a blackbody distribution, you have to have matter–something has to be emitting and absorbing the radiation. And that something can only emit and absorb where it has allowed transitions.
Ray – when you say Landau and Lifshitz, is this the one you mean?
I read the section in which they derive the characteristics of blackbody radiation. The approach is clever – dealing with the actual energy states and transitions of the matter directly would be impossibly complex, but if we can say that the matter is in equilibrium with a photon gas in a sufficiently large closed volume, we can take advantage of the simple properties of a “gas” to get what we want easily by just combining the distribution of available electromagnetic states with the distribution of photons among those states – and we get to the blackbody spectrum in just 3 or 4 equations.
Clever, but in a way unsatisfying, since it doesn’t provide any insightful ways of thinking about the matter that’s doing the radiating. Most of the time when we talk about blackbody radiation, we’re not talking about the conditions inside a cavity, but about some body which is radiating outward into space – and presumably not in equilibrium with a photon gas. Yet the radiation has the same distribution as in the L&L derivation – or at least is pretty close. It seem like we should be able to get to the same distribution by saying something about the number and spacing of states in the matter itself, and about the quantum distribution according to which they’re populated. So far the most concrete description I’ve found is
but I’m not quite sure whether my mental model is consistent with it or not.
“Clever, but in a way unsatisfying, since it doesn’t provide any insightful ways of thinking about the matter that’s doing the radiating.”
Then do the version of the calculations that have this problem instead:
“dealing with the actual energy states and transitions of the matter directly would be impossibly complex”
David Warkentin,
Yep! That’s the one. Feynman’s derivation is good as well. My preference for the L&L is mainly because of their emphasis on the properties of the photon gas (also present, but less explicit in Feynman). The assumption of equilibrium is actually pretty good as long as the system is strongly interacting with the radiation (closer to a true blackbody). Thus, Earth’s outgoing IR is seen to approximate radiation from a series of layers of different temperatures.
And, yes, I realize this is tremendously off topic. However, it is crucial to understanding the greenhouse effect and blackbody radiation does seem to be an area where many people have conceptual difficulties.
Hey! CFU, I mostly agree with your #1035. One clarification/question though (and possibly this is just a pristine distinction — or maybe not): My understanding is that it is the decreasing temperature resulting from the expansion of the universe soup which lengthens the peak wavelength of the Planck radiation. It is not exactly because the “edge” of the universe is sailing away and creating a red shift per se.
Rod B. says: “My understanding is that it is the decreasing temperature resulting from the expansion of the universe soup which lengthens the peak wavelength of the Planck radiation. It is not exactly because the “edge” of the universe is sailing away and creating a red shift per se.”
Actually, the redshift is because every point in the Universe is moving away from every other point in the Universe–you are creating space between each pair of crests in the electromagnetic wave corresponding to a CMB photon, and so increasing the wavelength. Voila. Red shift.
Remember, at the Big Bang, space was a geometric point, so space is being created continually.
Ray L, but “how it got there” is closer to the question at hand, so it does matter. And, I’m still not sure of the relevance to equilibrium, although it is related, I’m still working on it, and don’t materially disagree with your description per se. (David Warkentin might describe my thinking here better than I do.)
However, I think part of your post 1040 sheds light on our differences. You say, “…But they [photons] can only interact with matter if their energy corresponds to an allowed transition of the matter…” I agree with the precise wording of that, but the generation (or absorption) of blackbody-type radiation in matter has near infinite number of allowed energy transition levels. It is not limited to the vibration, rotation, and electron orbital transitions/levels. That was my contention and the crux of the disagreement. (Though it has morphed around a bit, I admit.) To wit: The photons absorbed or emitted by virtue of the internal energy levels of molecular vibration, rotation, and electron orbits ARE NOT the genesis of blackbody-type radiation — is not ‘how it got there.’ The (nearly) continuous blackbody spectrum is emitted instantaneously with the heating of the matter.
This doesn’t mean that Planck and related functions/equations can not be used as constructs to analyze the internal energy levels and their attendant emission and radiation. They can be and beneficially are. It is hard to model the few discrete (at least very very narrow) lines from vibration and/or rotation for example. But if you take the few lines over a slightly wider bandwidth (but still very narrow compared to a continuous spectrum) and average their intensities (as best one can) as if spread continuously over that narrow bandwidth, then the relatively simple Planck functions, which can have different emissivities and be constrained to a very narrow bandwidth by virtue of the limits of integration, can be used in the model more readily and yield pretty accurate results. But that doesn’t make the radiation blackbody type.
Ray (and CFU), actually the red shift explanation of CBR makes sense. But so does the lower temperature because of gaseous expansion. Maybe a little of both? Though it all raises many other fascinating questions. It’s all mind boggling. […and Hank is probably tearing his hair out…]