Thursday, April 19, 2012

Doctrine or Consensus?

Recently a reader of the blog asked some specific questions about the results in the survey of climate scientists. The questions are:


Is there one group of scientists who are unconvinced on many of the questions you ask, or are different scientists more unconvinced on different points?

If the issue is a very expensive and extensive mitigation strategy, what fraction of respondents actually feel forced to answer Yes because they are very convinced by all the sub issues that would make that strategy necessary? And what fraction might believe in most of them, but disagree on enough points that they would not see an extensive mitigation strategy as a good idea.

Are the people who disagree with details AGW consensus the same group throughout, or are there some people who disagree with some aspects but agree with others?

 
A comparison based on expertise (questions 4 – 6) of:

 
  • change mostly due to CO2
  • seriousness of issue
  • potential for catastrophe
  • mitigation vs adaptation
  • IPCC over or under estimates sensitivity
  • need for immediate action for mitigation
 

 In the following I attempt to answer some of these questions.
 

 
 
Is there one group of scientists who are unconvinced on many of the questions you ask, or are different scientists more unconvinced on different points?

The ‘groups’ of scientists are based on a rudimentary measure of ‘expertise’, namely the number of peer reviewed publications and participation as an IPCC lead author.

Number of peer reviewed publications

Do the IPCC reports tend to underestimate, accurately reflect (a value of 4) or overestimate the magnitude of the impacts resulting from change in temperature?

under estimate 1 2 3 4 5 6 7 over estimate


Table 1. Temperature x Number of peer reviewed papers 
 



Do the IPCC reports tend to underestimate, accurately reflect (a value of 4) or overestimate the magnitude of the impacts resulting from change in precipitation?


under estimate 1 2 3 4 5 6 7 over estimate


Table 2. Precipitation x Number of peer reviewed papers



Do the IPCC reports tend to underestimate, accurately reflect (a value of 4) or overestimate the magnitude of the impacts resulting from change in sea level rise?


under estimate 1 2 3 4 5 6 7 over estimate


Table 3. Sea Level Rise x Number of peer reviewed papers


Looking at the means it is probably safe to say that the number of peer reviewed publications makes no real difference in the assessment of IPCC tendency to over or under estimate aspects of climate change.


 
IPCC Lead Author

Do the IPCC reports tend to underestimate, accurately reflect (a value of 4) or overestimate the magnitude of the impacts resulting from change in temperature?

under estimate 1 2 3 4 5 6 7 over estimate

Table.4. Temperature x IPCC lead author



Do the IPCC reports tend to underestimate, accurately reflect (a value of 4) or overestimate the magnitude of the impacts resulting from change in precipitation?


under estimate 1 2 3 4 5 6 7 over estimate


Table 5. Precipitation x IPCC lead author




Do the IPCC reports tend to underestimate, accurately reflect (a value of 4) or overestimate the magnitude of the impacts resulting from change in sea level rise?


under estimate 1 2 3 4 5 6 7 over estimate


 
Table 6. Sea level rise x IPCC lead author
 
 
 

 
 
If the issue is a very expensive and extensive mitigation strategy, what fraction of respondents actually feel forced to answer Yes because they are very convinced by all the sub issues that would make that strategy necessary? And what fraction might believe in most of them, but disagree on enough points that they would not see an extensive mitigation strategy as a good idea.

The survey asked the respondent to agree or disagree that there is a great need for immediate policy decisions for immediate action to mitigate climate change.

Cross tabs of temperature estimates, precipitation estimates and sea level rise with the immediate need for mitigation decisions produces the expected. Those scientists who perceived the IPCC to underestimate the phenomena were more inclined to emphasize the urgency and immediacy of mitigation policies and actions. The frequencies can be seen in the following tables.

There is a great need for immediate policy decisions for immediate actions to mitigate climate change:

strongly disagree 1 2 3 4 5 6 7 strongly agree


Table 7. IPCC temperature estimate x Need for mitigation decisions

 



 

Table 8. IPCC precipitation estimates x Need for mitigation decisions




Table 9. IPCC sea level rise estimates x Need for mitigation decisions


 
The perceived need for mitigation measures increases with the perception that the IPCC under estimates climate change.


 

 
Are the people who disagree with details AGW consensus the same group throughout, or are there some people who disagree with some aspects but agree with others?

 
A comparison based on expertise (questions 4 – 6) of: Some of the questions are answered above so this section only presents what is new.

 
  • change mostly due to CO2
  • seriousness of issue
  • potential for catastrophe
  • mitigation vs adaptation

 Expertise is rudimentary measured as number of peer reviewed publications and being a lead author for IPCC

 
 Anthro attribution: : not at all 1 2 3 4 5 6 7 very much




 Tabkle 10. Change mostly due to CO2 x Number of peer reviewed publications





 
 Looking at the means, those with the higher number of publications are slightly more convinced that most of the recent or near future change will be a result of anthropogenic causes. The difference is minimal that a test of statistical significance is not really necessary.


How convinced are you that climate change poses a very serious and dangerous threat to humanity?

not at all 1 2 3 4 5 6 7 very much


Table 11. Seriousness of issue x Number of peer reviewed publications




Again, number of publications does not seem to influence opinion



If we do not do anything towards adaptation or mitigation, the potential for catastrophe resulting from climate change for other parts of the world are:  very low 1 2 3 4 5 6 7 very high


Table 12. Catastrophe in next 10 years x Number of peer reviewed publications

 
 
A slight difference is noticeable where as those with more publications are very slightly less convinced of impending catastrophe in the next 10 years.



Table 13. Catastrophe in next 50 years x Number of peer reviewed publications


Here a similar tendency as the 10 year estimate is evident.

Anthro attribution: not at all 1 2 3 4 5 6 7 very much








Table 14. Change mostly due to CO2 x IPCC Lead Author


 
 
Those scientists participating as a lead author for the IPCC seem to be slightly more convinced when it comes to attribution of CO2.



How convinced are you that climate change poses a very serious and dangerous threat to humanity?

not at all 1 2 3 4 5 6 7 very much








Table 15. Seriousness of issue x IPCC Lead Author






Here the is no difference between groups



Table 16. Catastrophe in next 10 years x IPCC lead author

 
Again, no real difference



Table 17. Catastrophe in next 50 years x IPCC lead author

 
And again, no real difference.


As for the conclusions drawn, I will leave that for your comments.

7 comments:

Hector M. said...

Very interesting.
I think a similar or even stronger consensus would have come from a survey of physicists in 1904 about the possibility that Newtonian theories about time, space and gravitation could be refuted within 10 years, that time and space are not absolute, that gravity is a consequence of the curvilinear nature of space-time, that gravity can bend light. On a 1-7 point agree-disagree scale, the average against such outrageous ideas would be nearly 7, with almost all scientists (especially those with more peer reviewed papers) indignantly rejecting such possibilities as absurd and baseless. Would such consensus in any way affect or enhance the validity of any one thesis or the opposite? Would it tell what to expect about scientific developments in the following 10 years?
Consensus statistics do not tell anything about physical reality, not even about future scientific knowledge, but only about the state of knowledge and belief among scientists at a given time.

Harry Dale Huffman said...

No, this is not interesting, because it is just an opinion poll, not a test of their actual knowledge. As an independent physical scientist, I found it all too easy to demonstrate the incompetence of the climate "consensus" on points of basic physics; since they are incompetent, why should I waste my time studying their, to me, worthless opinions--that is a job for psychologists and future historians of this period, who I guarantee will be asking, for a long time, "why didn't they see that...?".

Dennis Bray said...

@ Hector M

I think it would be a nice project to look at various consensuses throughout the history of science. “Would consensus in any way affect or enhance the validity of any one thesis or the opposite? Would it tell us what to expect about scientific developments in the following 10 years?” Well, I don’t know if there is a strong relationship between validity and consensus, I think it depends on the case. But I think there is an obvious relationship between consensus and power, both within and external to science. With the case of climate change, this also includes political power. The ‘consensus’ has also assisted in transforming a scientific issue into a major growth industry, where all and sundry can participate. Viewing science as simply another social entity (with a very strange market model) helps to put this into perspective, especially in times when technology seems to supersede science in terms of reward. What would it tell us about what to expect in the next 10 years – maybe the direction of funding, the requirements for prestige. The intention of the surveys is not to tell us anything about the physical reality and the future of scientific knowledge. It is as you have stated, to tell us ‘about the state of knowledge and belief among scientists at a given time.’ To this end, the survey has been repeated and will be conducted again, so as to provide a picture of how knowledge and beliefs among scientists might change over time, and hint at the influences that lead to consensus and how these might change over time as well.

@Harry Dale Huffmann

Dear Harry, is not a person’s opinion of what is correct knowledge, simply that, an opinion. In my opinion, 2+2 = 4. I think there is also a pretty good consensus on that account. But I would be hard pressed to explain it theoretically, other than it is an additive form of measurement.

I am not sure what an ‘independent physical scientist’ does, but if you can demonstrate all of this incompetence in climate science, why not do so. I would say I hold reserved opinions – I am not a physicist – I simply observe what is occurring among scientists within climate science. If you, however, hold knowledge that could provide definite and legitimate arguments that something is wrong with the science, then why not publish them? Is that not a moral duty if you are a scientist? After looking at your web site I get the impression that you are a supporter of intelligent design. If that is correct, perhaps you could answer a few questions:

1. is it the possibility that climate is changing that you do not accept? or
2. is it the possibility that climate change is an anthropogenic influence?
3. if your position is the former, do you deny climate change in the past?
4. if you accept the possibility of natural climate change, how does that fit into intelligent design?

It would be nice to know a bit about consensus of people with alternative world views. Of course, I am not in a position to question your ‘knowledge’, I simply want your opinion, which, after all, you are entitled to, valid or not.

Alex Harvey said...

Dear Dennis,

Why do you think these results are so different from the recent AMS survey?

In the AMS survey it was reported, for instance,

3. Do you think that the global warming that has occurred over the past 150 years has been caused...

Mostly by human activity - 59%

More-­‐or-­‐less equally by human activity and natural events - 11%

Mostly by natural events - 6%

I do not believe we (scientists) know enough yet to determine the degree of human or natural causation, even in the general terms stated in the categories above - 23%


This seems quite a different response than suggested by table 10 in your survey, for instance.

ghost said...

good survey, I think. I do not see a lot of bias in any group. Furthermore, I think, most are quite satisfied with the IPCC report.

I think, the questions refer mostly to the WGI report or physical questions (temperature, precipitation, sea level).

But, the biggest controversies in the IPCC report did not occur in the WGI report, but in the other parts. I think the other parts are more complicated because they are about the (regional) effects and the society. Physics is simple compared to humans and societies, IMHO. You can measure data, you can apply physical laws and mathematics to analyze data and to model the world. That is hard, but somehow you have a lot of objective constraints.

Dennis Bray said...

@ Alex

I am not too familiar with the AMS survey. If I am right, the survey was only of members of the AMS, and the membership it not entirely climate scientists. My survey was of an international sample of climate scientists. This might explain some of the difference. Also the time of the surveys (mine pre climategate, AMS post climategate) might act upon the results. That aside, results from the AMS results show that 59% of the respondents thought that global warming in the last 150 years has been caused mostly by human activity. First of all, the question does not give the respondent the opportunity to say that NO global warming has occurred.

Concerning table 10 in this posting, the full question is as follows:

How convinced are you that most of recent or near future climate change is, or will be, a result of anthropogenic causes? The possible response categories were: not at all 1 2 3 4 5 6 7 very much.

AMS asked Do you think that the global warming that has occurred over the past 150 years has been caused by

1. mostly human activity (69%)
2. more or less equally by human and natural events (11%)
3. mostly natural causes (6%)
4. I do not believe we know enough to determine causation (23%)

While the two questions both address attribution, they do so in a different manner.

If we allow some leeway in interpretation, we could suggest that S response category 1 (mostly human activity) corresponds with my survey categories 6, 7, (very much convinced of anthro causes) making the results AMS 69%, my survey appx 67%; not so different. But I am not sure if it would be worthwhile to do a similar comparison with other response categories.

Sorry, that is the best I can offer considering 2 the different questions, the different sampling techniques and the time and events between the two surveys.

Dennis Bray said...

@ ghost

The IPCC: The survey contained a set of questions specifically referring to satisfaction with the IPCC. The question was posed as “The IPCC reports tend to under estimate (1) accurately reflect (value of 4) or over estimate (7) the magnitude of the impacts resulting from changes in temperature, precipitation, sea level rise and extreme events. A second question was posed in the same way concerning projections made in the IPCC report. Combining response categories 3,4 and 5 as a representation of being satisfied with the IPCC assessments produces the following results:
Impacts: temperature – 91%; precipitation – 85%; sea level rise – 81%, extreme events – 76%. Future projections: temperature – 90%; precipitation – 63%; sea level rise – 81%; extreme events – 78%. From that I would conclude that in 2008 most of the sample was quite satisfied with the IPCC report, but less so, for example, concerning precipitation.

You are correct in assuming the questions were mostly in reference to the WG1 report, and the sample for the survey – climate change scientists from the natural scientists – was selected accordingly.

While I am not so sure that most of the controversy does not concern WG1, I do agree 100% that the socio-economic impacts and regional projections are extremely complicated to assess. My concern is that often those doing impact analysis and regional projections do not question what is presented in WG1 just as WG1 does not seem – although I am guessing here – to be overly concerned about the ‘scenarios’ driving the models. Both seem to accept the complimentary disciplines’ input without too much question. In the surveys of climate scientists I have tried to shed some light on the collective opinion of climate scientists concerning their assessment of their own science not only for the benefit of the physical component of the climate issue but also to demonstrate to the social component that there is no such thing as certainty and that the projections should not be taken as gospel. If the climate physical-science community is only 63% happy with the IPCC’s assessment of precipitation, for example, then the impact and socio-regional assessment community should be aware of this short coming and make it known in their results.