Climate Science Glossary

Term Lookup

Enter a term in the search box to find its definition.


Use the controls in the far right panel to increase or decrease the number of terms automatically displayed (or to completely turn that feature off).

Term Lookup


All IPCC definitions taken from Climate Change 2007: The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Annex I, Glossary, pp. 941-954. Cambridge University Press.

Home Arguments Software Resources Comments The Consensus Project Translations About Donate

Twitter Facebook YouTube Pinterest

RSS Posts RSS Comments Email Subscribe

Climate's changed before
It's the sun
It's not bad
There is no consensus
It's cooling
Models are unreliable
Temp record is unreliable
Animals and plants can adapt
It hasn't warmed since 1998
Antarctica is gaining ice
View All Arguments...

New? Register here
Forgot your password?

Latest Posts


Global warming theory isn't falsifiable

What the science says...

Climate Myth...

Global warming theory isn't falsifiable


1  2  Next

Comments 1 to 50 out of 87:

  1. As a skeptical person in general, I haven't really made my mind up either way.  But this idea of falsifiability seems particularly elusive.  I've been searching for the current definitions of falsfiability for global climate change theories with little luck.  Is there any resources that summarizes that aspect of the debate?  Google has been no help at all, in that regard.

  2. Falsification just means that it is possible to falsify a theory if it is false.  This means the theory must exclude the possibility of some event.  There is little more to it than that.  In the case of the basic element of AGW is the idea that CO2 is a greenhouse gas, the theory suggest that in the absence of any other change in the forcings that could explain it, a cooling trend on a centennial scale would be impossible while CO2 concontrations continue to rise at a non-negligible rate.  This means that the thoery is falsifiable because there is an observaton (centennial cooling trend) that is excluded by the theory.  Of course there could be quibbling about the length of the trend required and the rate of increase of CO2 etc., but the idea of falsifiability assumes that both parties are acting in a rational and fair manner to determine the exact details of the test.  Hope this helps.

  3. P.S. there is a good reason why Google doesn't show much of interest, which is that it anyone who understands falsifiability and the physics of the greenhouse effect can immediately see why it is directly falsifiable.  The question is largely used in the debate for rhetorical, rather than scientific reasons.

    I suspect Tom Curtis would be able to give a very satisfactory answer on this one.

  4. Dikran's list is sufficient, but there are several places where folks have discussed in more detail.  One is at Climate Sight.

  5. Jacklongley,

    In 1896 Arhennius predicted that the land would warm faster than the ocean, the Northern Hemisphere faster than the Southern, the night faster than the day, the Arctic faster than the rest of the Northern Hemisphere, and the winter faster than the summer.  If any of the predictions did not work out that might be considered a "falsification" by those who make this argument.  In the 120 years since his predictions, the measured data show he was correct on all of them.  Can you suggest an alternate theory that explains these observations?  Skeptics have provided no alternate thory.  There are myriad other more recent predictions of climate science that can be found, but Arhennius is so strong an argument that it seems not worth the trouble to look them up.  

  6. Much of this idea about "falsfiability" actually comes back to models. Those screaming about lack of proof conveniently ignore all the obvious direct tests made long ago because they dont falsify the theory as hoped. Instead this is an expectation that climate theory will predict tomorrow's weather and bitching because the failure of climate models to do this doesnt seem to count as falsification. It is important to realise that the theory could be right but models could be wrong, but more importantly, models are predicting long term trends (remarkably well) and dont pretend to have any skill at predicting decadal trends.

    Its worth having a look at the papers in "empirical evidence for global warming". The Evans 2006 paper (and similar by Harries, Chen etc) are pretty direct experiments that would have returned different results if AGW was wrong.

  7. jacklongley @1, falsifiability is a strange concept of limited use in science, despite its popularity.  The reason is that when we test any hypothesis, we must make background assumptions both about other conditions, and about how our instruments work.  These background assumptions then form auxilliary hypotheses which are tested alongside the hypothesis we actually desire to test.  As a consequence, if our test gives a negative result, we do not falsify any individual hypothesis (including the one we wanted to test).  Rather we falsify the conjunction of the hypotheses.  We show that not all of them can be true together.  This is known as the Duhem-Quine Thesis, after its two independent "discoverers".

    To illustrate this, consider Dikran Marsupial's test of "global climate change theories" from 2 above.  He claims that a centenial negative trend in temperatures would falsify the theory.  Of course, if that centenial trend coincided with a 50% reduction of solar physics, the theory would not be falsified.  Dikran is quite aware of this, and covered himself with the auxilliary clause that the trend occured "in the absence of any other change in the forcings that could explain it".

    In very simple theories, we can radically reduce the number of auxilliary hypotheses making the particular hypothesis of interest more amenable to falsification in a "crucial test".  We can also vary our experimental methods so that we are testing the theory with different auxilliary hypotheses.  Thus, for very simple hypotheses, we can reduce the impact of the Duhem-Quine Thesis, but we can never entirely avoid it.

    Because AGW is a complex theory with many auxilliary hypotheses, it is difficult to develop "crucial tests", ie, any individual test that will show it to be false.  In fact, in the very short term it is impossible.  What we can do is develop "crucial tests" for important elements of the theory, but not for the whole theory at once.  We can also measure relative likilihood with respect to competing theories.  Doing so, we can show that AGW easilly is a superior theory to its competitors.  But we cannot pick a single experiment to falsify the theory, so you will not find much discussion of falsification with respect to AGW.

    When you do, it is often for critics of AGW who take a farcically simplistic view of falsification to declare that "AGW is falsified".  Spencer and Christy played this game for a while, declaring the UAH satellite temperature index falsified AGW.  Then (on several occasions) they were embarrassed when it was shown that their auxilliary hypothesis that they had eliminated all significant errors from their temperature record was what was false, and that UAH tends to confirm rather than falsify AGW.

    Lucia Liljegren has played a similar game, several times declaring that the recent temperature record falsifies IPCC predictions.  She has neglected, however, the IPCC auxilliary hypothesis of neutral ENSO conditions*.  She has merely falsified the conjunction of hypotheses that (CO2 forcing is increasing & climate sensitivity is in the IPCC range & ENSO fluctuations do not effect global temperatures & ....).  As her third, tacit, auxilliary hypothesis is known to be false, her results are massively uninteresting.  (She also uses a very simplistic definition of falsification in which events with a 1in 20 probability of occuring in a theory are supposed to falify the theory; something Popper would have considered ludicrous.)

    Despite the limitations of the Duhem-Quine Thesis, it is possible to extend the emperical content of theories, and the climate science community is doing just that with the theory of AGW.  They are going to extraordinary lengths to do so.  Because of the complexity of climate science, in some areas that is difficult and slow going.  It is very noticable, however, that it is the people rejecting AGW, the Republicans in the US, the Harper government in Canada, who are defunding research.  They are so confident in their position, it appears, that they dare not put it to empirical test.

    There is a lot more that could be said on this very interesting subject, but I am out of time, and this post is long enough as is.  I'll respond later if you want to explore the subject further.


    (* The IPCC auxilliary hypothesis is not that ENSO does not effect temperatures, but that ENSO fluctuations cannot be predicted except in the very short term, and so cannot be included in IPCC projections.  That does mean that IPCC predictions in the short term cannot be properly tested unless the effect of ENSO is removed from the temperature record.  Lucia does not do that, thereby assuming as a tacit auxilliary hypothesis that ENSO does not effect global temperatures.)

  8. My theory that Tom could give a good answer was not falsified, thanks Tom!

    The thing that I would point out is that the attempts to falsify the models never seem to mention the decline in Arctic sea ice, which is much more obviously wrong in the models than the tropical trophospheric hotspot (both of which the climate modellers are well aware and quite happy to discuss).  I wonder why that might be ;o)

    Falsifiability is a useful concept in science in the same way that frequentist hypothesis tests are, both have severe limitations and both are widely misunderstood.  Used properly, however, they are both good methods of providing basic sanity checks that help us not to jump to conclusions (i.e. they help us to be self-skeptical - which is the most important form of skepticism, if you haven't got that, you haven't got any other sort!).

  9. Dear @Tom Curtis,

    Falsifiability may very well seem like a "strange concept of limited use" to an AGW believer, but nevertheless it happens to be the very definition of science. Be sure to note that you happen to be in an absolute minority in your way of looking for an alternative definition for science.

    Also I am pleased to learn that the AGW community is looking for ways to formulate the AGW theory in a (more) scientific manner, but I think it is very inappropriate for Mr Curtis to use ad hominem political classifications to those who happen to disagree with his beliefs. I read somewhere that these types of comments have no place on this site.


    [JH] You assert that falsibility "happens to be the very definition of science."  Please document the source of your statement.

  10. PanicBusiness:  You are incorrect that falsifiability is "the very definition of science."  That is something you would know if you had gotten past introductory science classes in college.  (It is sad that such fundamentals of science are inadequately taught at the introductory course level.)  It happens that there is a recent and excellent post on falsifiability of anthropocentric global warming, by Hans Custers.

  11. IMO going on about falsifiability without any other support is just another tired rhetorical argument against the available evidence.

    It's the philosophy-of-science version of "pounding the table" as per what I understand is a well-known joke among lawyers. (*)

    (*) To whit:

    If you have the facts on your side, pound the facts.
    If you have the law on your side, pound the law.
    If you don't have either the facts or the law on your side, pound the table.

  12. PanicBusiness, as I showed at 2, AGW is a theory that is every bit as falsifiable as pretty much any science where designed experiments are not possible.  The caveats that Tom gives are equally applicable to many other sciences.  The idea that AGW is not falsifiable is an obvious canard, and you would do better by learning something about the philosopgy of science from Tom.

  13. In answer to the moderator inline @9, Karl Popper did, or at least something very like it.  In fact, he wrote:

    "According to my proposal, what characterizes the empirical method is its manner of exposing to falsification, in every conceivable way, the system to be tested. Its aim is not to save the lives of untenable systems but, on the contrary, to select the one which is by comparison the fittest, by exposing them all to the fiercest struggle for survival."

    (Logic of Scientific Discovery, page 20)

    The odd thing is that a great deal of science was done before Popper published that "definition" of science in 1959.  Some of the very best of that science (the development of heliocentrism, of Newton's laws of motion and graviation, of the principle of conservation of energy) was done by methods which do not meet the imprimature of Popper's methods.  Either what the great scientists of the ages has been doing was not, after all, science - or Popper was wrong.

    However, even if Popper was right, that would not justify the naive methodological falsificationism to which PanicBusiness appeals.  As Popper writes:

    "Every test of a theory, whether resulting in its corroboration or falsification, must stop at some basic statement or other which we decide to accept. If we do not come to any decision, and do not accept some basic statement or other, then the test will have led nowhere. But considered from a logical point of view, the situation is never such that it compels us to stop at this particular basic statement rather than at that, or else give up the test altogether. For any basic statement can again in its turn be subjected to tests, using as a touchstone any of the basic statements which can be deduced from it with the help of some theory, either the one under test, or another. This procedure has no natural end.  Thus if the test is to lead us anywhere, nothing remains but to stop at some point or other and say that we are satisfied, for the time being."

    (Logic of Scientific Discovery, page 86, my emphasis)

    That may be a bit obscure to people unused to the philsophy of science, but what Popper is saying is that when we have a purported falsifying instance (ie, the "basic sentences" mentioned in this passage), the truth or otherwise of that sentence can also be called into question.  And if called into question, it can be tested and potentially falsified.  (This is just the Duhem-Quine thesis in a different guise.)

    The obvious question then becomes, what happens to your falsification if the basic sentence which falsified the theory is itself falsified?

    Regardless of how we answer that question, Popper is very clear that the acceptance that a basic sentence falsifies a theory is a matter of pragmatic convention.  Clearly that convention must be heavilly influenced by empirical facts, and Popper specifies certain methodological conventions to ensure that this is so.  Those conventions cannot be determinative, however.  They must always leave room for essentially subjective choice.  If they did not, then Popper would be claiming to have solved Hume's problem of induction, something he was certain could not be solved, and which his theory was intended to side step be developing a science without induction. 

    The consequence is that a truly Popperian science would proceed much as I described @7.  So, PanicBusiness has not only mistaken Popperian science for naive falsificationist science, and but he(?) has mistaken a statement of how Popperian science must, logically proceed as a rejection of Popper.

  14. Is there supposed to be an actual article associated with this?  All I can see is the Myth (Global warming theory isn't falsifiable) followed immediately by the comments.  I see the same thing on an iPhone, an iPad, and an iMac (using both Safari and Firefox).  I don't notice this on other posts.  Don't get me wrong: the comments have been very interesting, but I'm wondering if I'm missing something (and why that might be).


    [SkS] The author team is looking into this matter. 

  15. @Tom Dayton, @Composer99, @JH and others. It is a pleasure to introduce you to the scientific method.

    Scientific theories are testable and make falsifiable predictions.1

    Now, I am not claiming that the current widespread AGW belief system cannot be formulated in a scientific manner, quite on the contrary I urge the AGW/CAGW community to commit themselves to do so.

    In particular, as the AGW belief system happens to be strongly tied to devastating2 political regulations3, I further urge the community to propose experimets that could potentially exclude models and theories with high climate sensitivity i.e. higher than, say 1.5°C.

    In science, scientists are actively looking for experiments that would potentially falsify their theories and they proudly announce if they find such. Because this makes their theories scientific. Yet the vast majority of commenters here attempt to bend the definition of science to make "climate science" a better fit.

    Again, I urge the AGW community to stop this dishonest behavior and start discussing potential ways to falsify AGW/CAGW theories and announce it when they reach a "consensus". Until then you not only cannot assert that AGW/CAGW is settled science, you can not even assert that it is science.

  16. SteveS @14  no I don't see an article either !

    Here <sarc>is the proof that Einsteins general theory of relativity is unfalsifiable. Why don't "skeptics" complain about that too ? </sarc>

  17. PanicBusiness, Please read the comments policy; your previous post is in contravention of the comments policy in a number of respects: (1) sloganeering - you are just repeating a point that has already been addressed by Tom (2) accusation of dishonesty: (3) inflamatory tone.  I suspect if you continue in this way, your posts will soon attract the attention of the moderators (I would already have taken action if not for the fact I have already participated in the discussion).


  18. Dikran:

    Was there an OP posted when you commented on Jan 24?

  19. no, IIRC it was blank at that time as well.

  20. Panicbusiness,

    I noticed that you skipped the content of the first two paragraph of the link you have.  The theory of AGW meets all the descriptions in your link, including the falsifiability provision.

    I listed 5 falsifiable predictions of AGW in my post above.  Several other posters have either listed what you need to show or linked to other sites which list a variety of additional experiments you can do to show your point.  It is not really our responsibility to repost these links every time a new person appears who does not want to read the posts already on this thread.

    When scientists have been working on a theory for 180 years, they solve most of the problems with the theory.  This has happened with the theory of AGW.  All the major issues were addressed 50-150 years ago.  The falsibility tests have been done and the theory has correctly predicted what would be measured in the new experiments.  AGW has stood the test of time.  Deniers do not want to acknowledge these results so they claim that is impossible to falsify the theory.

    You are correct in one way though: it is impossible to falsify a theory that is correct.

  21. PanicBusiness:  Following up on Michael Sweet's reply, I suggest you research "theory evaluation."  Here is one of a great many excellent explanations.  I found this one at the top of my Google search results page just now:

  22. PanicBusiness:  A caveat about the theory evaluation explanation I just gave you.  Note that the term "disprove" in the first paragraph there is in quotes.  That's important, because science does not really disprove nor prove anything, per Tom Curtis's comment above.  It really all comes down to subjective probability, eventually.

    Often people say "proof is for math, not science."  But even math does not have absolute proof, because mathematicians can make mistakes.  Otherwise there would be no "proofs" disproving other "proofs."

  23. panicbusiness

    " I further urge the community to propose experimets that could potentially exclude models and theories with high climate sensitivity i.e. higher than, say 1.5°C."

    OK. We examine past climates to estimate the climate sensitivity actually was. If past climates commonly indicate a CS below 1.5 then the idea of a CS above that may well be falsified.

    Rohling et al 2012

    They look at several dozen sttudies, examining periods over many millions of years estimating CS.

    The key graph is this (you can find it on page 6)


    They are quoting CS as K/W/m2

    To convert to the more common usage of CS as per doubling of CO2 multiply by 3.7. The center points for all the estimates then averages around 3. None are lower than 2. And several are significantly higher.

    That CS is above 1.5 is not falsified. However the the counter assertion, that CS is below 1.5 is falsified.

    This is how proper science is done, including the appropriate use of pepperian falsifiability. And the results from Rohling et al were included in the latest IPCC report.

    panicbusiness. There is no dishonesty among the scientific community regarding AGW. But some parts of the blogosphere are awash with dishonesty. Best to not get ones opinions from those sources don't you think.

  24. It's worth taking note that PB's citations @15 consist of Wikipedia, The Daily Caller and Forbes' Larry Bell.

    PanicBusiness... If you want to be taken seriously you're really going to have to up your game just a touch. It's fine to use Wiki to get a general gist of things, but you'd probably want to dig into the references and read those in addition. 

    As for the Daily Caller and Larry Bell, these are not reliable sources at all. Anything you read there should be followed up with actual research to check for accuracy. 

    Being "skeptical" means putting in some real leg work and being ready to have your own position challenged. Be ready to adjust your beliefs based on what you learn. 

  25. Glenn...  When PB said, "I further urge the community to propose experimets that could potentially exclude models and theories with high climate sensitivity i.e. higher than, say 1.5°C."

    My interpretation was that he's saying "the community" should find ways to only show low sensitivity in the range that he prefers. Note the predetermined conclusion for what CS should be.

    PanicBusiness... Researchers come up with novel methods for how to estimate climate sensitivity, collect the data, run the numbers, and then let the cards fall where they may. You can't frontload the results you want.

  26. As to my references @Rob Honeycutt, I checked, all of them are sufficiently accurate, and I decided that these will do the best with the crowd at hand.


    [JH] Please lose the snark.

  27. Ok we are never getting there so I have to ask directly: How much more discrepancy between the GCM based simulations and real measured surface temperature anomalies is required before it is safe to announce that those models have been falsified*? If you say "it doesn't matter simulations are still true", then you right there failed to accompany the scientific method(falsifiability).

    *real skeptics say they already are.

  28. Ppanic,

    Tamino demonstrates here that there is not any discrepancy between the trend in increasing temperature over the past 30 years and GCM models.  There is an issue with cherry picking the extreme heat year 1998 as start year.  This cherry pick causes it to be not yet statisticly certain that it is still warming for the period of the cherry pick.  In the past 15 years there are no years that are below one standard deviation from the expected trend line and there are two years that are above one standard deviation from the trend line.  

    This is a scientific blog.  Data is required to support all claims.  I have provided data to support my claim that no discrepancy exists.  Please provide data to support your wild claim that there is a discrepancy between observed and GCM simulations.  Your unsupported assertion that such a discrepancy exists is insufficient to support your claim, data is required here at SkS.

  29. PB @26...  

    1) The Daily Caller states that, "United Nations climate chief Christiana Figueres said that democracy is a poor political system for fighting global warming. Communist China, she says, is the best model."

    In fact, Figueres said nothing of the sort. She didn't say anything suggesting anything about political systems at all. She stated, "China, the top emitter of greenhouse gases, is also the country that’s “doing it right” when it comes to addressing global warming."

    The Daily Caller has taken extreme liberties with her intended meaning.

    So, strike one on this source being "sufficiently accurate."

    Strike two would be that none of this has any bearing on what you're apparently citing which is "political regulations (sic)." The best you could possibly claim is it's tangential.

    2) The Larry Bell article is related to a 2011(?) GAO report regarding climate spending and bear no resemblence to what you seem to be attempting to cite, which is "devastating" (?). Devastating what? Spending? Larry Bell is making an absurd case in this article (as per usual) since the climate budget according to the GAO document was $8B out of a $3.8T budget. That amounts to 0.2% of the US budget.

    Strike three. You're out.

  30. Tamino's graph of recent global warming

    Here is a copy of the graph.

    It is your responsibility to provide data to support your wild claim that a discrepancy exists.  I do not see any data links in any of your posts. If you cannot provide data, they are all empty political statements.  Provide data to support your wild claims or stop making them.

  31. Michael Sweet @28 & @30, technically Tamino's graph shows the continuation of the trend 1975-1999 trend for Cowtan and Way's krigged HadCRUT4 temperature index.  That trend is 0.155 +/- 0.11 C per decade.  That is, the trend is just under one SD below the AR4 model mean prediction, and 1.4 SD below the AR5 model mean trend.  Consequently there will be several years more than one SD below mean expected values as determined by the models, and may even be one or two that are two SD below the mean expected value as determined by the models.

    That does not constitute a falsification of the models.  Not even on a naive falsificationist model of science with a naive frequentist approach to statistics does it constitute a falsification.  But the balance of evidence at the moment is that current trends are running 15% below AR4 expectations, and 26% below AR5 expectations.  That has been occuring for long enough now that is is probable that the model mean is running too hot, but only by a small amount.  Certainly not sufficiently hot to warrant setting 1.5 C per doubling as an upper bound on climate model sensitivity.  Indeed, a climate sensitivity 1.5 C per doubling has also not been falsified by observations, but it is a lot closer to being so falsified than is a sensitivity of 3, or even 4.5 C per doubling.

    PanicBusiness' comment about setting 1.5 C as an upper bound on model climate sensitivity (which, as it happens would exclude all models) removes any doubt as to his agenda.  He is not a true falsificationist in science.  Rather, he is using a language he does not appear to understand to claim that any observed values in disagreement with the values he preffers for political reasons are wrong, and on the putative basis that they are unfalsifiable.  Meanwhile his confidence in a low climate sensitivity is so unshakable by observation that he does not see that Rohling et al falsify it (as much as anything can be falsified).  Falsification, obviously, in his practise, is a standard that only applies to beliefs he does not want to accept.

    Clearly there is no point in pretending he is capable of rational discussion.  DNFTT 

  32. PanicBusiness: Your question is poorly formed and even odd: "How much more discrepancy between the GCM based simulations and real measured surface temperature anomalies is required before it is safe to announce that those models have been falsified*?"

    I don't know what exactly "falsifying" GCMs means. It is not as clear cut as "falsifying" the theory that CO2 absorbs and emits infrared radiation at certain wavelengths.

    If what you mean is demonstrating that GCMs' temperature projections are inaccurate, you need to specify what amount of accuracy is sufficient, because of course GCMs' projections are not and never will be 100% accurate, but that fact is not relevant to anything at all.

    When evaluating any theory, it is necessary to specify the framework/goals within which it is evaluated--to what purposes is the theory being put, and is the theory sufficiently useful for those purposes? For example, in the world of physics, string theories so far are not "falsifiable" insofar that nobody has even conceived of a way they could be empirically distinguished from each other or from non-string theories--not even in principle. So string theories rate "0" in the falsifability attribute of theories. (Not exactly 0, because somebody might come up with a way someday. They are not unfalsifiable in such a deeply principled way to get a true 0.) But string theories rate high in the fruitfulness attribute--they have sparked creative thinking even within non-string theories.

    You failed to provide that framework for evaluating GCMs, so I can't answer your question.

  33. Hi, @Rob Honeycutt

    It is remarkable how you fall for the number one communist fallacy. You seem to think that the cost of restricting free markets manifests itself in the cost of creating and enforcing the regulations plus incubating otherwise unhealthy business models. NO. The vast majority of the cost of the climate regulations is in the fact that on the long run free markets are the best possible model for efficient use of the available resources(yes, free markets are the most sustainable when you account for the actual goods and services produced). Now for the actual cost, see what we got here. $8B? really?

    "The Small Business Administration estimates that compliance with such regulations costs the U.S. economy more than $1.75 trillion per year — about 12%-14% of GDP, and half of the $3.456 trillion Washington is currently spending. The Competitive Enterprise Institute believes the annual cost is closer to $1.8 trillion when an estimated $55.4 billion regulatory administration and policing budget is included. CEI further observes that those regulation costs exceed 2008 corporate pretax profits of $1.436 trillion"

    original article

    As for Ms Figueres, here's another reference if you think the last one was an unbalanced interpretation. But I don't think you have room to question that the AGW/CAGW beliefs are strongly tied to devastating political regulations. Here's another:

    "The political divide in the U.S. Congress has slowed efforts to pass climate legislation and is “very detrimental” to the fight against global warming, she said."

    Bloomberg sustainability

    Most of what I say is widely accepted the reason why I still provide references is to give opportunity for readers to catch up on the issue.

    note: Rob's confused comment calls for an answer but I do recognize it is not very beneficial continue ths discussion and I am committed to put an end to it.

  34. PanicBusiness, I continued my response to you on the more appropriate thread Models Are Unreliable.

  35. Okay Tom Daytons' 32 is an excellent start for discussing falsifiability in the framework of GCM based simulations but one thing I want to point out is that if you have a theory you are the one obliged to provide a possibility of falsifiability as to mature your theory into a scientific theory. To stick with tom's example string theorists are indeed having hard time to finding testable predictions (with current technologies) but they are also under heavy pressure from the rest of the theorethical physics community to do so. For a catch up on the controversy you can read Woit's blog.

    Now I do want to make sure that at this point everyone understands the need for a theory to be falsifiable. You can write books about it but this little text sums it up best from wikipedia about Pauli

    "[...] However, this was not his most severe criticism, which he reserved for theories or theses so unclearly presented as to be untestable or unevaluatable and, thus, not properly belonging within the realm of science, even though posing as such. They were worse than wrong because they could not be proven wrong. Famously, he once said of such an unclear paper: "It is not even wrong!""

    P.S. I like the automatic helper on the sidebar that helps people with keywords. It makes writing to a wider audience a lot easier.



    [JH] You are now skating on the thin ice of excessive repitition which is prohibited by the SkS Comments Polcy. All of your assertions about falsifiability have been thoroughly addressed many times over by more than one SkS contributor. Please cease and desist replowing the same ground. If you do not, your future posts will be summarily deleted.   

  36. PanicBusiness:  You implied that I said GCMs are not falsifiable.  That is incorrect.  GCMs of course have the ability to be empirically demonstrated wrong.  But there are many varieties of "wrong," so there are many varieties of my answer to your question.  I can't respond with my answer to your question until you specify which variety of "wrong" you are asking me about.  For example, do you want to know my criteria for concluding, based solely on the match of GCM projections to observations, that the presence of greenhouse gases does not make the Earth warmer than it would be otherwise?  Or do you want to know my criteria for concluding that GCMs over-predict temperature trend by 2 times?  Or what?  Exactly?

  37. PanicBusiness, you missed the point of my example of string theories.  String theories certainly qualify as scientific theories as opposed to supernatural explanations.  That was true despite them not being falsifiable when they were thunk up originally, as is true of a great many theories when they are originally thunk up but have not been subjected to empirical tests capable of falsifying them, and even when nobody has thought of a way to falsify them.  Theories can be very valuable in other ways (e.g., fruitfulness) despite that.  Initially, theories can get a pass on the falsifiability attribute of theory quality, as they can for low scores on other attributes.  They are still scientific theories, and may even be potentially great theories because there is good reason to hope they can be improved in their low-scoring attributes.  Eventually the theories might be abandoned because despite a great deal of work, folks have given up improving their low-scoring attributes or their total low score summed across all the weighted attributes.  They might be "abandoned" meaning set aside because they are not as good as competing theories, rather than being "discarded" because they have been "falsified."  Even a theory that appears to be falsified can be resurrected by realizing that some of its auxiliary theories were the pieces that were wrong. 

    "Falsifiability" is not an unambiguous, monolithic attribute.  It depends on context--framework.  Newtonian physics has been "falsified."  But Newtonian physics has not been falsified in contexts where relativity is not strong enough to be important. Consquently, Newtonian physics is used a lot more than relativistic physics is. Newtonian physics still is a legitimate scientific theory.  It scores poorly in certain attributes, but in most contexts of use those attributes are not very important.

    Falsification is not the only attribute of scientific theories.  It is not necessarily the most important attribute.

  38. PanicBusiness - You have been referring to GCMs and falsifiability; this is however an inappropriate terminology. Global circulation models are simulations of physics, evaluations of how those physics and the climate state might evolve over time, but they are not in themselves either hypotheses or theories. 

    GCMs are models, and models in general are always 'wrong' in that they do not contain the entirety of the physics, the details, and in that there will always be errors. The question with models is whether or not they are useful. It may very well be that the current generation of GCMs are incomplete in aspects of the climate that make them inaccurate - insufficient accounting for variability or modes thereof, or (as in the case of many models) run with inaccurate forcings or temperatures. There's certainly a significant literature pointing in that direction, as with England et al 2014 wrt variability or Contan and Way 2014 wrt temperature measures.

    The only judgement you can make based solely on model output is whether or not they are accurate enough to be useful.

    Global warming theory and the anthropogenic influence, on the other hand, is entirely falsifiable. Predictions include night warming faster than day, winter warming faster than summer, warming troposphere and cooling stratosphere, polar amplification, the changes in top of atmosphere forcings with changing GHGs, the sum climate energy increase as seen in ocean heat content, etc. If these predictions failed, there would be evidence against the theories. 

    Those predictions have, however, been validated within the limits of the data available - while falsifiable, they have not been. And you have not been discussing the theories themselves in any fashion whatsoever.

  39. It appears PanicBusiness "real" skeptics were fooled by strawman predictions, cherry-picked starting point and false baselining. What happened to the "skepticism" when assessing those claims?

  40. Okay I will try to be as clear as possible.

    I personally find it very likely that in the coming five years there will be no significant warming or there will even be significant cooling. If that happens I want the CAGW community to not come up with additional excuses, and hand-waving like it was totally expected.

    I want the AGWers (who are on supposed consensus) to state their predictions now and clearly as to how much surface temperature warming will happen in function of CO2 emissions with confidence intervals in the future. This is how you make predictions.

    If, for example, there will be 0.4°C cooling in the next 30 years I want to be able to say this is what they said and they were wrong. Or if they do not have the courage to make any meaningful prediction or just extremely vague ones, I want to be able to say they are not even wrong.

    thank you for reading.

  41. My thoughts on how to falsify anthropogenic global warming.
    Like any other theory, you falsify by demonstrating that observations are at variance with the tenets of theory. So let look at the tenets how they could be falsified:
    1/ the increase in CO2 is due to human emissions.
    Falsified if a/ showing insufficient emissions to account for the increase. b/ show the isotopic signature is inconsistant with emission.
    2/ The rising CO2 concentration is causing increased LW irradiation of surface and consequent change in outgoing LW; and that radiation will have the spectral signature of CO2.
    Falsified by ground or space spectrometers readings inconsistent with calculated signature. Pyrgeometers not registering increased LW.
    3/ In accordance with conservation of energy and Planck's Law, the increased LW irradiation will warm the oceans and land surfaces. Because most of the surface is covered by liquid however, and ocean/atmospheric processes have a big influence on surface temperatures, an equilibrium temperature will take 100s of years to be obtained. While this is happening though, the ocean heat content will rise at a rate consistent with the energy imbalance.
    Falsified by OHC decreasing or stable; falling or stable 30-year surface temperature trends.

    Models may not be perfect, they are always wrong in some ways, but they remain the most skillful way to predict the future climate that we have. If you dont like models, then you have to fallback on less skillful methods.

  42. I want to have a document from 2014 stating what the "settled science" actually says. As a reference. And see what happens.

  43. Okay then PanicBusiness:  What you are asking for is what the GCMs produce, and which are described (and should be commented on) in the post Models Are Unreliable.  See my comments to you there.

  44. "I want the AGWers (who are on supposed consensus) to state their predictions now and clearly as to how much surface temperature warming will happen in function of CO2 emissions with confidence intervals in the future. This is how you make predictions."

    Firstly, what part of "models have no skill at decadal level prediction" do you not understand?  However, if you understand that models do not predict ENSO, then I think it is possible to make reasonable predictions for a 5 year trend given a value for the ENSO index. Ie, to make predictions of the form "if ENSO value is this, the aerosols are Y, TSI is Z, then temperature will be T +/-"

  45. I would expect climate scientists to say, yes we were wrong and to continue to investigate to find out why.

    I would also ask the corrallary of your question. I would expect the next El Nino with an index of 1.8 or more to register record global temperatures on all current indexes including tropospheric satellite measurements. If this happens, will you change your stance and ask yourself what measures you would support to reduce global emissions?

  46. PB @33... I've continued this conversation on a more appropriate thread. Here.

  47. PB: "I personally find it very likely that in the coming five years there will be no significant warming or there will even be significant cooling. If that happens I want the CAGW community to not come up with additional excuses, and hand-waving like it was totally expected."

    Can you tell me how you arrived at that prediction?  I'd like to repeat the method to see if I get the same result.

  48. DSL - I agree - I would love to hear the basis on which PB forms his/her opinion.

  49. All: PanicBusiness has been banned from further posting on the SkS website. The person behind the PanicBusiness screen is the same person the was behind the Elephant In The Room screen. Sock puppetry is strictly prohibited by the SkS Comments Policy. Persons engaging in sock puppetry automatically lose their posting privileges.

  50. We need a volunteer to write an OP for this comment thread.  

1  2  Next

Post a Comment

Political, off-topic or ad hominem comments will be deleted. Comments Policy...

You need to be logged in to post a comment. Login via the left margin or if you're new, register here.

Link to this page

The Consensus Project Website


(free to republish)

© Copyright 2020 John Cook
Home | Links | Translations | About Us | Privacy | Contact Us