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 Support

Twitter Facebook YouTube Pinterest MeWe

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


How reliable are climate models?

What the science says...

Select a level... Basic Intermediate

Models successfully reproduce temperatures since 1900 globally, by land, in the air and the ocean.

Climate Myth...

Models are unreliable

"[Models] are full of fudge factors that are fitted to the existing climate, so the models more or less agree with the observed data. But there is no reason to believe that the same fudge factors would give the right behaviour in a world with different chemistry, for example in a world with increased CO2 in the atmosphere."  (Freeman Dyson)

Climate models are mathematical representations of the interactions between the atmosphere, oceans, land surface, ice – and the sun. This is clearly a very complex task, so models are built to estimate trends rather than events. For example, a climate model can tell you it will be cold in winter, but it can’t tell you what the temperature will be on a specific day – that’s weather forecasting. Climate trends are weather, averaged out over time - usually 30 years. Trends are important because they eliminate - or "smooth out" - single events that may be extreme, but quite rare.

Climate models have to be tested to find out if they work. We can’t wait for 30 years to see if a model is any good or not; models are tested against the past, against what we know happened. If a model can correctly predict trends from a starting point somewhere in the past, we could expect it to predict with reasonable certainty what might happen in the future.

So all models are first tested in a process called Hindcasting. The models used to predict future global warming can accurately map past climate changes. If they get the past right, there is no reason to think their predictions would be wrong. Testing models against the existing instrumental record suggested CO2 must cause global warming, because the models could not simulate what had already happened unless the extra CO2 was added to the model. All other known forcings are adequate in explaining temperature variations prior to the rise in temperature over the last thirty years, while none of them are capable of explaining the rise in the past thirty years.  CO2 does explain that rise, and explains it completely without any need for additional, as yet unknown forcings.

Where models have been running for sufficient time, they have also been proved to make accurate predictions. For example, the eruption of Mt. Pinatubo allowed modellers to test the accuracy of models by feeding in the data about the eruption. The models successfully predicted the climatic response after the eruption. Models also correctly predicted other effects subsequently confirmed by observation, including greater warming in the Arctic and over land, greater warming at night, and stratospheric cooling.

The climate models, far from being melodramatic, may be conservative in the predictions they produce. For example, here’s a graph of sea level rise:

Observed sea level rise since 1970 from tide gauge data (red) and satellite measurements (blue) compared to model projections for 1990-2010 from the IPCC Third Assessment Report (grey band).  (Source: The Copenhagen Diagnosis, 2009)

Here, the models have understated the problem. In reality, observed sea level is tracking at the upper range of the model projections. There are other examples of models being too conservative, rather than alarmist as some portray them. All models have limits - uncertainties - for they are modelling complex systems. However, all models improve over time, and with increasing sources of real-world information such as satellites, the output of climate models can be constantly refined to increase their power and usefulness.

Climate models have already predicted many of the phenomena for which we now have empirical evidence. Climate models form a reliable guide to potential climate change.

Mainstream climate models have also accurately projected global surface temperature changes.  Climate contrarians have not.

Various global temperature projections by mainstream climate scientists and models, and by climate contrarians, compared to observations by NASA GISS. Created by Dana Nuccitelli.

A 2019 study led by Zeke Hausfather evaluated 17 global surface temperature projections from climate models in studies published between 1970 and 2007.  The authors found "14 out of the 17 model projections indistinguishable from what actually occurred."

There's one chart often used to argue to the contrary, but it's got some serious problems, and ignores most of the data.

Christy Chart

Basic rebuttal written by GPWayne

Update July 2015:

Here is a related lecture-video from Denial101x - Making Sense of Climate Science Denial

Additional video from the MOOC

Dana Nuccitelli: Principles that models are built on.

Last updated on 9 September 2019 by pattimer. View Archives

Printable Version  |  Offline PDF Version  |  Link to this page

Argument Feedback

Please use this form to let us know about suggested updates to this rebuttal.

Further reading

Carbon Brief on Models

In January 2018, CarbonBrief published a series about climate models which includes the following articles:

Q&A: How do climate models work?
This indepth article explains in detail how scientists use computers to understand our changing climate.

Timeline: The history of climate modelling
Scroll through 50 key moments in the development of climate models over the last almost 100 years.

In-depth: Scientists discuss how to improve climate models
Carbon Brief asked a range of climate scientists what they think the main priorities are for improving climate models over the coming decade.

Guest post: Why clouds hold the key to better climate models
The never-ending and continuous changing nature of clouds has given rise to beautiful poetry, hours of cloud-spotting fun and decades of challenges to climate modellers as Prof Ellie Highwood explains in this article.

Explainer: What climate models tell us about future rainfall
Much of the public discussion around climate change has focused on how much the Earth will warm over the coming century. But climate change is not limited just to temperature; how precipitation – both rain and snow – changes will also have an impact on the global population.


On 21 January 2012, 'the skeptic argument' was revised to correct for some small formatting errors.


Prev  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  Next

Comments 501 to 550 out of 1297:

  1. Relevant to this discussion on skill/validation of models is Hargreaves 2010
  2. Scaddenp brought me here to answer his question, so here I am. I said that extraordinary claims require extraordinary evidence. Sccadenp asks "Sorry, what is extraordinary about the claims of climate theory?" Here are two examples: "And that is what is at stake: our ability to live on planet Earth, to have a future as a civilization." Al Gore, An Inconvenient Truth. "On the long run, if this [increase in greenhouse gases] continues for centuries, that's it for all the species on this planet." John Hansen, The Runaway Greenhouse Effect (YouTube video) I call these claims extraordinary and I expect an extraordinary amount of model validation before I accept the consequences, which are also extraordinary. And, for hearing too many failed season-long forecasts from our Canadian chief meteorologist, I do not believe a minute that we understand all there is to understand about the physics of climate. Read this for fun: Balmy winter takes climate experts by surprise
    Response: Regarding your last paragraph, you need to understand that weather is not climate. See Scientists can’t even predict weather
  3. Perhaps, Manny, if these are indeed "claims of climate theory", as you propose, you could furnish us with the references in the peer-reviewed literature where either of these claims are made. I contend that you can't do this, as the first claim is made by a politician in a film, and the second is a personal opinion by (James) Hansen in a youTube video. See if you can find either statement in the published literature or in the IPCC reports. What you will find in the published literature: extensive discussions of empirical evidence documenting the causes and impacts of climate change, with uncertainties attached; discussions, with evidence and uncertainty, of how sensitive our climate is to change (e.g. 2C-4.5C warming per doubling CO2); discussion of model representations of the climate system, with assessment of their strong and weak points. Models validate the core propositions of our theory of climate very well. Seasonal weather forecasts are an irrelevant distraction, being both notoriously unreliable and bearing precious little in common with climate modelling. I wonder why 'skeptics' would wish to conflate the two?
  4. "I call these claims extraordinary". Why? Would "firing off our entire nuclear arsenal at once might lead to mass extinctions", or "a 10km asteroid impact, will lead to mass extinctions" be extraordinary claims? I would say no - both a consistent with known science. The context for that original quote and its basis from Laplace, refer to ideas that are in breach of all known science. The denialist community is making the extraordinary claim that modifying our atmosphere with a known radiative gas is somehow, in defiance of quantum mechanics and laws of thermodynamics, not going to result in a warming climate. The question over model skill is whether they are better than predicting the future than a naive assumption.(eg that man cant affect climate). The models demonstrably have that skill. Instead, Manny, you seem prepared to bet the future on the basis that known physics is wrong. I rather doubt you make similar bets against in science in other spheres (eg what you Dr tells you). Would this be because you perceive that any solution would violate your political ideals?
    Response: TC: Text edited to change all capitals into bolded. The comments policy applies for everyone.
  5. Manny's argument, if it can be called one, cuts both ways. Here is an extraordinary claim: We can burn fossil fuels without regard to our impact on the global environment. Where is the extraordinary evidence that supports this extraordinary claim? Or do we just carry on with business as usual, on the sole basis of 'that's what we've always done'? Particularly when we know without question that our actions do indeed impact the environment in both directions - see ozone, smog, the Black Triangle, the Clean Air Act, Asian brown clouds, etc.
  6. Continued from here In other words, Clyde has no evidence what so ever, only conjecture and assumption. A question for you, Clyde: Which would you find more believable as a climate modeler, a climate scientist, or rather team of scientists, who are also competent at building a computer model of the physics and chemistry of the coupled land-ocean-atmosphere system, or an expert at computer modeling who knows absolutely nothing about the underlying physics and chemistry of the coupled land-ocean-atmosphere system that is being modeled? (Why should they, they are not climate scientists, they're computer modelers, while practically all physical scientists work intensively with computers.) And why did you fail to consider that climate scientists work with actual computer modelers when developing climate models, as I suggested, before leaping to defend your assertion that climate scientists have no expertise in the use of computers?
  7. Jim Eager 506# I've never seen any credit given to computer modelers. Would you want heart surgery done by a doctor who has performed say 100 successful operations or by a doctor who has used a computer to project/predict how to do the operation? Would you bet your life on the future global warming projections/predictions coming from computers? I read a few other blogs. As i said earlier the science is above my head. I know enough to understand a failed prediction. Do you want me to post the links to the failed predictions/projections? From my brief time of reading this blog Rodger Pielke Sr is not one of the favorites around here. Why doesn't somebody ( you if your qualified) refute his claims on climate model predictions/projections? I'm not saying Rodger is right or wrong. He has an open challenge to prove him wrong & nobody has taken him up on it. He admits when he is wrong. He had to eat crow after a discussion he had with dana1981. You can input all the physics & chemistry you want into a computer. Doesn't mean what comes out is accurate. Have a nice day
  8. "heart surgery done ... by a doctor who has used a computer to project/predict how to do the operation?" What nonsense. Medicine uses computer imaging, based on models of how the body interacts with magnetic fields, sound and/or radiation. Oil exploration uses computer modeling based on seismic techniques. Is there a car or airplane that can function without a computer? Why is climate science held to the false standard that 'computer models don't work'? "You can input all the physics & chemistry you want into a computer. Doesn't mean what comes out is accurate." No, but it does mean that we can eliminate things that can't be verified by models. We can eliminate the idea that climate change is purely natural.
  9. Clyde, what on earth makes you think modelling teams dont include heavy-duty modelling folk? As to "failed predictions" that pick up off blogs, there are a couple of things to check. First, check the source of the prediction. The usual denialist stuff is make claims about a prediction that are not actually made and since its a straw man, (take note of error bars) then its easy to demolish. All models are wrong, but some predictions are far more robust than others. A converse page of robust model predictions together with papers that do the prediction and papers that confirm it can be found here Second, modellers usually judge models by skill. Ie the ability of models to make better predictions than some simpler method (ie that nothing is changing). Climate models have no skill for instance at decadal-level predictions. This is common "skeptic" ploy. As to Piekle, perhaps you should follow the discussion with the modellers at Realclimate? In short, you cannot make naive comparisons of models and observations. If you still think there is clear case of model "being wrong" supported by papers, then by all means post links.
  10. Clyde: What is your definition of a "computer modeler"? On what basis do you claim that any particular "climatology expert" is not knowledgeable about computer modeling, and how would this affect the work that they are doing? I think that you are creating a strawman "computer modeling expert", in a futile attempt to pretend that climatologists can't do "computer modeling". Many climate modelers have physics and mathematics in their background, and at least one I know personally works in a mathematics department. I would consider myself to be a "climatologist", and I have written "climate models" (microclimate) from scratch. My background is physical geography, but - I have studied numerical methods, including finite difference solutions to partial differential equations. - I have coded numerical solutions to radiative transfer, atmospheric turbulent heat transfer, and soil thermal diffusion problems. - I know from study and experience the issues related to floating-point arithmetic, and how to avoid them. - I have coded the required root-finding procedures for complex, non-linear systems. - I wrote my first computer program before I took my first computer course, and before most people had a clue what "Computer Science" was (or would be). I never took a second computer course, but I have written serious code in three different languages, simple code in a few more, and can read read quite a few more than that. And I know that "Computer Science" students often don't get exposed to half the stuff that I have learned that is necessary to do "climate modeling" correctly. - my first numerical programming was done on punch cards, fed into a mainframe computer. I've been doing programing for almost 35 years. What else would I need in my background to convince you that I know enough about "computer modeling"? I seemed to convince my PhD thesis examining board that the modeling work I did was valid - and that board included an engineer that asked me why I used a secant root finding method instead of Newton-Raphson, and a physicist that said that when he read my thesis he discovered a field of study that he could have been very happy doing instead of physics. I agree that the science is above your head. It doesn't have to stay that way if you are willing to learn. Start by giving me your definition of "computer modeler", and what makes you think that people doing climatology aren't capable of it...
  11. Clyde: you said "I know enough to understand a failed prediction. Do you want me to post the links to the failed predictions/projections? Please do so, but keep in mind that here at SkS you will be expected to back your position up with references to real scientific literature (not just blog posts). Before you start to post your own stuff, though, you may want to review the series of posts found using the Lessons from Predictions search item. There is a button that will do this search for you near the top left of the SkS page (just above "Most Used Climate Myths").
  12. Last comment for now... Clyde: if you want to see why the name of Pielke Sr. gets the reaction it does here, read the discussions available here, where he has participated in some comment threads, and there have been numerous blog posts commenting on his arguments. Note that Pielke Sr. does not seem to allow comments at his own blog, so it isn't easy to engage in a discussion with him about them. I've already done some searching for you (the search box is in the upper left corner of the SkS page). You can try reading these blog posts and discussions: Pielke Sr and scientific equivocation: don't beat around the bush, Roger Response to Roger Pielke Sr. One-Sided 'Skepticism' Chasing Pielke's Goodyear Blimp SkS Responses to Pielke Sr. Questions Pielke Sr. Agrees with SkS on Reducing Carbon Emissions Pielke Sr. and SkS Disagreements and Open Questions Pielke Sr. and SkS Warming Estimates Pielke Sr. and SkS Dialogue Final Summary Pielke Sr. Misinforms High School Students
    Response: (Rob P) Removed dead link
  13. Just on the RC discussion you might like to look at this comment and surrounding context.
  14. Clyde, you appear to be attempting to make a case against climate science (or at least, echoing Dr Pielke Sr's case) on the basis of its use of computer models to project future states of the climate. This is likely to be an exercise in futility on your part. As I intimated on the Bob Carter thread, the mainstream scientific position on climate is the result of an intertwining web of: (1) Physics & chemistry theory (ranging from quantum-mechanic radiative properties of IR-trapping gases, to the physics of blackbodies, to the chemistry of ocean buffering of CO2, to many other strands of theory besides); (2) Lab or computer experiment (starting with Tyndall's experiments demonstrating the atmospheric IR-trapping gases and confirming the atmospheric IR-trapping 'greenhouse' effect all the way to the elaborate atmosphere-ocean-surface coupled models of the present); (3) Empirical observation (the surface, satellite & sea surface temperature datasets, measured ocean heat content, measured crysophere melt, shifting wind currents & atmospheric cells, shifting animal & plant distribution, and so on). It it this intertwining of these various strands of evidence which has led to the formation of the generally-accepted scientific consensus on climate change, as expressed by the IPCC, the US NAS, the UK Royal Society, and virtually every major national, transnational, or intranational scientific organization. Even supposing you could chip away at the reliability & accuracy of computer models, you would still have an enormous task ahead of you to knock down enough of the theory, experiment & observation supporting the mainstream consensus to cause a substantial re-think.
  15. Clyde asks: "Would you want heart surgery done by a doctor who has performed say 100 successful operations or by a doctor who has used a computer to project/predict how to do the operation?" This is an obviously bogus analogy. Heart surgery is performed routinely these days, so it is easy to find heart surgeons who have performed 100 successful operations. However climatology as a field is only a couple hundred years old and it takes at least 30 years worth of observations to get a clear picture of what the climate is doing, thus there is no climatologist that has made 100 successful (independent) predictions of future climate. Sadly observations of future climate are not available at the current time, so projections based on computer simulations are the best tool we have at the moment for exploring the consequences of a given course of action (or inaction). If Clyde has a better solution, then lets hear it. BTW, if computer simulations were of no value, then one would ask why computer simulation is widely used in training surgeons?
  16. Clyde wrote: "I've never seen any credit given to computer modelers." How would you know this? What, exactly, is your definition of a computer modeler? What is your expertise in determining who is a competent computer modeler? "Would you want heart surgery done by a doctor who has performed say 100 successful operations or by a doctor who has used a computer to project/predict how to do the operation?" Your analogy actually better applies to you. It seems you would trust the expert modeler who has no understanding of the physical climate than the climatologists who actually study the real climate. Where do you think climatologists working with computer modeling learn enough about the the physics and chemistry of the real climate to model it? "Would you bet your life on the future global warming projections/predictions coming from computers?" Yes, I would and I am. Would you bet yours, or more telling, your children's and your grandchildren's lives on going forward with business as usual without understanding the possible impacts of tampering with the atmosphere and greenhouse effect and ocean chemistry and without bounding the probability of those impacts by modeling them? "Do you want me to post the links to the failed predictions/projections?" Do you mean failures like successfully predicting... ... that global mean temperature would warm, by about how fast, and by about how much ... the rising of the tropopause and the effective radiating altitude ... that the troposphere would warm while the stratosphere would cool ... that night time surface temperatures would increase more than daytime temperatures ... that winter surface temperatures would increase more than summer temperatures ... that higher latitudes would warm faster than temperate and equatorial latitudes (polar amplification) ... that the Arctic would warm faster than the Antarctic because the two poles are physically and geographically quite different ... the magnitude (~0.3 C) and duration (~two years) of the aerosol cooling caused by the 1991 Mt. Pinatubo eruption ... that modeled hindcasts for Last Glacial Maximum sea surface temperatures was inconsistent with the paleo evidence, and then better paleo evidence showed that the models were right ... a trend significantly different and differently signed from the UAH satellite temperature record, and then a bug was found in the UAH satellite data ... a tropospheric temperature trend significantly different and differently signed from the balloon radiosond temperature record, and then it was found that the thermometers used on the balloons were not properly shielded from direct sunlight ... the ~4% increase in absolute humidity as the atmosphere warms (water vapor feedback) ... the increase in both number and intensity of record high temperature events ... the increase in both drought intensity and intense precipitation events ... the response of southern ocean winds to the ozone hole ... the northward expansion of Hadley cell circulation ... the expanded range of hurricanes and tropical cyclones, poleward movement of storm tracks, and the increase in average cyclone & hurricane energy intensity It's "failures" like these that give me confidence that the models are useful. Not that they are right, but useful.
  17. And, finally, Clyde, remember that any claim you make contrary to these computer models must be based on a model itself. It may be your intuition, or it may be someone else's computer model, but it's still a model. Unless, of course, you give voice to thoughts which have no rational origin, and I don't think you'd admit to that (or your admission wouldn't mean anything if it were true). Thus, just saying "models suck" is not good enough. You need to be able to defend your own model (or the model you currently accept) against the IPCC's model set, or you're just barking loudly without any teeth.
  18. Ben Santer has an interesting lecture where he discusses climate modeling HERE. Santer states that no model is perfect, they're good but not perfect. He says that some models are better than other models, but ensembles of the models is better than any given model. And they also find that ensemble models that are weighted for the better individual models perform even better than that. And you have to realize, the models are not being put out there without checking them against empirical results. They, in fact, are tested against actual results, but as Dikran says, it takes a long time to test the models against the broad, long term climatic response to forcing. What I think is going on in the media and the blogs is, it's a meme that plays on people's lack of understanding of a complex science. It's the "hey, they can't even predict the weather next week" idea, which is a completely false analogy. So, people get the idea that climate change is all based on models and models don't even work. It's the mother of all red herrings in the climate debate.
    Response: TC: I think Clyde has more than sufficient responses to his comments now. Any further responses before he responds would by "dogpiling", and hence in contravention of the comments policy. In order to to avoid overwhelming Clyde with weight of respondents rather than weight of argument, I also request that only Bob Loblaw and Jim Eager respond to his future posts, unless they wish to deffer to some other person.
  19. I appreciate all the help everybody is giving. I can't reply to everybody. Thanks to the moderator who has ask the "dogpiling" be kept down. Bob Loblaw 510 Which is harder/more complicated - Writing code for GCM or writing HTML codes for a website?
  20. Bob Loblaw 512 I have no idea if Pielke Sr is right or wrong when he issued the challenge. All i know is if anybody has taken him up on it he hasn't posted least not yet. The link below is not the challenge (I'll try to find it later) but has several peer reviewed papers casting doubt on predicting future climate based on computer models. Read more here. I've read articles that Kevin Trenberth has a theory where the "missing heat" went. Another example of the models not being very accurate. How can you just a model[s] that missed 10 years of heat? My first try with the hyperlink deal. Hoping this works.
  21. Jim Eager 516 I guessing if you had a peer reviewed paper you would give credit to all who helped in writing of the paper IE any folks who had been computer modelers by trade. My definition of a computer modeler would be somebody who's primary job was making computer models. I might have more trust in the models if i knew the scientist running them had expertise in computer modeling. So yes i would prefer the scientist give the data they want modeled to somebody who's primary job is computer modelling. No i wouldn't bet my life on computer model predictions. My using the betting of life is not a good thing to do IMO. My apologies to all. (-Snip-)
    Response: [DB] Off-topic snipped.
  22. Sorry about the post above. More evidence the computer models are not good at predicting future climate change. One minor "type O" & things come out wrong.
  23. Well, since Clyde refuses to acknowledge all the things that climate models have gotten right, and that they are useful for understanding how climate behaves in the present and will behave in the future, or that scientists can even be competent climate modelers, and is now directing people to Jo Nova's site to support his argument, as per the moderator's request I hereby defer my right of reply to anyone else wishing to waste their time continuing the discussion with him.
  24. Clyde, your (now deleted) Arctic comment has been responded to on a more appropriate thread.
  25. Clyde#521: "I might have more trust in the models if i knew the scientist running them had expertise in computer modeling." How do you know they do not have such expertise - or do not have available those who do? It would appear you missed this model quality-check from the Intermediate version of this argument. If your argument with computer models is that 'scientists don't know how to program,' that song is old and tired. I suggest you look here or here. If your argument is 'you can't trust a computer model,' I suggest you look here to see how computer models impact medicine and here to see how pervasive computer models are. If you distrust modeling so much, be prepared to give up a lot of what we now take for granted as part of our 'quality of life.'
  26. Clyde, so far, your case is entirely evidence-free regarding: 1: what you regard as a specialist computer modeller? 2: what expertise such a person would have that it is impossible for a climate modeller to acquire? (who will spend most of their research life from at least PhD onwards writing code, usually beginning with a background in physics or earth/environmental science). 3: why somebody who has spent many years, perhaps even decades, coding climate models would not then be an expert in coding? Are there some secrets that they never let out of the computer science department? I've coded small climate models, and I've worked with a number of people who professionally code much more sophisticated ice sheet and climate models - you baselessly insult their intelligence and hard-won expertise. I've seen physicists attempting to code ice sheet models and making a pigs ear of it in the first instance because they didn't have the understanding of earth systems, glacier dynamics and climate to make such a thing realistic (they learned, gradually!). I've seen the reverse - the physics expertise is equally hard-won. I have not seen a computer scientist even have a go because they would need to get to grips with two major things: how a climate system works, and the mathematics of the thermodynamics/mechanics required to calculate components of the model. Quite how you think the pure coder who can write very tidy Fortran/C++/Python could do this better than the existing experts is remarkable. It is much easier to begin with an understanding of climate and physics, and graduate onto writing computer code, which is fundamentally not that difficult to do, than the alternative.
  27. Clyde @552 links to an atrocious analysis by David Evans, who by all accounts (particularly his own) is an expert in computer modeling. Evans criticizes two models which are supposedly representative of IPCC model predictions, the 1988 prediction by Hansen, and the projections by the IPCC First Assessment Report (FAR). He says of them:
    "The climate models have been essentially the same for 30 years now, maintaining roughly the same sensitivity to extra CO2 even while they got more detailed with more computer power."
    Oddly, in the IPCC Third Assessment Report (TAR) we read:
    "IPCC (1990) and the SAR used a radiative forcing of 4.37 Wm-2 for a doubling of CO2 calculated with a simplified expression. Since then several studies, including some using GCMs (Mitchell and Johns, 1997; Ramaswamy and Chen, 1997b; Hansen et al., 1998), have calculated a lower radiative forcing due to CO2 (Pinnock et al., 1995; Roehl et al., 1995; Myhre and Stordal, 1997; Myhre et al., 1998b; Jain et al., 2000). The newer estimates of radiative forcing due to a doubling of CO2 are between 3.5 and 4.1 Wm-2 with the relevant species and various overlaps between greenhouse gases included. The lower forcing in the cited newer studies is due to an accounting of the stratospheric temperature adjustment which was not properly taken into account in the simplified expression used in IPCC (1990) and the SAR (Myhre et al., 1998b). In Myhre et al. (1998b) and Jain et al. (2000), the short-wave forcing due to CO2 is also included, an effect not taken into account in the SAR. The short-wave effect results in a negative forcing contribution for the surface-troposphere system owing to the extra absorption due to CO2 in the stratosphere; however, this effect is relatively small compared to the total radiative forcing (< 5%). The new best estimate based on the published results for the radiative forcing due to a doubling of CO2 is 3.7 Wm-2, which is a reduction of 15% compared to the SAR. The forcing since pre-industrial times in the SAR was estimated to be 1.56 Wm-2; this is now altered to 1.46 Wm-2 in accordance with the discussion above. The overall decrease of about 6% (from 1.56 to 1.46) accounts for the above effect and also accounts for the increase in CO2 concentration since the time period considered in the SAR (the latter effect, by itself, yields an increase in the forcing of about 10%)."
    (My emphasis) A 15% reduction in estimated climate sensitivity is not "roughly the same sensitivity". What is more, early climate models included very few forcings. Evan's comment on that in his video saying (falsely) that they only include CO2, and do not include natural forcings. However models used in the Third and Fourth Assessment reports most certainly used natural forcings, as well as a wide range of anthropogenic forcings. Therefore the claim that "[t]he climate models have been essentially the same for 30 years now" is simply false. More troubling is the graphic Evan's uses: First we have the label indicating the projections dependent on CO2 emissions as if CO2 was the only forcing modeled by Hansen. Indeed, in the video, Evans explicitly states just that, ie, that CO2 was the only modeled forcing. In fact Hansen included five different anthropogenic gases in each model run, so checking just CO2 emissions does not check how well reality conformed with any particular scenario. Far worse, he labels scenario A as "CO2 emissions as actually occurred". What actually occurred, and entirely unpredicted by Hansen, was that the Soviet Union collapsed resulting in a massive reduction of very polluting Soviet Block industry, with a consequent massive reduction of CO2 emissions from the Soviet Block: As a result, current CO2 levels (ignoring seasonal variation) are only 390.5 ppmv, which compares to the 391 ppmv projected by Hansen for 2011 in scenario B. In other words, Evans is claiming that CO2 emissions followed scenario A whereas in reality they have not yet caught up to scenario B. Here are the current concentrations of the other GHG used in Hansen's model: Gas | Actual__ | Hansen (Scenario) ___CH4 | 1810 ppb | 1920 (Scenario C) ___NO2 | _323 ppb | _330 (Scenario B) _CFC11 | _240 ppt | _275 (Scenario C) _CFC12 | _533 ppt | _961 (Scenario B) So, for every gas modeled, the actual 2011 concentration is greater than the projected scenario B concentration, often much greater. In two cases, even the scenario C projected concentration is greater than the actual concentration; yet Evans says that Scenario A emissions is what happened. Given the size of the discrepancies, there are only two possibilities. Either Evans did not bother looking up the data before making his assertion - an assertion he has made repeatedly while strongly emphasizing his expertise. Or he is flat out lying. Seeing Clyde introduced Evans' rubbish to this discussion, he now needs to answer several questions: Do experts make assertions about data which they have not bothered looking up? Do they lie? And why, given that they are supposedly so skeptical, have no fake "skeptics" picked up on these errors and criticized Evans for them? Finally, for a proper analysis of those predictions, I recommend the posts by Dana on Hansen's 1988 predictions, and on the predictions of the First Assessment Report. I don't think Dana claims to be an expert on climate modeling, but at least he treats the data with integrity.

    TC: I again request that we avoid dogpiling. As Jim Eager has withdrawn, I ask that only Bob Loblaw and Skywatcher, as people with directly relevant experience to the topic make further responses to him. I ask that posters forgive my slight hypocrisy also responding, but I am sure you will understand my distaste for the introduction of Evans' effort in what hopes to be a conversation guided by evidence and aiming at truth. In the event, future responses to Clyde other than by Bob Loblaw or Skywatcher will be deleted as dogpiling, unless either withdraws, in which case Muoncounter can take up the cudgels.

    [DB] Lest some think that this is moderation by fiat by TC, this action has the full support of the moderation staff of SkS. TC has merely implemented a jointly discussed and approved action.

  28. Tom Curtis “How reliable are climate models?” (# 527: 10.38 am 1 June 2012) To Tom Curtis: I’m seeking clarification here, so I hope these question does not fall into the category of “dogpiling” (a term I’m loathe to use, given its homophonic noun alternative). I note your criticism of David Evans’ chart overlaying Hansen’s 1988 predictions with NASA satellite data of global air temperatures, where you demonstrate that the current CO2 level corresponds closely to Hansen’s prediction for Scenario B. Is your main point in that part of your analysis, that Evans should have noted that Scenario B “actually occurred”? Or is there more that I have missed? I’m a bit confused by your paragraph “So, for every gas modeled, the actual 2011 concentration is greater than the projected scenario B concentration, often much greater. In two cases, even the scenario C projected concentration is greater than the actual concentration; yet Evans says that Scenario A emissions is what happened.” Does the four row table above that paragraph, show the 1988 figures used in Hansen’s modelling? My last question is whether there is any issue with the Evans chart of the actual NASA air temperatures – the black line? I’d appreciate your help. Thank you.
  29. (snip) As to JoNova/David Evans misinformation - well look around Skepsci for take downs, (eg hot spot and Evans (snip) Hansen's 1984 model - yes it had sensivity wrong for well understood reasons. see Lessons from past predictions 1981 (and rest of that series for interest). And yes, climate sensitivity is still uncertain, but very unlikely to be less than 2 (or more than 4) - but claiming a past prediction is falsified by data on sensitivity doesnt fly when sensitivity wasnt a robust prediction. (snip)
    Response: TC: With regret, comments sniped for compliance with the no dogpiling rule. Explicit discussion of Hansen 1988 was retained as being a relevant response to my post,rather than to Clyde. If either of the two current respondents to Clyde with to step aside in your favour, I shall restore your comment (if the html code is working as it should). In that respect I note that your expertise is in computer modelling of petroleum basins.
  30. TC - Jim Eager stepped aside 523 so I thought I would continue. Your excellent post was composed as I was composing mine.
    Response: TC: It happens. If it were me I was asking you to stand aside, I would step aside in that your post was more directly relevant to Clyde's argument, and you are far more qualified on this topic than I. As it happens, Skywatcher beat you to the punch. I consider my post an aside to rebuff the use of outrageously flawed misinformation.
  31. With Jim Eager stepping back and several others chiming in, and another request to avoid dogpiling, for the moment I will restrict my comments to a few direct statements from Clyde. First "Which is harder/more complicated - Writing code for GCM or writing HTML codes for a website? I consider this to be a rather misleading question. It's like asking "which is harder to edit? A book written in English on the growth of multinational corporations, or a book written in German on how to rebuild the engine of a Leopard tank? Both require editing skills, but one benefits from a knowledge of economics and business (and the English language), while the other is easier if you have a detailed understanding of internal combustion engines, tools (and can read and write German). ...but, to answer your question: - given the specifications of the required procedures, any competent programmer can likely implement the algorithms. - the major stumbling block is in determining the algorithms to use, and someone who knows climatology and numerical methods will do better at a GCM, and someone that knows HTML and graphics displays (or whatever the web page is supposed to do) will do better on the web page. In my personal experience (writing web pages by creating HTML in a text editor), the programming skills I developed in my climatology career made HTML a trivial exercise (for relatively simple web pages), and programmers that only know web development do a poor job at any sort of scientific/numerical programming.
  32. Peter42 @528, the "no dogpiling" rule is designed to avoid people, normally new questioners and/or skeptics, from facing an overwhelming number of responses, thus unintentionally intimidating them. It is certainly not intended to stop side conversations. That said, discussion of Hansen's 1988 predictions are probably off topic here, and should probably be carried across to Dana's excellent post on the topic. My point was not specifically about Hansen's predictions except that Evans completely misrepresents them, either through ignorance where he claimed knowledge, or through willfull deceit; and that therefore reference to his claims has no place in any intelligent discussion of this topic. That said, none of the three scenarios actually occurred, although what actually occurred more closely approximates to B than either other scenario. As Dana says:
    Total Scenario B greenhouse gas radiative forcing from 1984 to 2010 = 1.1 W/m2 The actual greenhouse gas forcing from 1984 to 2010 was approximately 1.06 W/m2 (NASA GISS). Thus the greenhouse gas radiative forcing in Scenario B was too high by about 5%."
    Note that Dana uses a higher value for CO2 concentration than I do, presumably because he got his data from a different source. And yes, there are issues with Evans relative placement of predictions and temperatures (to compare trends, the trend of the different projections should be centered on the trend line of the data at the initial point of the graph to avoid misleading visual cues) and his choice of HadCRUT3 data rather than the more accurate GISTemp or NCDC temperature records.
  33. Re: Clyde's mention of Pielke Sr. and his "challenge": I have no interest in going to Pielke's web site to find out what sort of "challenge" he has issued. He does not allow comments at his blog (last I visited), and I have no interest in trying to "engage" in a one-sided conversation completely in his control. If you wish to place a comment here describing what you understand the challenge to be, then I would be willing to discuss it with you. Pielke Sr. has participated in discussions here at SkS (in some of the blog posts I linked to above), and at Real Climate, and I have debated with him during those discussions. He is free to return here where we can debate on even terms. Since you don't have the time to read the many blog posts I referred you to, I will only suggest that you read this comment of mine on one of those threads, which may explain why I have no respect for Pielke Sr. as a scientist.
  34. Re: Clyde's definition of "computer modeller" and pointers to blogs that purportedly show model weaknesses. You have utterly failed to provide a useful definition of "computer modeller". It is the equivalent of telling me that a "frobnitz gleabinator" is someone who can "gleabinate a frobnitz". As for your link to JoNova's site: I see no point in going to a blog written by someone with no basic understanding of climatology. Others have already posted critiques of that information, and I see no need to add to them now. As I mentioned in my earlier request, please provide links to real scientific literature (Pielke Sr. also doesn't count in this area of expertise), or at least web sites where real scientific information is presented. When you post such links, please provide at least a short description (in your own words) of just what it is I should expect to find there. Now, to continue this discussion, can you please provide me with answers to the following questions that I have already posed to you: 1) What is your definition of a "computer modeler"? 2) On what basis do you claim that any particular "climatology expert" is not knowledgeable about computer modeling, and how would this affect the work that they are doing? 3)What else would I need in my background to convince you that I know enough about "computer modeling"? (My background was presented in this comment.)
  35. Tom, that first graph you show is even more egregious for another reason - the projections are shown by Evans as being initialised from a single high point in the noise of the temperature record in 1988. In reality, Hansen's model runs begin before 1960, and the individual runs are already diverging by 1988, depending on the different settings [e.g. scenario A has no volcanic forcing after 1988] of the model: The scenarios A, B and C are spread over ~0.2C around 1988, and A does not cross B or C after this point (contrast Evans' A and B separation with Hansen's A and B). By doing this, Evans greatly exaggerates discrepancy between modelled and observed temperatures, a discrepancy not actually present (see in detail here at RealClimate, 3rd fig). Below is my estimate of the positions in 1988 of Hansen's Scenarios A, B and C (shaded grey circles), and Evans' start point for all three (blue square), based on GISS (used by Hansen) and UAH (used by Evans) data, with the temperature plots offset so they overlap in WoodForTrees plotting package: The reality is that much of the visual discrepancy in Evans' chart is a consequence of his misrepresenting the positions of the model runs w.r.t. to 1988 temperature. He thus shifts all three model runs much too high, compared to the temperature. Readers are left to ask the question why Evans chose to start all three model run plots from the same spot, a positions higher than any of the model runs as they were actually presented in Hansen's 1988 paper. They can then ask the question why Clyde thinks this is a good example of models not reproducing reality...
  36. Bob Loblaw 533 I read the comment you requested & understand your feelings about Pielke Sr. To the best of my knowledge his reason for turning of the comments on his blog was to avoid dealing with name calling/childish behavior. So far i haven't had that problem with you nor anybody else. This not the only time he has issued the challenge & not the exact page i was looking for. It does have the info needed if anybody wants to refute his claim. I would think if the models are as good as some say this should be an easy task. Read more here. This label, of course, can be avoided if the researchers provide quantitiative model and observational comparisons of multi-decadal regional and local predictions of changes in climate statistics, and show them to be skillful in terms of what metrics are needed by the impacts community. I invite anyone who has published such a study to present a guest post on this weblog alerting us to such a robust scientific study.
  37. Bob Loblaw 534 First let me say i was only curious as to which was more complicated, GCM or HTML coding. I have no experience in GCM (big surprise i know -_^) coding. In the very little HTML coding I've been involved with its a pain in the butt. 1) What is your definition of a "computer modeler"? Somebody who can write the code & has the computer to run the code. 2) On what basis do you claim that any particular "climatology expert" is not knowledgeable about computer modeling, and how would this affect the work that they are doing? I'm going by say a doctor. A heart surgeon can operate on a heart, but that doesn't mean they can write the code & run it on a computer. If my comment left the impression i don't think any scientist has the ability to do both, that wasn't what i meant. 3)What else would I need in my background to convince you that I know enough about "computer modeling"? (My background was presented in this comment.) I don't recall saying you don't know enough about "computer modeling." I only "know" you from the brief interaction we've had here. The reason i asked you the question about the different coding was because you said you have written scientific code before.
  38. I hope this is not off topic for this thread. The reason i think its not is the paper gives more evidence of past warming be equal to or greater than today's. Making climate models in my view not reliable enough to pass new laws & regulations. You only get a small part & have to pay to read the full paper. No i didn't pay to read the full paper. Another paper with evidence of past warming being equal to if not more than today's. My apologies again for this mistyped hyperlink.
    Response: [DB] Your link is indeed off-topic for this thread. Future off-topic comments will receive moderation.
  39. Clyde - that natural fires occur is not evidence against arson, but it would be better to direct comments to Climate has changed before. Computer models do "predict" past warming - its just that the forcing are different.
  40. Clyde #537 - I see you have attempted to answer Bob and I's questions (both are similar). However, your answers consciously avoid any statement of why the expert in a field (heart surgery or climate) cannot become an expert modeller of a process in that field. Why is it that somebody, who has attained skill in understanding the processes of how something works, is precluded from encapsulating that knowledge in computer code? What I want to know is this: What is unique about a "modeller", that means neither a climatologist or a heart specialist can ever become one? How, in your opinion, do you become a "modeller"? Exactly what are the unique skills a modeller has? You see, fundamentally, what a "modeller" is, in this context, is someone who has the ability to generate computer code that results in an approximate representation of one or more processes in the climate system. They will have the ability to test that code, and to validate that code against expected results using synthetic data, as well as against real-world data. They can then correct their code or adjust the uncertainties accordingly. They will be able to estimate the uncertainties in their results and evaluate the strengths and weaknesses of their model. This is a technical skill, but one that is eminently achievable by physical/environmental scientists. By doing so, they become specialist climate modellers. You still have provided not one shred of justification as to why such scientists cannot do this. I don't actually believe you are willing to answer these questions adequately. Your subsequent casual comment concerning a Greenland climate paper equally shows you have little understanding of climatology, palaeoclimate, forcings, and regional versus global variations, to add to your evident failure to substantiate your original disparaging claims about climate modellers. Did you think warming/cooling was globally monotonic?
    Response: [DB] Fixed link.
  41. skywatcher 540 They will have the ability to test that code, and to validate that code against expected results using synthetic data, as well as against real-world data. They can then correct their code or adjust the uncertainties accordingly. That's part of my reason for not trusting models. Correct their code or adjust uncertainties. If laws & regs are passed based on current models that will need adjustments & corrections then why pass said laws & regs? Most of the adjustments & corrections I've read about are always to make the temp higher. (-Snip-) I've answered your other questions in my 537 post. You feel their not "adequate." I noticed some jumped all over JoNova & nobody has refuted the papers in my 520 comment. (-Snip-)

    [DB] Imputations of impropriety snipped.

    Off-topic snipped.

    Please construct comments in better compliance with the Comments Policy.

  42. The link to "Only In It For The Gold is old." It takes you to a page that redirects you to the site below.
  43. Clyde @ 541... "Most of the adjustments & corrections I've read about are always to make the temp higher." Um, I would suggest that's clearly not the case. In general, climate sensitivity estimates have come down slightly. Back in the 80's Hansen was estimating 4.2C for climate sensitivity (based on models and empirical research) and since then that's been adjusted down closer to 3C for 2XCO2. Even more recently research is showing that some of the very high estimations of CS are less likely thereby pushing the most likely CS down a smidge from that, to around 2.9-2.8C.
  44. Clyde - I pointed you to RC over your 520 "paper". The reason why action is needed, even with uncertainties, is because low end of uncertainties are bad and high end is very very bad (uncertainty cuts both ways). Heard of the precautionary principle? Its great that you are interested enough in truth to come here rather than just haunting disinformation sites, but it appears you have some predetermined opinions which are really seriously uninformed. Please take time to look for the real answers (backed by published science) rather than just assuming things (like climate scientists arent competent modellers, that models cant explain past climate change etc). Take a good look over the skeptical argument list - top left button).
  45. Clyde @537: 1) You are still just providing a circular definition of "computer modeller". If you don't know what a circular definition is, look it up in the dictionary under "definition, circular". Or admit that you don't have a definition. 2) I'm not interested in analogies with heart surgeons or doctors. I want you to identify an actual, real "climate expert" that you know of, and explain why that person is not "knowledgeable about computer modeling, and how would this affect the work that they are doing" (to quote my original question). In other words, what is it you think that they are doing that is weakened by your belief that they have insufficient knowledge of "computer modelling"? Or admit that you don't actually have any specifics that you can use to back up your claim. 3) You said that I "don't know enough about computer modeling" in this comment here, where you said "Why is it that folks who critique AGW are dismissed if their not experts in climate science, but we should just accept a climate scientist's work on models when their not experts in computer modeling?" You've cast a pretty wide net with that general claim, and as the old saying goes "I resemble that remark". - I have studied climatology through a Physical Geography program (B.Sc. and Ph.D.). - I have taught climatology in a major Canadian research universty (in a Geography department) - I have published journal papers on my research in reputable scientific journals - my research included writing/coding and using "climate models" I think this is sufficient to be called a "climate expert" - I took one first year "computer science" course in the 1970s. - I stopped taking mathematics course after first year calculus and algebra. I think that makes me someone that you might think of as "not an expert in computer modelling" Yet, somehow I still wrote computer models of climate. Please, tell me what it is you think I need in my background to convince you that I actually knew what I was doing? Surely, with my weak "computer training", I must be an easy target for you to criticize. If you can't argue that I fit your broad, sweeping generalization, then who does? (Which takes us back to point 2.) Back up your claim, instead of just avoiding it. Or admit that you're wrong.
  46. Clyde: I'm not interested in going to Pielke's web site. Please provide a short description of what you think his "challenge" is, and I will discuss it with you here. ("Here" being subject to the assumption that it is relevant to this particular topic, which is the reliability of climate models. If it isn't, please pick another thread and point me to it.)
  47. Clyde @ 541: You say "That's part of my reason for not trusting models. Correct their code or adjust uncertainties." Are you really telling me that if I write a model, and I find that there are difference between it and measurements, and I either - figure out what my model is doing incorrectly, and make it better ("correct the code") - decide that this means that the uncertainties in my model are greater than I thought they were when I had the more limited (and less different) measurements to compare it to ("adjust uncertainties") ..that you would decide that I am a bad scientist and not to be trusted? What actions or characteristics would make you trust a scientist faced with data that differs from a model?
  48. Clyde #541: And with that, you show unequivocally that you really don't have an understanding of what a modeller does, and how a modeller goes about their work. In your #537, you exactly did not answer the specific questions, as you stated that a modeller is someone who can "write the code". Climate modellers around the world can "write the code"! That part is easy! The hard part is validating the code. But you have, as yet, given absolutely no explanation as to why all these people who can "write the code" cannot write and validate a good climate model. You additionally, as Bob says, give no explanation as to why checking/changing a model, having found a discrepancy with real-world data, is anything other than good science. I wonder if you can furnish us with a specific example of the occasions where adjustments "make the temp higher", because to me it sounds like you are confusing temperature reconstructions with climate models. You also are, by this statement indirectly attributing deliberate motivations to the approaches of scientists. Do you actually believe anybody wants temperature to be higher? In other respects, I concur entirely with what Bob says
  49. I told you you would be wasting your time. Clyde isn't really answering your questions and he isn't allowing himself to be pinned down. When you do so he just switches to another argument (which is then deleted). He's not here to learn.
    Response: TC: Indeed. If Clyde does not very shortly answer some of the questions directed at him with answers that would actually substantiate his initial claims, or else acknowledge those initial claims to have been in error, or misinformed, this discussion will be in danger of violating the "no excessive repetition clause" of the comments policy.
  50. Perhaps some personal experience may be illuminating. Lest this be perceived as "dogpiling", I'm happy to respond in the context of skywatcher's claim "It is much easier to begin with an understanding of climate and physics, and graduate onto writing computer code, which is fundamentally not that difficult to do, than the alternative." I am part of a team of three people that writes scientific software (not climate-related). It involves modelling, calibration, error estimation, and 3D graphics, and the consequences of mistakes can be extremely serious. Two of us -- myself included -- are computer science graduates. The third has a PhD in the field that the software is actually used in. My CS degree was very heavy in mathematics (an option I took because I love maths) and the software is very maths-intense, which is obviously an advantage. I understand how the software works, and can explain it to others. However, the scientific innovations in the software usually come from the guy with the PhD in the field. I normally take his working implementation and optimise the hell out of it, as well as do all the 3D graphics stuff, etc., but he usually comes up with the core algorithms. He had no formal computer science training, and learnt most of his coding "on the job". His implementation is still often far from perfect, especially performance-wise, and he isn't aware of a pretty large body-of-knowledge about how to implement things well, but it still works. If he didn't have us, I believe he could still have produced working software that would have done the job, although it would have been orders of magnitude slower, less "fancy" from a user's point of view, and probably much harder to maintain and difficult to understand. It certainly wouldn't have less trustworthy just because it wasn't written by somebody with a CS degree. OTOH, if we didn't have him, we could still have written some software (I know, because we had more primitive software 15 years ago when he joined) but it wouldn't have been as sophisticated and it certainly would have taken us a lot longer to think up the algorithms that he has developed over the years. Writing computer code can either be extremely easy or the most difficult thing a human being can attempt to do. It depends on the nature and complexity of the code. Scientific code is generally not that complex from a computer-science point of view -- the important parts are simply direct transcriptions of mathematical expressions -- and so speaking as a computer scientist, it doesn't bother me in the least if no computer scientists are involved in the writing of GCMs. What they are possibly missing out on is optimised, multi-threaded implementations with wizz-bang 3D GUIs and easy-to-maintain code, but that doesn't change the correctness or reliability of the models. I also have no problem categorising people with a few decades of experience of writing code without CS degrees as "computer modellers". My PhD supervisor, like virtually all CS academics of his generation, had degrees in other disciplines (physics, in his case). It would be pretty absurd to classify me as a computer modeller but not the people who taught me!

Prev  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  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 2022 John Cook
Home | Links | Translations | About Us | Privacy | Contact Us