The different colours show the different studies reviewed. These used different and independent methods and Skeptical Science concludes they “provide multiple lines of evidence that humans are the dominant cause of global warming over the past century, and especially over the past 50 to 65 years.”
These studies considered different human and natural causes:
Human greenhouse gas (GHG) emissions,
Solar activity,
Volcanic activity,
Human aerosol emissions,
The El Niño Southern Oscillation (ENSO).
GHG and human aerosol (primarily sulphur dioxide, SO2) are the two largest human influences, and solar and volcanic activity and ENSO are the dominant natural influences on global temperature. The figure below illustrates the breakdown of their contributions in these studies.
Clearly, human inputs have provided the largest effect on global temperature over recent years. Greenhouse gas emission have had a positive effect and aerosol emissions a negative, partly balancing, effect. This negative human factor is decreasing as humanity cleans up its act in terms of SO2 and other aerosol pollution.
“A wide variety of statistical and physical approaches all arrived at the same conclusion: that humans are the dominant cause of the global warming over the past century, and particularly over the past 50 years.
I had a glance at these studies. To summarise:
Tett et al. 2000 compare temp with the output of HadCM3 model runs. They demonstrate that the model output fits better to the observations if you include GHG, and the models show a significant contribution from GHG.
Meehl et al. 2004 shows that in an ensemble of GCMs the effect of the forcing can be regarded as additive.
Stone et al. 2007 provides a method for estimating the impacts of separate forcings from the output of GCMs run only using a combined set of forcings.
Lean and Rind 2008 fit a linear model to of various forcings to the observations and show significant trends. They don’t test that the assumptions required for fitting a linear model apply (although they do observe that their error margins do not include uncertainties in either the temperature observations or the forcings to the time series), and consequently they overstate the accuracy of their estimates.
Huber and Knutti 2011 use an abstraction of an ensemble of GCMs constrained to surface temperatures. The link between the forcings and energy budget as determined from the model are reported.
Gillett et al. 2012 adjusts the output of an ensemble of runs from a GCM to fit temps and attribution of that to forcings, and then uses that to forecast future temps.
In then main these are papers about GCMs. Only Lean and Rind 2008 model observations directly, but fail to test the assumptions of their model.
If you believe that GCMs adequately model our current climate then you might draw the inferences, however the uncertainty in GCMs is understated (and even here I note that the caveats in the original papers go unreported).
I would rather hope that this isn’t the robust scientific evidence that humans are the dominant cause of global warming.
I wouldn’t wipe my arse with your “summary”.
I don’t get my science from some nobody on the internet.
If you have something to say about the science then stop wasting your time commenting on blogs and do some work.
Publish your findings.
Go ahead and challange the scientific consensus. Claim that Nobel Prize.
Everything else is just idle waffle.
Ken, I think you are asking me two separate questions here. The first is how good do I think the individual papers are, and second what do I think of Skeptical Science’s conclusions and the evidence it has gather together to support them.
On the papers they in the main involve interesting use of GCMs and related models to run scenarios of the relationships between forcings and (in the main) temp. In doing this the models are used in interesting ways, but they all fail to deal with the uncertainties robustly and caveats about the scenario nature of the findings (that one would expect from better science) get lost in claims about the real world.
Skeptical Science has pulled this rather eclectic group of papers together with a political end in mind – simply to remind us that “This robust scientific evidence is why there is a consensus amongst scientific experts that humans are the dominant cause of global warming.”
Why Skeptical Science choose papers that basically rely on GCMs for their conclusions and then implicitly suggest this represents a “wide variety of statistical and physical approaches” does rather suggest someone carried away by their own rhetoric. This is a narrow range of papers if one is looking at attribution, and, as I’ve said, they are pretty firmly rooted in modelling rather than physics or stats.
But perhaps it doesn’t sound as good to be honest and say ‘a wide variety of scenarios run using GCMs all arrived at the same conclusion: GHGs are modelled as the dominant cause of the global warming over the past century, although these results are subject to significant uncertainty …”.
The trouble with this stuff is that it is so value laden that being honest about the science that was actually done and its quality is hard.
Perhaps it is easier to see these issues looking at it all from a different domain. If a number of economists ran computable general equilibrium models of the NZ economy (a simpler model than a GCM) fed in various scenarios about taxation, and then (assuming they could construct a model that showed this) claimed “A wide variety of statistical and physical approaches all arrived at the same conclusion: that high taxes are the dominant cause of child poverty over the past century”, what would you think?
I hope you wouldn’t accept this. I hopefully imagine hearing you say that the model only showed what it output, it was only as good as its assumptions, and that model output isn’t empirical evidence. I also trust that if you were peer reviewing the paper you would I trust put a line through the speculation about causality.
In doing this the models are used in interesting ways, but they all fail to deal with the uncertainties robustly and caveats about the scenario nature of the findings (that one would expect from better science) get lost in claims about the real world.
Talk is cheap. Your hand waving is worthless.
Skeptical Science has pulled this rather eclectic group of papers together with a political end in mind…
Says who?
You?
(yawn)
This is a narrow range of papers if one is looking at attribution, and, as I’ve said, they are pretty firmly rooted in modelling rather than physics or stats.
Your eggs are addled.
Find out about how science works in the real world before you embarrass yourself on the internet.
Global Climate Models have successfully predicted:
* That the globe would warm, and about how fast, and about how much.
* That the troposphere would warm and the stratosphere would cool.
* That nighttime temperatures would increase more than daytime temperatures.
* That winter temperatures would increase more than summer temperatures
* Polar amplification (greater temperature increase as you move toward the poles).
* That the Arctic would warm faster than the Antarctic.
* The magnitude (0.3 K) and duration (two years) of the cooling from the Mt. Pinatubo eruption.
* They made a retrodiction for Last Glacial Maximum sea surface temperatures which was inconsistent with the paleo evidence, and better paleo evidence showed the models were right.
* They predicted a trend significantly different and differently signed from UAH satellite temperatures, and then a bug was found in the satellite data.
* The amount of water vapor feedback due to ENSO.
* The response of southern ocean winds to the ozone hole.
* The expansion of the Hadley cells.
* The poleward movement of storm tracks.
* The rising of the tropopause and the effective radiating altitude.
* The clear sky super greenhouse effect from increased water vapor in the tropics.
* The near constancy of relative humidity on global average.
* That coastal upwelling of ocean water would increase.
Seventeen correct predictions? Looks like a pretty good track record to me.
Are there problems with the models, and areas where they haven’t gotten it right yet? Sure there are. The double Inter-Tropical Convergence Zone which shows up in some coupled models, ENSO variability, insufficiently sensitive sea ice, diurnal cycles of moist convection, and the exact response of climate to clouds are all areas of ongoing research. But the models are still the best thing we have for climate prediction under different scenarios, and there is no reason at all to think they’re getting the overall picture wrong. http://bartonpaullevenson.com/ModelsReliable.html
ropata, climate models are no doubt useful (and in the end that is the criteria to judge any model against). The point of this interchange is that Skeptical Science claims that a set of papers relying upon GCMs are the robust scientific evidence that humans are the dominant cause of global warming.
I’ve simply been making the point that GCMs by themselves have problems as proof, and that in fact they may not be fit for purpose given the much wider uncertainty in results from them than are typically reported in the literature.
A case in point can be seen from the web reference you quote from in your comment. Somewhat ironically I note that it quotes Lean and Rind 2008 as evidence for the ability of climate models to predict global warming. This is the paper quoted by Skeptical Science, and as I noted this is the only paper so quoted that doesn’t involve the use of GCMs.
The first irony is this paper says the GCMs (used by the other papers) may “lack – or incorrectly parameterize – fundamental processes by which surface temperatures respond to radiative forcings. Cloud responses, which affect the latitude response structure, are known to be uncertain in the models”.
The next irony is that the paper argues that “Empirical models that combine natural and anthropogenic influences (at appropriate lags) capture 76% of the variance in the CRU monthly global surface temperature record, suggesting that much of the variability arises from processes that can be identified and their impact on the global surface temperature quantified by direct linear association with the observations” (my emphasis). In fact the paper doesn’t report that they tested the time series in question to see if they conform to the assumption required to make this assertion.
The third irony is that the error ranges reported in Table 1 strongly suggest the error ranges from the linear fit do assume the series are better behaved than they are. If we knew the error ranges the inferences we might draw could be different.
My point remains, these modelling based studies don’t necessarily provide the robust evidence that is claimed.
Simon Arnold:…models only work insofar as we have good data. Why not examine the data itself, then, if you’re not satisified with the models themselves? (Though…it’s not good to keep data behind a paywall, especially on such contentious issues like climate change.)
…Skeptical Science claims that a set of papers relying upon GCMs are the robust scientific evidence that humans are the dominant cause of global warming….
No, that won’t do.
It’s not just a website.
There is a global scientific consensus covering all the Earth Sciences.
Climate change is real and yes, it’s anthropogenic.
I’ve simply been making the point (…) problems as proof(…)they may not be fit for purpose(…) much wider uncertainty(…) “lack – or incorrectly parameterize – fundamental processes(…)are known to be uncertain(…) doesn’t report that they tested the time series in question(…)strongly suggest the error ranges from the linear fit do assume the series are better behaved than they are(..)
Your “point” is to hem and haw and to insert as much doubt and uncertainly as the English language will allow you. You never get around to saying anything definitive. It’s all about drawing pudgy, little question marks in the margins on other people’s work with a crayon. Find out about how science works before you offer a comment.
NASA is a good place to start.
Models don’t need to be exact in every respect to give us an accurate overall trend and its major effects – and we have that now. If you knew there were a 90% chance you’d be in a car crash, you wouldn’t get in the car (or at the very least, you’d wear a seatbelt). The IPCC concludes, with a greater than 90% probability, that humans are causing global warming. To wait for 100% certainty before acting is recklessly irresponsible. http://www.skepticalscience.com/climate-models-intermediate.htm http://www.skepticalscience.com/its-not-us-basic.htm
“Are you able to accept the conclusions that Skeptic Science made?
Or do you seek to dismiss or undermine them.”
Unfortunately you don’t seem to want to provide clear answers. But in your fluffing around you do seem to want to raise doubts. I suspect that in fact “manufacturing doubt” is your motive. Which answers the second question and explains you unwillingness to be clear about the first.
Later you say: “The point of this interchange is that Skeptical Science claims that a set of papers relying upon GCMs are the robust scientific evidence that humans are the dominant cause of global warming.”
Well, if you object to the word “robust” what about raising that at Skeptical Science? (I note you have not participated in that discussion there yet). It’s not a word I normally use, but I actually do not think it inappropriate in a popular presentation like this.
Your main point seems to be that models are not reality – they are inevitably imperfect reflections of reality. Who disagrees with that? In fact. I think the point is so obvious it does not require such elaboration (unless one is trying to raise straw men). That is the very nature of models. We are, after all attempting to model reality. And for a purpose. We never expect our predictions to always be completely and always accurate
Our models while inevitably imperfect will improve as we fill knowledge gaps. They are currently the best we have and we are entitled to use them appropriately. Especially as in this case we are not using individual models to predict a future situation.
Are you suggesting that we should never use our imperfect knowledge to make decisions? That we should wait until our knowledge corresponds completely with reality? if you are that is just a plan for inaction and putting your head in the sand. You will still be doing that when our planet is engulfed by the sun.
You suggest Sceptical science should have instead said: “a wide variety of scenarios run using GCMs all arrived at the same conclusion: GHGs are modelled as the dominant cause of the global warming over the past century, although these results are subject to significant uncertainty …”.
Well, I guess there is a difference in popular science writing and detailed reports or papers. But your formulation suggests you have entirely missed the whole point – and in fact it is quite erroneous. Models are not being used individually to determine the causes of global warming. They are being used in an attempt to reproduce the global temperatures that have been recorded.
Despite limitations and variabilities it turns out that the model predictions do largely correspond to observed global temperatures, especially when the results from a whole group of models are combined.
But here is the thing. While we are able to successfully model the recorded global temperature changes over a period – this is only possible if human inputs like greenhouse and aerosol emissions are included. If only natural inputs are included then the models just cannot predict the changes over the last 50 years.
Now, it seems to me that the IPCC conclusion (that human inputs are most likely to be the major cause of recent global warming) is pretty robust. It accommodates the fact that there is room for some doubt (very small room), but it is based on multiple models. One would expect that if the results were not so robust we would be able to find a set of models capable of accurately predicting global temperatures without inclusion of human inputs.
Or, have you actually found such a set of models, Simon?
Cathy, models can fail with good data, but your point about looking at the data is valid. Lean and Rind 2008 quote their data sources in general terms but I couldn’t see any supplementary info describing their methodology or specific dat.
Ken, you raise a number of points. But first a bit of context.
You post with favourable comments a summary of a post made at another web site. You link your post to Sciblogs which is designed to create an hub for scientific analysis etc. It seems to me that you are therefore seeking to expose your views to some scientific scrutiny, particularity when papers in the literature are being quoted as evidence for the views you endorse.
The point of all this is that I’m responding to you, not Skeptical Science.
Now when I start to question how good those papers are, and how well they do support the conclusions you’ve drawn, your response has not been to engage in the substance of the issues I raise (I might well have it wrong – these papers might be the best climate science has to offer when it comes to attribution, perhaps they do show uncertainty robustly).
Instead you seem personally aggrieved that I might be raising doubt. This gets to the point where you slip into talking about “Our models ..” (Which is the reason I suggested to elucidate the issues we talk about something less charged, like economic modelling).
Anyway unfortunately there is doubt and uncertainty in climate science, particularly about attribution and sensitivity. The good scientists know this and are working to try to reduce it. And it is important that it isn’t allowed to be swept under the carpet both for the sake of the science (that’s its bread and butter) and for the politics (uncertainty makes decision making hard).
To expand on this latter point the problem with not being clear about uncertainty is that you end up doing the wrong thing. A simple example is if you are overly confident about climate sensitivity (I personally think IPCC AR4 is) then you are likely to over invest in mitigation, when the rational economic and political response might well be to invest in reducing that uncertainty.
Anyway that’s a long way of getting around to answer your question.
Skeptical Science’s conclusions don’t follow from the papers they quote because those papers don’t deal with uncertainty appropriately, including the uncertainty that arises from the use of GCMs as proxies for the real world.
And to repeat, not to pay attention to the uncertainty (and frankly to actively try and suppress any discussion of it) is poor science and poor politics.
Simon, SciBlogs is a collection of scientists who blog. Some of us (who blogged before SciBlogs was formed) have chosen to syndicate – I am not linking.
SciBlogs is not a ” hub for scientific analysis” – it is an aggregate of blogging scientists – not a collection of scientific papers.
No, Simon. I am not “personally aggrieved.” Not at all. Contrarians commonly use the tactic of “manufacturing doubt.” I am used to it. I am just recognising it. It happens. You read far too much into “our models.” I am just talking about scientific models in general. Surely that is clear. I have worked with models – but not climate models – not my field.
Of course there is doubt and uncertainty in climate science – no climate scientists are “sweeping anything under the carpet.” The question of uncertainties and doubts are often discussed and there are continuai9ng attempts by scientists to make this situation clear to the lay reader.
But to suggest somebody is sweeping things under the carpet or ignoring uncertainties is a tactic of “manufacturing doubt.” As is the suggestions that no-one invests in reducing uncertainty. What the hell do you think the money invested in climate research is used for?
Your intention is clear with this last statement “not to pay attention to the uncertainty (and frankly to actively try and suppress any discussion of it) is poor science and poor politics.”
Strawmannery.
Anyway – thanks to the link to Judith Curry’s blog post (again not a scientific paper). She is a common reference for contrarians – but I will have a look at her post as clearly you are basing you objection the IPCDC conclusions on it.
Simon, I think I have. But I did ask you a question which you have ignored:
Are you aware of any group of models, or individual models, which have been demonstrated to track global temperatures well without including human inputs?
Your answer to that is important to me as if you can produce such a group of models I would find your opposition to Skeptical Science’s use of “robust” more credible.
It is the fact that none of the models appear to be able to track recent global temperature changes without including human inputs that I find so interesting. I understand the question of variations and uncertainties but we are talking about an overall compliance here.
Unless I am mistaken – and I am sure if I am you will answer that question clearly.
Ken, and I shouldn’t probably have to tell you this given your scientific background, but every time you test a model against the null of being due to chance, you are testing it against the hypothesis that chance is a better model.
That is why being careful about the calculation of uncertainty is so important.
So to get you engaged in dealing with my objections to the papers in question, do you think that Lean and Rind 2008 do adequately demonstrate that their linear regression model meets the requirements to make the assertions they do about the parameters being significantly different from the chance model?
A quite acceptable answer under the circumstances would be “we can’t tell”.
Seems to me it’s key. If a group of models are able to track global temperatures over recent years without human inputs it really does suggest “robust” is an exaggeration.
If there aren’t any thatius very telling, isn’t it!
Strange, Simon, because this was the message in my article and I thought you were objecting to it.
No one, least of all me, overestimates the ability of individual models. Uncertanties and variability will always be with us. However we must work with what we have, while we also work to improve them.
And it seems to me very telling that the models appear to be able to track global temperatures (the proof of the pudding is in the eating), but don’t for the last 50 years if human inputs are not included.
If the models were no better than chance, which you seem to be suggesting, they would not track like that.
And your inability to identify models which track over the last 50 years without human inputs does suggest to me that the evidence for human effects is pretty robust.
This was the IPCC conclusion and Skeptical Science was updating that with recent studies.
Given you have nothing to say about the particular papers cited I take it we’re agreed that they aren’t the robust evidence humans are the dominant cause of global warming?
Other evidence would have been better to use perhaps?
Ken I came back to see how this had gone over night and it occurs to me that it’s worth one last go – if you’ll agree that resorting to impugning motives and ad hominem is for the children.
The problem is that the models quoted by you are not very good. They are probably no better than chance. They don’t demonstrate that models with GHG do better than other models. The model that Lean and Rind 2008 work with shows the trend coefficients on non-GHG parameters changing sign when modelling the periods with and without global warming.
You obviously feel there are other models that will do the job better, but they aren’t the ones you quoted.
If the models are no better than chance how come they can track the observed global temperatures (even if imperfectly)? In combination they seem very good. And why would we bother with such models if we could obtain the same results by flipping coins?
You have taken the known limitations of scientific models, their known gaps (eg clouds and aerosols), uncertainties, etc., and extrapolated to “no better than chance”. That is political – “manufacturing doubt.”
Real scientists deal with the real world. They use the tools at hand to the best they can, keeping in mind limitations, etc. That’s why we make use of real scientific models.
The deniers/contrarians etc want to hold us back. Demand that no interim conclusion be drawn or action taken until models are perfect (will never happen and they know it).
Governments understand the limitations of models and provisional nature of scientific knowledge. But they have to govern and probably appreciate that the models and knowledge provided by climate science are far better than they normally are used to dealing with.
I don’t think we will agree because you have expressed a position that nothing be done until uncertainties are removed. I don’t think that’s sensible and am sure you don’t do that in your day to day life. Nevertheless you seem to be trying hard to justify that approach.
Net human and natural percent contributions to the observed global surface warming over the past 50-65 years according to Tett et al. 2000 (T00, dark blue), Meehl et al. 2004 (M04, red), Stone et al. 2007 (S07, green), Lean and Rind 2008 (LR08, purple), Huber and Knutti 2011 (HK11, light blue), and Gillett et al. 2012 (G12, orange).
Percent contributions of various effects to the observed global surface warming over the past 100-150 years according to Tett et al. 2000 (T00, dark blue), Meehl et al. 2004 (M04, red), Stone et al. 2007 (S07, green), Lean and Rind 2008 (LR08, purple), Stott et al. 2010 (S10, gray), and Huber and Knutti 2011 (HR11, light blue).
Open Parachute 
Pingback: Our fingerprints are all over it! | Open Parachute | Secularity
“A wide variety of statistical and physical approaches all arrived at the same conclusion: that humans are the dominant cause of the global warming over the past century, and particularly over the past 50 years.
I had a glance at these studies. To summarise:
Tett et al. 2000 compare temp with the output of HadCM3 model runs. They demonstrate that the model output fits better to the observations if you include GHG, and the models show a significant contribution from GHG.
Meehl et al. 2004 shows that in an ensemble of GCMs the effect of the forcing can be regarded as additive.
Stone et al. 2007 provides a method for estimating the impacts of separate forcings from the output of GCMs run only using a combined set of forcings.
Lean and Rind 2008 fit a linear model to of various forcings to the observations and show significant trends. They don’t test that the assumptions required for fitting a linear model apply (although they do observe that their error margins do not include uncertainties in either the temperature observations or the forcings to the time series), and consequently they overstate the accuracy of their estimates.
Huber and Knutti 2011 use an abstraction of an ensemble of GCMs constrained to surface temperatures. The link between the forcings and energy budget as determined from the model are reported.
Gillett et al. 2012 adjusts the output of an ensemble of runs from a GCM to fit temps and attribution of that to forcings, and then uses that to forecast future temps.
In then main these are papers about GCMs. Only Lean and Rind 2008 model observations directly, but fail to test the assumptions of their model.
If you believe that GCMs adequately model our current climate then you might draw the inferences, however the uncertainty in GCMs is understated (and even here I note that the caveats in the original papers go unreported).
I would rather hope that this isn’t the robust scientific evidence that humans are the dominant cause of global warming.
LikeLike
I guess then, Simon, you have managed to convince yourself to ignore the findings of these workers?
LikeLike
Ken, no I wouldn’t ignore them, but equally I try and understand their limitations – don’t you?
LikeLike
Of course Simon. And the positives as well. All good science is like that.
But you seem to be trying to find any excuse to dismiss the work.
If not – good. Are you able to accept the conclusions that Skeptic Science made? Or do you seek to dismiss or undermine them.
LikeLike
I had a glance at these studies. To summarise:
I wouldn’t wipe my arse with your “summary”.
I don’t get my science from some nobody on the internet.
If you have something to say about the science then stop wasting your time commenting on blogs and do some work.
Publish your findings.
Go ahead and challange the scientific consensus. Claim that Nobel Prize.
Everything else is just idle waffle.
LikeLike
Ken, I think you are asking me two separate questions here. The first is how good do I think the individual papers are, and second what do I think of Skeptical Science’s conclusions and the evidence it has gather together to support them.
On the papers they in the main involve interesting use of GCMs and related models to run scenarios of the relationships between forcings and (in the main) temp. In doing this the models are used in interesting ways, but they all fail to deal with the uncertainties robustly and caveats about the scenario nature of the findings (that one would expect from better science) get lost in claims about the real world.
Skeptical Science has pulled this rather eclectic group of papers together with a political end in mind – simply to remind us that “This robust scientific evidence is why there is a consensus amongst scientific experts that humans are the dominant cause of global warming.”
Why Skeptical Science choose papers that basically rely on GCMs for their conclusions and then implicitly suggest this represents a “wide variety of statistical and physical approaches” does rather suggest someone carried away by their own rhetoric. This is a narrow range of papers if one is looking at attribution, and, as I’ve said, they are pretty firmly rooted in modelling rather than physics or stats.
But perhaps it doesn’t sound as good to be honest and say ‘a wide variety of scenarios run using GCMs all arrived at the same conclusion: GHGs are modelled as the dominant cause of the global warming over the past century, although these results are subject to significant uncertainty …”.
The trouble with this stuff is that it is so value laden that being honest about the science that was actually done and its quality is hard.
Perhaps it is easier to see these issues looking at it all from a different domain. If a number of economists ran computable general equilibrium models of the NZ economy (a simpler model than a GCM) fed in various scenarios about taxation, and then (assuming they could construct a model that showed this) claimed “A wide variety of statistical and physical approaches all arrived at the same conclusion: that high taxes are the dominant cause of child poverty over the past century”, what would you think?
I hope you wouldn’t accept this. I hopefully imagine hearing you say that the model only showed what it output, it was only as good as its assumptions, and that model output isn’t empirical evidence. I also trust that if you were peer reviewing the paper you would I trust put a line through the speculation about causality.
LikeLike
In doing this the models are used in interesting ways, but they all fail to deal with the uncertainties robustly and caveats about the scenario nature of the findings (that one would expect from better science) get lost in claims about the real world.
Talk is cheap. Your hand waving is worthless.
Skeptical Science has pulled this rather eclectic group of papers together with a political end in mind…
Says who?
You?
(yawn)
This is a narrow range of papers if one is looking at attribution, and, as I’ve said, they are pretty firmly rooted in modelling rather than physics or stats.
Your eggs are addled.
Find out about how science works in the real world before you embarrass yourself on the internet.
Global Warming: It’s Not About the Hockey Stick
LikeLike
Global Climate Models have successfully predicted:
* That the globe would warm, and about how fast, and about how much.
* That the troposphere would warm and the stratosphere would cool.
* That nighttime temperatures would increase more than daytime temperatures.
* That winter temperatures would increase more than summer temperatures
* Polar amplification (greater temperature increase as you move toward the poles).
* That the Arctic would warm faster than the Antarctic.
* The magnitude (0.3 K) and duration (two years) of the cooling from the Mt. Pinatubo eruption.
* They made a retrodiction for Last Glacial Maximum sea surface temperatures which was inconsistent with the paleo evidence, and better paleo evidence showed the models were right.
* They predicted a trend significantly different and differently signed from UAH satellite temperatures, and then a bug was found in the satellite data.
* The amount of water vapor feedback due to ENSO.
* The response of southern ocean winds to the ozone hole.
* The expansion of the Hadley cells.
* The poleward movement of storm tracks.
* The rising of the tropopause and the effective radiating altitude.
* The clear sky super greenhouse effect from increased water vapor in the tropics.
* The near constancy of relative humidity on global average.
* That coastal upwelling of ocean water would increase.
Seventeen correct predictions? Looks like a pretty good track record to me.
Are there problems with the models, and areas where they haven’t gotten it right yet? Sure there are. The double Inter-Tropical Convergence Zone which shows up in some coupled models, ENSO variability, insufficiently sensitive sea ice, diurnal cycles of moist convection, and the exact response of climate to clouds are all areas of ongoing research. But the models are still the best thing we have for climate prediction under different scenarios, and there is no reason at all to think they’re getting the overall picture wrong.
http://bartonpaullevenson.com/ModelsReliable.html
LikeLike
ropata, climate models are no doubt useful (and in the end that is the criteria to judge any model against). The point of this interchange is that Skeptical Science claims that a set of papers relying upon GCMs are the robust scientific evidence that humans are the dominant cause of global warming.
I’ve simply been making the point that GCMs by themselves have problems as proof, and that in fact they may not be fit for purpose given the much wider uncertainty in results from them than are typically reported in the literature.
A case in point can be seen from the web reference you quote from in your comment. Somewhat ironically I note that it quotes Lean and Rind 2008 as evidence for the ability of climate models to predict global warming. This is the paper quoted by Skeptical Science, and as I noted this is the only paper so quoted that doesn’t involve the use of GCMs.
The first irony is this paper says the GCMs (used by the other papers) may “lack – or incorrectly parameterize – fundamental processes by which surface temperatures respond to radiative forcings. Cloud responses, which affect the latitude response structure, are known to be uncertain in the models”.
The next irony is that the paper argues that “Empirical models that combine natural and anthropogenic influences (at appropriate lags) capture 76% of the variance in the CRU monthly global surface temperature record, suggesting that much of the variability arises from processes that can be identified and their impact on the global surface temperature quantified by direct linear association with the observations” (my emphasis). In fact the paper doesn’t report that they tested the time series in question to see if they conform to the assumption required to make this assertion.
The third irony is that the error ranges reported in Table 1 strongly suggest the error ranges from the linear fit do assume the series are better behaved than they are. If we knew the error ranges the inferences we might draw could be different.
My point remains, these modelling based studies don’t necessarily provide the robust evidence that is claimed.
LikeLike
Simon Arnold:…models only work insofar as we have good data. Why not examine the data itself, then, if you’re not satisified with the models themselves? (Though…it’s not good to keep data behind a paywall, especially on such contentious issues like climate change.)
LikeLike
…Skeptical Science claims that a set of papers relying upon GCMs are the robust scientific evidence that humans are the dominant cause of global warming….
No, that won’t do.
It’s not just a website.
There is a global scientific consensus covering all the Earth Sciences.
Climate change is real and yes, it’s anthropogenic.
I’ve simply been making the point (…) problems as proof(…)they may not be fit for purpose(…) much wider uncertainty(…) “lack – or incorrectly parameterize – fundamental processes(…)are known to be uncertain(…) doesn’t report that they tested the time series in question(…)strongly suggest the error ranges from the linear fit do assume the series are better behaved than they are(..)
Your “point” is to hem and haw and to insert as much doubt and uncertainly as the English language will allow you. You never get around to saying anything definitive. It’s all about drawing pudgy, little question marks in the margins on other people’s work with a crayon. Find out about how science works before you offer a comment.
NASA is a good place to start.
NASA: Climate Change; A Warming World (HD)
LikeLike
Models don’t need to be exact in every respect to give us an accurate overall trend and its major effects – and we have that now. If you knew there were a 90% chance you’d be in a car crash, you wouldn’t get in the car (or at the very least, you’d wear a seatbelt). The IPCC concludes, with a greater than 90% probability, that humans are causing global warming. To wait for 100% certainty before acting is recklessly irresponsible.
http://www.skepticalscience.com/climate-models-intermediate.htm
http://www.skepticalscience.com/its-not-us-basic.htm
LikeLike
Actually, Simon, my questions were:
“Are you able to accept the conclusions that Skeptic Science made?
Or do you seek to dismiss or undermine them.”
Unfortunately you don’t seem to want to provide clear answers. But in your fluffing around you do seem to want to raise doubts. I suspect that in fact “manufacturing doubt” is your motive. Which answers the second question and explains you unwillingness to be clear about the first.
Later you say: “The point of this interchange is that Skeptical Science claims that a set of papers relying upon GCMs are the robust scientific evidence that humans are the dominant cause of global warming.”
Well, if you object to the word “robust” what about raising that at Skeptical Science? (I note you have not participated in that discussion there yet). It’s not a word I normally use, but I actually do not think it inappropriate in a popular presentation like this.
Your main point seems to be that models are not reality – they are inevitably imperfect reflections of reality. Who disagrees with that? In fact. I think the point is so obvious it does not require such elaboration (unless one is trying to raise straw men). That is the very nature of models. We are, after all attempting to model reality. And for a purpose. We never expect our predictions to always be completely and always accurate
Our models while inevitably imperfect will improve as we fill knowledge gaps. They are currently the best we have and we are entitled to use them appropriately. Especially as in this case we are not using individual models to predict a future situation.
Are you suggesting that we should never use our imperfect knowledge to make decisions? That we should wait until our knowledge corresponds completely with reality? if you are that is just a plan for inaction and putting your head in the sand. You will still be doing that when our planet is engulfed by the sun.
You suggest Sceptical science should have instead said: “a wide variety of scenarios run using GCMs all arrived at the same conclusion: GHGs are modelled as the dominant cause of the global warming over the past century, although these results are subject to significant uncertainty …”.
Well, I guess there is a difference in popular science writing and detailed reports or papers. But your formulation suggests you have entirely missed the whole point – and in fact it is quite erroneous. Models are not being used individually to determine the causes of global warming. They are being used in an attempt to reproduce the global temperatures that have been recorded.
Despite limitations and variabilities it turns out that the model predictions do largely correspond to observed global temperatures, especially when the results from a whole group of models are combined.
But here is the thing. While we are able to successfully model the recorded global temperature changes over a period – this is only possible if human inputs like greenhouse and aerosol emissions are included. If only natural inputs are included then the models just cannot predict the changes over the last 50 years.
Now, it seems to me that the IPCC conclusion (that human inputs are most likely to be the major cause of recent global warming) is pretty robust. It accommodates the fact that there is room for some doubt (very small room), but it is based on multiple models. One would expect that if the results were not so robust we would be able to find a set of models capable of accurately predicting global temperatures without inclusion of human inputs.
Or, have you actually found such a set of models, Simon?
LikeLike
Cathy, models can fail with good data, but your point about looking at the data is valid. Lean and Rind 2008 quote their data sources in general terms but I couldn’t see any supplementary info describing their methodology or specific dat.
ropata, unfortunately there is more uncertainty here than the last IPCC report suggests. http://judithcurry.com/2010/10/24/overconfidence-in-ipccs-detection-and-attribution-part-iii/ sets out some of the issues (just ignore the actual assessments in the post if you like, but read it to gain an impression of the issues involved).
Ken, you raise a number of points. But first a bit of context.
You post with favourable comments a summary of a post made at another web site. You link your post to Sciblogs which is designed to create an hub for scientific analysis etc. It seems to me that you are therefore seeking to expose your views to some scientific scrutiny, particularity when papers in the literature are being quoted as evidence for the views you endorse.
The point of all this is that I’m responding to you, not Skeptical Science.
Now when I start to question how good those papers are, and how well they do support the conclusions you’ve drawn, your response has not been to engage in the substance of the issues I raise (I might well have it wrong – these papers might be the best climate science has to offer when it comes to attribution, perhaps they do show uncertainty robustly).
Instead you seem personally aggrieved that I might be raising doubt. This gets to the point where you slip into talking about “Our models ..” (Which is the reason I suggested to elucidate the issues we talk about something less charged, like economic modelling).
Anyway unfortunately there is doubt and uncertainty in climate science, particularly about attribution and sensitivity. The good scientists know this and are working to try to reduce it. And it is important that it isn’t allowed to be swept under the carpet both for the sake of the science (that’s its bread and butter) and for the politics (uncertainty makes decision making hard).
To expand on this latter point the problem with not being clear about uncertainty is that you end up doing the wrong thing. A simple example is if you are overly confident about climate sensitivity (I personally think IPCC AR4 is) then you are likely to over invest in mitigation, when the rational economic and political response might well be to invest in reducing that uncertainty.
Anyway that’s a long way of getting around to answer your question.
Skeptical Science’s conclusions don’t follow from the papers they quote because those papers don’t deal with uncertainty appropriately, including the uncertainty that arises from the use of GCMs as proxies for the real world.
And to repeat, not to pay attention to the uncertainty (and frankly to actively try and suppress any discussion of it) is poor science and poor politics.
LikeLike
Simon, SciBlogs is a collection of scientists who blog. Some of us (who blogged before SciBlogs was formed) have chosen to syndicate – I am not linking.
SciBlogs is not a ” hub for scientific analysis” – it is an aggregate of blogging scientists – not a collection of scientific papers.
No, Simon. I am not “personally aggrieved.” Not at all. Contrarians commonly use the tactic of “manufacturing doubt.” I am used to it. I am just recognising it. It happens. You read far too much into “our models.” I am just talking about scientific models in general. Surely that is clear. I have worked with models – but not climate models – not my field.
Of course there is doubt and uncertainty in climate science – no climate scientists are “sweeping anything under the carpet.” The question of uncertainties and doubts are often discussed and there are continuai9ng attempts by scientists to make this situation clear to the lay reader.
But to suggest somebody is sweeping things under the carpet or ignoring uncertainties is a tactic of “manufacturing doubt.” As is the suggestions that no-one invests in reducing uncertainty. What the hell do you think the money invested in climate research is used for?
Your intention is clear with this last statement “not to pay attention to the uncertainty (and frankly to actively try and suppress any discussion of it) is poor science and poor politics.”
Strawmannery.
Anyway – thanks to the link to Judith Curry’s blog post (again not a scientific paper). She is a common reference for contrarians – but I will have a look at her post as clearly you are basing you objection the IPCDC conclusions on it.
LikeLike
Ken, you didn’t seem to deal with the substance of my objections to the papers in question.
LikeLike
Simon, I think I have. But I did ask you a question which you have ignored:
Are you aware of any group of models, or individual models, which have been demonstrated to track global temperatures well without including human inputs?
Your answer to that is important to me as if you can produce such a group of models I would find your opposition to Skeptical Science’s use of “robust” more credible.
It is the fact that none of the models appear to be able to track recent global temperature changes without including human inputs that I find so interesting. I understand the question of variations and uncertainties but we are talking about an overall compliance here.
Unless I am mistaken – and I am sure if I am you will answer that question clearly.
LikeLike
Ken, and I shouldn’t probably have to tell you this given your scientific background, but every time you test a model against the null of being due to chance, you are testing it against the hypothesis that chance is a better model.
That is why being careful about the calculation of uncertainty is so important.
So to get you engaged in dealing with my objections to the papers in question, do you think that Lean and Rind 2008 do adequately demonstrate that their linear regression model meets the requirements to make the assertions they do about the parameters being significantly different from the chance model?
A quite acceptable answer under the circumstances would be “we can’t tell”.
LikeLike
Simon, are you avoiding my question?
Seems to me it’s key. If a group of models are able to track global temperatures over recent years without human inputs it really does suggest “robust” is an exaggeration.
If there aren’t any thatius very telling, isn’t it!
LikeLike
Ken, you are fighting another battle not of my making.
Do you think the papers cited by you and Skeptical Science are the robust evidence humans are the dominant cause of global warming?
If you do let’s talk about it, if you don’t or can’t tell because of problems with the papers, we are agreed.
LikeLike
Strange, Simon, because this was the message in my article and I thought you were objecting to it.
No one, least of all me, overestimates the ability of individual models. Uncertanties and variability will always be with us. However we must work with what we have, while we also work to improve them.
And it seems to me very telling that the models appear to be able to track global temperatures (the proof of the pudding is in the eating), but don’t for the last 50 years if human inputs are not included.
If the models were no better than chance, which you seem to be suggesting, they would not track like that.
And your inability to identify models which track over the last 50 years without human inputs does suggest to me that the evidence for human effects is pretty robust.
This was the IPCC conclusion and Skeptical Science was updating that with recent studies.
LikeLike
Given you have nothing to say about the particular papers cited I take it we’re agreed that they aren’t the robust evidence humans are the dominant cause of global warming?
Other evidence would have been better to use perhaps?
LikeLike
Come off it, Simon. That’s childish.
You have not produced a single model capable of tracking recent global temperature using only natural inputs.
I think that tells us a lot about the evidence for human causes of recent global temperature increases. Seems pretty robust to me.
I appreciate that we don’t agree but I suspect your motives all along have been to manufacture doubt one way or the other.
No skin of my nose – you aren’t the only person who thinks that way.
LikeLike
You have not produced a single model capable of tracking recent global temperature using only natural inputs.
Should be easy enough.
Stop being a tool, Simon.
This Year’s Model
LikeLike
Ken I came back to see how this had gone over night and it occurs to me that it’s worth one last go – if you’ll agree that resorting to impugning motives and ad hominem is for the children.
The problem is that the models quoted by you are not very good. They are probably no better than chance. They don’t demonstrate that models with GHG do better than other models. The model that Lean and Rind 2008 work with shows the trend coefficients on non-GHG parameters changing sign when modelling the periods with and without global warming.
You obviously feel there are other models that will do the job better, but they aren’t the ones you quoted.
LikeLike
If the models are no better than chance how come they can track the observed global temperatures (even if imperfectly)? In combination they seem very good. And why would we bother with such models if we could obtain the same results by flipping coins?
You have taken the known limitations of scientific models, their known gaps (eg clouds and aerosols), uncertainties, etc., and extrapolated to “no better than chance”. That is political – “manufacturing doubt.”
Real scientists deal with the real world. They use the tools at hand to the best they can, keeping in mind limitations, etc. That’s why we make use of real scientific models.
The deniers/contrarians etc want to hold us back. Demand that no interim conclusion be drawn or action taken until models are perfect (will never happen and they know it).
Governments understand the limitations of models and provisional nature of scientific knowledge. But they have to govern and probably appreciate that the models and knowledge provided by climate science are far better than they normally are used to dealing with.
I don’t think we will agree because you have expressed a position that nothing be done until uncertainties are removed. I don’t think that’s sensible and am sure you don’t do that in your day to day life. Nevertheless you seem to be trying hard to justify that approach.
So I think we just have to agree to disagree.
LikeLike
The problem is that the models quoted by you are not very good.
Who says so?
You?
Oh.
They are probably no better than chance.
Who says so?
You?
Hmm.
(yawn)
LikeLike