Man Made Global Warming

[cybercoma]

Of course. We should be reducing emissions of all kinds. The key is, as you said, not to harm the economy. Weather GW is happening or not, whether or not we are causing it, we should reduce our impact on the planet. That said, there is no reason to panic. The best thing we can do is adapt to what is happening and move forward on all fronts.

This may seem like splitting hairs, but I did not say "not to harm the economy". I said, "without significantly affecting our economies". I feel some affect to the economy may be tolerable, but to what extent is up for debate.

My point, and you supported it to some extent, is that we know carbon dioxide is a greenhouse gas. That's not up for debate. It traps heat and has an impact on our environment. We know that we are pumping carbon dioxide into the atmosphere through our activities. So, the only logical thing to do is determine how we can significantly reduce or stop that without causing other problems that could be just as devastating as the environmental ones we're facing.

[bush_cheney2004]

My point, and you supported it to some extent, is that we know carbon dioxide is a greenhouse gas. That's not up for debate. It traps heat and has an impact on our environment. We know that we are pumping carbon dioxide into the atmosphere through our activities. So, the only logical thing to do is determine how we can significantly reduce or stop that without causing other problems that could be just as devastating as the environmental ones we're facing.

So is water vapor, and to a much larger degree...shall we begin efforts to limit water vapor "emissions" as well?

[Smallc]

This may seem like splitting hairs, but I did not say "not to harm the economy". I said, "without significantly affecting our economies". I feel some affect to the economy may be tolerable, but to what extent is up for debate.

I would agree that some extent is tolerable were we not currently in what appears to be a global slowdown. Now is not the time to shoot ourselves in the foot IMO.

[Riverwind]

It traps heat and has an impact on our environment.

Trapping heat is not necessarily a bad thing. Human civilization has generally flourished during warm spells in the past. CO2 is the one of the only "pollutants" which can be put into the environment and provide positive as well as negative effects. This makes the cost-benefit analysis for reducing CO2 production a lot more complicated.

So, the only logical thing to do is determine how we can significantly reduce or stop that without causing other problems that could be just as devastating as the environmental ones we're facing.

The only things that we should do are things which would make sense even if the risk from CO2 is minimal. For example, building nuclear plants instead of coal plant makes a lot of sense. As do tighter energy efficiency requirements. R&D and funding of large scale alternative energy projects makes sense too.

What does not make sense are arbitrary carbon pricing measures like cap and trade and and carbon taxes which do not take into account that CO2 emissions are unavoidable for many activities. The inability to discriminate between different types of activities means that carbon pricing measures will have either set a price that is too low to have any effect on emissions or accept that many kinds of economic activities will be effectively outlawed.

That is why the I think the conservative plan makes a lot of sense. It targets specific industries and sets realitistic goals for those industries. It provides for cap and trade, however, the proceeds must go back into the industry to fund initiatives to reduce CO2 output (i.e. the money is used to solve the problem instead of funding new social programs). These measures will increase costs for consumers but those cost increases will be manageable because the regulations take into account what is technically possible for each industry. Dion's carbon tax will be too low to change behavoir but will increase prices across the board.

Edited September 2, 2008 by Riverwind

[Smallc]

Trapping heat is not necessarily a bad thing. Human civilization has generally flourished during warm spells in the past. CO2 is the one of the only "pollutants" which can be put into the environment and provide positive as well as negative effects. This makes the cost-benefit analysis for reducing CO2 production a lot more complicated

Thats very true. If we are increasing CO2 levels and Earth is heating, maybe we're preventing an ice age.

[marksman]

I assume you don't have much experience with engineering or any other discipline that uses numerical models similar to climate models to make decisions that can have safety or monetary implications. People working in these fields immediately recognize that any data available while the model is being developed affects the choices of the model makers. This means it is impossible to determine whether a match between a model and the real data is the result of a correct theory or a result of tuning. The only way to demonstrate that the theory of the model is actually correct is to compare the model predictions to data that was not known to the model makers while they developed the model. Now this criteria could be met if a rigorous process was followed and some historical data was kept secret, however, there is no reason to believe that this kind of rigourous process was followed with the IPCC climate models. Therefore, we must use future data to determine whether the models have any predictive ability.

I am not saying the model makers are being deliberately deceptive. I am saying that the model makers can't help themselves. They have certain expectations and they have a lot of tunable parameters. If the model output does not match their expectations they will adjust the tuneable parameters. The model makers would only rethink their theory if it was impossible to get a match without pushing these tunable parameters outside of "realistic" values. However, in climate models the range of "realistic" values for things like aerosols is huge which means model makers can easily hide deficiencies in the model theory with 'scientifically justifiable" tuning.

I also realize that the model makers have little choice when it comes to modelling climate and they have likely done the best with what they have. However, they are wrong to claim that hindcasts give us any indication of how the models may perform in the future. The only reasonable claim they can make is that the hindcasts show that they have a reasonable but unproven hypotheses, however, there is a significant probability that they have messed something up and the model predictions will be flat out wrong.

Bottom line: sometimes admitting that we simply cannot answer certain questions is better than using unrealiable techniques that give the illusion of certainty when there is none.

Agreed - but how do we sort out the good from the bad. My concern is the hockey stick has demonstrated that the traditional climate science community and the IPCC cannot be trusted to do this sorting so who do we turn to?

Your assumption wouldn't be correct. But I think I must not have been clear about the fact that future predictions are the ultimate check on these models. Obviously the model can only be validated by making future predictions. That isn't where the problem is. My problem was with you calling the entire process flawed. There is a danger that people will consciously or unconsciously skew their model based on available data but that doesn't invalidate the process of tuning the model. What point is there in making future predictions if you've got no idea how well your model reflects reality? You're basically saying that models should be created in a vacuum and their deficiencies should only be exposed after years of gathering data. But you're more likely to get better predictions if you can demonstrate your model already has some degree of accuracy.

In this way "hindcasts" do give indications of how a model will perform in the future. If you can't explain the last 100 years there's no point in thinking you can explain the next 100. If you can explain the last 100 then this doesn't mean you're automatically correct about the next 100 but it's more likely that your model is closer to giving accurate predictions. You've said that "there is a significant probability that they have messed something up and the model predictions will be flat out wrong" but if you've got an unchecked model then there's an even higher probability that you've messed up and your predictions will be flat out wrong. By arguing against this method you're arguing for a lower probability of accuracy in future predictions.

To use a historical example look at Newton and Einstein. Einstein developed his gravity theory knowing that Newton's theory didn't explain Mercury's orbit. He developed a theory that explained Mercury's orbit and its been subsequently proved in other ways. Planetary orbits are obviously different than a planet's climate system but to say that using historical data to improve a model somehow invalidates the entire model is unrealistic and impractical.

You've also assumed that any tuning is done by using unrealistic values. There's no reason to think this is always the case. I'm sure some people do this but it is equally possible that historical data shows that some factors were overestimated or underestimated and realistic tuning is possible.

I don't like the term climate science community because it doesn't reflect what we're really talking about - those who believe in human caused climate change and those who don't. You're using the hockey stick debate to say the pro side can't be trusted. If this is the case then we can't trust either side because some on the anti side are guilty of even more explicit underhanded tactics:

http://www.ucsusa.org/news/press_release/E...ng-tobacco.html

The IPCC isn't some conspiracy organization out to prove human caused climate change at any cost. It's trying to sort through mountains of evidence and dealing with something on this scale it's inevitable that mistakes will be made. It's also inevitable that scientific conclusions will change and be refined. But to dismiss all of the good evidence they've looked at based on some criticism of one piece of evidence isn't scientific and certainly isn't balanced if you're willing to consider evidence from the other side when the other side has done equally questionable things.

[Riverwind]

You're basically saying that models should be created in a vacuum and their deficiencies should only be exposed after years of gathering data. But you're more likely to get better predictions if you can demonstrate your model already has some degree of accuracy.

You have misunderstood what I am saying. I am not saying that hindcasts should not be done. I am saying that hindcasts tell us nothing about the predictive skill of the model nor do they tell us that the theory built into the model is correct. The only way to show that the theory built into the model is correct is to use the model to predict outcomes that were not known when the model was developed. That does not mean unvalidated models are useless. In fact, poor models can tell us many things. But unvalidated models cannot be used to make claims like "only CO2 can explain the rise in temperatures since 1960" because that claim presumes the model theory is correct and we cannot know if it is correct until they have been validated against future data. That is why I say the IPCC attribution studies are useless exercises in circular logic.

To use a historical example look at Newton and Einstein. Einstein developed his gravity theory knowing that Newton's theory didn't explain Mercury's orbit. He developed a theory that explained Mercury's orbit and its been subsequently proved in other ways. Planetary orbits are obviously different than a planet's climate system but to say that using historical data to improve a model somehow invalidates the entire model is unrealistic and impractical.

That NOT what I am saying. What I am saying is it would be wrong to make expensive public policy decisions based on Einstein's 'model of gravity' until it has been validated by real experiments.

You've also assumed that any tuning is done by using unrealistic values. There's no reason to think this is always the case. I'm sure some people do this but it is equally possible that historical data shows that some factors were overestimated or underestimated and realistic tuning is possible.

Again. You are completely misrepresenting what I said. I said that the range of realistic values is so wide that the model developers have a lot of latitude and can easily tune their models without using unrealistic tuning. This occurs because we have no data that describes many important climate parameters so this data has to be made up.

But to dismiss all of the good evidence they've looked at based on some criticism of one piece of evidence isn't scientific and certainly isn't balanced if you're willing to consider evidence from the other side when the other side has done equally questionable things.

For starters there is no 'other side'. There are many different individuals with many different opinions. For example, Roger Peilke Sr's opinion on climate change is quite different from Roy Spencer's. However, these scientists claim that the IPCC/climate science community is unreasonably ignoring their work and, under normal conditions, I would assume that this is because the work does not have much merit. However, the fiasco over the hockey stick tells me that the IPCC cannot be trusted to vet scientific work objectively and the scientists complaining that their work is being unreasonably ignored probably have a point.

This has led me to look at the work by these scientists and others. I have found a lot of junk. But I have also found many plausible hypotheses which could explain a portion of the warming since 1960 but are not included in the climate models. If these hypotheses were include then the modellers would need to reduce their estimates for CO2 sensitivity which would, in turn, reduce the magnitude of future warming.

The phenomena which I think have been left out of the models are:

- long term variations in ocean currents (i.e. ocean weather)

- long term variations in cloud cover (i.e. random or caused by cosmic rays)

- large scale land use changes.

- the of aerosols like black carbon on snow and ice.

The links I gave earlier are a starting point for finding research on these topics.

Edited September 2, 2008 by Riverwind

[marksman]

You have misunderstood what I am saying. I am not saying that hindcasts should not be done. I am saying that hindcasts tell us nothing about the predictive skill of the model nor do they tell us that the theory built into the model is correct. The only way to show that the theory built into the model is correct is to use the model to predict outcomes that were not known when the model was developed. That does not mean unvalidated models are useless. In fact, poor models can tell us many things. But unvalidated models cannot be used to make claims like "only CO2 can explain the rise in temperatures since 1960" because that claim presumes the model theory is correct and we cannot know if it is correct until they have been validated against future data. That is why I say the IPCC attribution studies are useless exercises in circular logic.

That NOT what I am saying. What I am saying is it would be wrong to make expensive public policy decisions based on Einstein's 'model of gravity' until it has been validated by real experiments.

Again. You are completely misrepresenting what I said. I said that the range of realistic values is so wide that the model developers have a lot of latitude and can easily tune their models without using unrealistic tuning. This occurs because we have no data that describes many important climate parameters so this data has to be made up.

For starters there is no 'other side'. There are many different individuals with many different opinions. For example, Roger Peilke Sr's opinion on climate change is quite different from Roy Spencer's. However, these scientists claim that the IPCC/climate science community is unreasonably ignoring their work and, under normal conditions, I would assume that this is because the work does not have much merit. However, the fiasco over the hockey stick tells me that the IPCC cannot be trusted to vet scientific work objectively and the scientists complaining that their work is being unreasonably ignored probably have a point.

This has led me to look at the work by these scientists and others. I have found a lot of junk. But I have also found many plausible hypotheses which could explain a portion of the warming since 1960 but are not included in the climate models. If these hypotheses were include then the modellers would need to reduce their estimates for CO2 sensitivity which would, in turn, reduce the magnitude of future warming.

The phenomena which I think have been left out of the models are:

- long term variations in ocean currents (i.e. ocean weather)

- long term variations in cloud cover (i.e. random or caused by cosmic rays)

- large scale land use changes.

- the of aerosols like black carbon on snow and ice.

The links I gave earlier are a starting point for finding research on these topics.

As I posted models that can predict historical data are more likely to give accurate predictions for the future compared to models that can't even predict historical data. We both agree that only future predictions will validate a model but I still think you're dismissing the value of "hindcast" models too quickly.

I'm not sure how the IPCC is using circular logic. The models say to the best of our knowledge human released greenhouse gasses explain warming with low uncertainty values. Using different scenarios of humans releasing more greenhouse gasses those models are used for future predictions. How is that circular logic?

Sorry I wasn't aware your criticism of tuning was based on cost benefit analysis. But we don't make cost benefit decisions based on certain data. You take the cost of a possible risk and the probability that the risk will happen to determine the cost for the purpose of the analysis. If you haven't seen it then search for the Stern report for an analysis of the costs of ignoring climate change. You'll also find criticisms that Stern's discount rate was too low but even with a higher discount rate the costs of ignoring climate change are quite high.

As for the tuning your post implied that scientists were hiding deficiencies in their models by picking variables. If the choice of inputs are scientifically justifiable then noone is really hiding a model's deficiency. Hiding deficiencies implies picking inputs that are unjustifiable so that you don't have to change the underlying theories. Models also include uncertainties and ranges precisely because data maybe lacking or maybe uncertain.

The IPCC goes through the work of thousands of scientists from across the globe and you're going to ignore all of that work because they didn't do an indepth analysis of the statistical methods used in one source? It's the job of the community and the peer review process to do that sort of checking. Given that the US National Research Council looked into the matter and found that there were multiple other sources that confirmed the overall conclusion I don't think you can blame the IPCC for still using the overall conclusion.

The four things you think should be included in models are in many cases accounted for to some degree. Studies have been done by organizations like NASA and the US National Climatic Data Center to account for urban and rural temperature reporting. One study removed all urban sources to remove bias. Another examined temperature data over 100 years in rural and urban areas to try to see how much of an effect urban areas have. So that's being looked at. More complicated models definitely take into account aerosols. It'd be foolish not to given the research done into global dimming and how aerosols played a role in that phenomenon. From what I've read the research on cosmic rays causing clouds and affecting climate is extremely weak not least because there's no clear evidence that there's been an increase in cosmic rays reaching the lower atmosphere over the past few decades in order to cause increased cloud formation. Modelers continue to refine how to model clouds in these climate models because so far there is no consensus technique agreed upon. But this area is being looked at so you can't say people are ignoring clouds. Models also look at oceans and in early climate models poor ocean modeling caused climate drift where the model would start giving unrealistic results. Oceans aren't easy but people are working on modeling oceans better. Recent models use ocean models that are more realistic including things as basic as the fact that there's a thermocline.

Your list of 4 things may not be as precise as everyone would like but they're certainly included to some degree.

[Riverwind]

I still think you're dismissing the value of "hindcast" models too quickly.

I am dismissing the value of the hindcast because the real data collected over the last 8 years suggest that the models over predict the amount of warming. I realize that 8 years of data does not prove the models wrong but it is enough for me to question whether the models have any useful predictive skill.

If you haven't seen it then search for the Stern report for an analysis of the costs of ignoring climate change. You'll also find criticisms that Stern's discount rate was too low but even with a higher discount rate the costs of ignoring climate change are quite high.

Stern's report is one opinion. Other economists have looked at the issue and concluded that the cost of mitigation is too high to justify the up front costs and that R&D is a more cost effective strategy. I tend to agree with the latter opinion because I think the technology to eliminate CO2 emissions does not exist and any attempt to force reductions before the technology exists will either accomplish nothing or kill the economy. In either case mitigation wastes a lot of money for little benefit.

If the choice of inputs are scientifically justifiable then noone is really hiding a model's deficiency. Hiding deficiencies implies picking inputs that are unjustifiable so that you don't have to change the underlying theories.

Choosing a value for the aerosol forcing that happens to give the correct trend does hide the model deficiencies because we cannot know if the choice for aerosol forcing is correct and the model theory is correct or if the model theory is wrong and there happened to be a aerosol forcing choice that compensated for the error in the model theory. You can't call a bad model correct simply by claiming the adjustments were scientifically justified. I am puzzled that you do not immediately see why your approach to validating models is deficient. I have noticed that some scientists who support AGW use similar arguments and it is one of the reasons why am not willing to blindly trust their judgement.

The IPCC goes through the work of thousands of scientists from across the globe and you're going to ignore all of that work because they didn't do an indepth analysis of the statistical methods used in one source?

Again you are completely misunderstanding what I have said. The hockey stick could have been story about one scientist who screwed up and it would have been long forgotten. What created the problem is the reaction of many in the climate science community who, even today, still deny that there was anything wrong with it and continue to use it. The apparent unethical behavior on the part of some journal editors and the willingness of the IPCC to break its own rules in order to create the illusion that the hockey stick was still valid further demonstrates that the community can not be trusted vet science objectively.

Your list of 4 things may not be as precise as everyone would like but they're certainly included to some degree.

They are not included to the degrees that the proponents claim they should be. Peike claims that CO2 can only account for ~25% of the warming. Shaviv suggests 50% due to cosmic rays. Ramasthan said black carbon could account for up to 60%. Spencer claims he has experimentally shown that CO2 sensitivity is much lower than previously claimed which makes the land use/black carbon/cosmic ray effects more plausible.

Obviously, I realize that they all can't be right and the modellers still could be right even if the models have not been validated. That is why I put such an emphasis on validation of the models against future data. If the models have left things out then we would expect to see some warming but not as much as the models predict. So far, the real data is supporting sciensts like Peike and Spencer. I will revise my opinion, if the current trends reverse and the warming increases to where it should be according to the latest set of model predictions. Until that happens I think we should only consider measures that would make sense even if CO2 is not a threat (e.g. build nukes, improve efficiency, invest in R&D).

Edited September 3, 2008 by Riverwind

[peter_puck]

The funny thing is that a scientist who supports human caused global warming is a genius and a respected scientist, and any scientist who offers an opposing viewpoint on global warming is a kook, or on the fringe of the scientific community.

I'm not saying that humans aren't causing global warming. However, there are very intelligent scientists who oppose the theory and get no air time, because the left-wing dominated media will only show what they want you to believe. There are two strong sides to this debate, but we are only hearing one side.

By the way, Al Gore is the biggest hypocrite you'll ever find.

I think the point is that many of the people who oppose AGW were kooks and on the fringe of science BEFORE the AGW debate.

Yes, there are a handful of serious people, but not that many.

[peter_puck]

So is water vapor, and to a much larger degree...shall we begin efforts to limit water vapor "emissions" as well?

The amount of water in the atmosphere is fixed (for a given temperature). If I put a gallon of water into the atmosphere, in a short period of time a gallon of water will fall on my head. I can put as much water in the air as I want, but the concentration will quickly return to normal. The air can hold only a limited amount of water vapor, so the rest is quickly turned into water.

Remember that the amount is fixed for a given temperature. The real scary global warming predictions say that when you increase the temperature, you increase the amount of water vapor...which being a greenhouse gas increases the temperature.....

C02 is the reverse. The Earth can only hold a limited amount of C02 (plant life, corals, free CO2 in the ocean, so what is left over gets put into the air. If I put a tonne of C02 into the air, the earth will have an extra tonne for years to come. (yes, its is more complicated than that, but I don't think you want a 10 page science lesson.

To make an analogy. If I drink 500 barrels of water in a week then wait for a few days, the amount of water in my system will the same as a twin who didn't drink any water (assuming their is no kidney damage).

On the other hand, if I eat 500 barrels of jelly beans (which contain carbon) my ass will be fat for years.

[Riverwind]

The real scary global warming predictions say that when you increase the temperature, you increase the amount of water vapor...which being a greenhouse gas increases the temperature.....

What the model makers seem to forget is water vapour has a habit of turning into clouds and rain - things that actually cause cooling. This kind of negative feedback is very common in virtually every stable natural system. Yet the climate modellers woud like us to believe that the climate is a highly unstable system dominated by positive feedback. Here is a link to a peer reviewed paper that found experimental evidence that these negative feedbacks via clouds do exist in the tropics: http://www.weatherquestions.com/Spencer_07GRL.pdf Edited September 4, 2008 by Riverwind

[peter_puck]

What the model makers seem to forget is water vapour has a habit of turning into clouds and rain - things that actually cause cooling. This kind of negative feedback is very common in virtually every stable natural system. Yet the climate modellers woud like us to believe that the climate is a highly unstable system dominated by positive feedback. Here is a link to a peer reviewed paper that found experimental evidence that these negative feedbacks via clouds do exist in the tropics: http://www.weatherquestions.com/Spencer_07GRL.pdf

"Model makers" don't forget such things. They have Phd's in Physics and atmospheric sciences - not in economics or political science. Some people specialize only in a tinny area of cloud formation. They debate these things at a higher level than either of could participate in

There are all sorts of feedbacks, positive and negative. The bitch is that some of the negative feedbacks turn into positive ones once you hit a certain tipping point.

Right now the arctic and the oceans are carbon sinks. They are net absorbers of C02. Get the earth hot enough, this changes - they become positive feedbacks. That is a very dangerous tipping point.

[Riverwind]

Right now the arctic and the oceans are carbon sinks. They are net absorbers of C02. Get the earth hot enough, this changes - they become positive feedbacks. That is a very dangerous tipping point.

And higher CO2 levels increase the amount of plant growth - particularily in the ocean - which results in a negative feedback. The carbon cycle is poorly understood. Scientists don't know why the planet is consistently absorbing 1/2 of human emissions even as these emissions are increasing. This increase in absorption is even more puzzling if one assumes that higher temperatures reduce the amount of CO2 that the oceans can absorb.

[marksman]

I am dismissing the value of the hindcast because the real data collected over the last 8 years suggest that the models over predict the amount of warming. I realize that 8 years of data does not prove the models wrong but it is enough for me to question whether the models have any useful predictive skill.

You've dismissed hindcast modeling for a number of different reasons. First they were circular logic but I'm not sure you've explained how. You've described the method as crappy. Then tuning models was inappropriate because there are "too many tuneable parameters" but this type of reasoning means we might as well give up on trying to figure out any complex system.

You've used aerosols as an example of how tuning is wrong but the link you provided concludes in the executive summary that even though more research is necessary the models currently used in the IPCC 4th report take into account a vastly improved understanding of how aerosols behave and how much is out there. In other words - new models have been tuned to take into account better information. That not only seems reasonable but seems like how science is supposed to work.

Observed temperatures do fall within the range of estimates in the IPCC 3rd report. After 8 years I'm not ready to declare these estimates or the IPCC 4th report estimates right or wrong. When an estimate uses a 20 year interval as its base for a 100 year prediction then 8 years isn't enough to question the model.

To question a model's predictive skill before a statistically significant period of time has passed is premature.

Stern's report is one opinion. Other economists have looked at the issue and concluded that the cost of mitigation is too high to justify the up front costs and that R&D is a more cost effective strategy. I tend to agree with the latter opinion because I think the technology to eliminate CO2 emissions does not exist and any attempt to force reductions before the technology exists will either accomplish nothing or kill the economy. In either case mitigation wastes a lot of money for little benefit.

What part of mitigation are you against? Mitigation includes R&D, improving efficiency and using power sources not based on fossil fuels - like nuclear, solar etc. These are all things you say you are for.

You've fallen into at least 3 strawman arguments. We're not talking about eliminating co2 emissions - elimination and reduction are different. Your desire to build nuclear power plants shows the claim that technology doesn't exist is false. Efforts that accomplish nothing or destroying the economy aren't the only 2 options on the table - economically reasonable measures can be taken that will affect greenhouse gas emissions.

Choosing a value for the aerosol forcing that happens to give the correct trend does hide the model deficiencies because we cannot know if the choice for aerosol forcing is correct and the model theory is correct or if the model theory is wrong and there happened to be a aerosol forcing choice that compensated for the error in the model theory. You can't call a bad model correct simply by claiming the adjustments were scientifically justified. I am puzzled that you do not immediately see why your approach to validating models is deficient. I have noticed that some scientists who support AGW use similar arguments and it is one of the reasons why am not willing to blindly trust their judgement.

I'm puzzled by a number of things in your posts. You keep saying that you can't use past data to validate these models except that I wasn't saying that. You say that my approach to validating models is deficient even though I've stated that in order to validate models you've got to have verified future predictions - which is the same as what you're saying. So if my approach is deficient then so is yours!

The process of creating a model, giving it historical inputs and then seeing how the results compare to historical data is useful even if its future predictions don't pan out. It helps refine the model rather than just shooting in the dark. Checking against historical data doesn't guarantee future success but it does increase the chance that the predictions will be more accurate.

It's true that in science 2+2=4 can mask the fact that you should be using 1+3=4 but this is true for everyone. Unless you're making the same criticisms against everyone then I don't see why this is a climate model problem. In fact you should be criticising one of your quoted sources for this very problem that I'll explain below.

Again you are completely misunderstanding what I have said. The hockey stick could have been story about one scientist who screwed up and it would have been long forgotten. What created the problem is the reaction of many in the climate science community who, even today, still deny that there was anything wrong with it and continue to use it. The apparent unethical behavior on the part of some journal editors and the willingness of the IPCC to break its own rules in order to create the illusion that the hockey stick was still valid further demonstrates that the community can not be trusted vet science objectively.

How am I misunderstanding your characterization of the hockey stick graph debate? You're using one example to say that you don't trust the IPCC. And the example isn't even as bad as it's made out to be since there is some validity to the statistical methods used and even your link pointed out that the low R2 value means that the graph might not have predictive value - which doesn't make the graph wrong. Other studies besides the A&W one agreed with the overall conclusion of the graph meaning it isn't invalid. If you're worried about the reactions of the community why is it that you aren't distrustful of the anti-human caused climate change crowd when some in that group actively promote papers which distort the conclusions of other scientists? That seems blatantly unethical compared to using questionable statistical methods.

They are not included to the degrees that the proponents claim they should be. Peike claims that CO2 can only account for ~25% of the warming. Shaviv suggests 50% due to cosmic rays. Ramasthan said black carbon could account for up to 60%. Spencer claims he has experimentally shown that CO2 sensitivity is much lower than previously claimed which makes the land use/black carbon/cosmic ray effects more plausible.

Obviously, I realize that they all can't be right and the modellers still could be right even if the models have not been validated. That is why I put such an emphasis on validation of the models against future data. If the models have left things out then we would expect to see some warming but not as much as the models predict. So far, the real data is supporting sciensts like Peike and Spencer. I will revise my opinion, if the current trends reverse and the warming increases to where it should be according to the latest set of model predictions. Until that happens I think we should only consider measures that would make sense even if CO2 is not a threat (e.g. build nukes, improve efficiency, invest in R&D).

So 1st the 4 phenomenon weren't included in models and now it's that they're just not included to the degree you and their proponents think they should be. That's a valid debate of course but the fact that there is debate doesn't make the models invalid.

At least 1 of your sources - Spencer - isn't convincing. It didn't take long to find an analysis of some of his anti human caused climate change work showing that it's not useful or scientific.

http://www.realclimate.org/index.php/archi...e-easy-lessons/

You complain about people tuning their models but it looks like Spencer is guilty of this very thing. He's messed around with numbers quite a bit to show climate change has nothing to do with greenhouse gasses. And he didn't even require a lot of variables. He's definitely tuning his model to cover up his model's deficiencies. Not to mention the nice postlude on the site that points out how Spencer continued to promote obviously wrong data.

You've added another link recently to one of his papers showing clouds can have an affect on climate. But that isn't anything new just further proof of what people already knew. Models showing human caused climate change do take these things into consideration.

[Riverwind]

In other words - new models have been tuned to take into account better information. That not only seems reasonable but seems like how science is supposed to work.

There is no accurate data for aerosols distributions in the past. Any historical data that is fed into the models is basically made up. Now the people making up the data may be doing so in going faith and they may have a reasonable rational for the choices they made but that does not change the fact that the data does not exist and had to be fabricated. This is same problem exists for numerous other parameters which can have a significant effect on the output of the models. This made up data comes with large error bars which gives the modellers a lot of flexibility when it comes to deciding what values they actually use within their models. This means we cannot possibly know whether the correct hindcasts means the model theory is correct or whether the model theory is wrong and it was simply covered up with parameter choices that compensated for the error in the model theory.

That is why the IPCC attribution studies are exercises in circular logic. They take models that have been tuned to show that their CO2 hypotheses can match the temperature trends and then use those same models to claim that no other hypotheses can explain the temperature trends. It is a meaningless comparison.

Observed temperatures do fall within the range of estimates in the IPCC 3rd report. After 8 years I'm not ready to declare these estimates or the IPCC 4th report estimates right or wrong. When an estimate uses a 20 year interval as its base for a 100 year prediction then 8 years isn't enough to question the model.

Only for models that have been validated. These models have never been validated so they cannot be presumed to be right.

To question a model's predictive skill before a statistically significant period of time has passed is premature.

It is premature to assume that a model is valid until it has demonstrated the ability to predict outcomes over a statistically significant period of time.

What part of mitigation are you against? Mitigation includes R&D, improving efficiency and using power sources not based on fossil fuels - like nuclear, solar etc. These are all things you say you are for.

I am against any attempt to regulate energy sources out of existance before alternatives have been proven. It is possible that large scalar solar or wind projects will be proven viable in 20-30 years. Until then we need to build as many coal fired plants as necessary to ensure we can meet our growing energy needs. We cannot build nuclear plants fast enough to make up the gap.

So if my approach is deficient then so is yours!

You fail to see that tuning - even when scientifically justified - can hide a deficiencies in a model.

The process of creating a model, giving it historical inputs and then seeing how the results compare to historical data is useful even if its future predictions don't pan out. It helps refine the model rather than just shooting in the dark. Checking against historical data doesn't guarantee future success but it does increase the chance that the predictions will be more accurate.

No disagreement from me. The only difference is I am saying models cannot be used to drive policy decisions until they have been validated against future data. The IPCC is not doing that. They are taking models that have only been verified with hindcasts and claiming that they are reliable enough to for policy decisions.

And the example isn't even as bad as it's made out to be since there is some validity to the statistical methods used and even your link pointed out that the low R2 value means that the graph might not have predictive value - which doesn't make the graph wrong. Other studies besides the A&W one agreed with the overall conclusion of the graph meaning it isn't invalid.

From the Wegman report: bad method + correct result = bad science. I am appalled that so many people who are supposedly trained in science don't get that.

Not to mention the nice postlude on the site that points out how Spencer continued to promote obviously wrong data.

I can't find spenser's rebuttle to the real climate post at the moment but the entire thing is a strawman that does not address the point he was making. The smear against him over the errors in the satillite data are extremely ironic give real climate's defence of the hockey stick. The contrast between Spencer and Mann is interesting. Both tried to develop new methods of extracting new insights from existing data. Spencer published his method and his data and allowed others to verify it. They found an error and Spencer acknowledged it, fixed it and moved on. Mann, on the other hand, refused to release his methods or data and refused to acknowledge his errors even when they were so obvious that they could be denied. Edited September 5, 2008 by Riverwind

[marksman]

There is no accurate data for aerosols distributions in the past. Any historical data that is fed into the models is basically made up.

The link you posted suggested otherwise - that while still uncertain the data used in current models is much more accurate than data used in earlier models. All I was saying was that your argument against tuning was flawed because most tuning is done with improved knowledge not randomly picking possibly justifiable values.

This means we cannot possibly know whether the correct hindcasts means the model theory is correct or whether the model theory is wrong and it was simply covered up with parameter choices that compensated for the error in the model theory.

This applies to all models - both models that find human caused climate change and models that find against human caused climate change. But you only argue against those models that show human caused climate change.

That is why the IPCC attribution studies are exercises in circular logic. They take models that have been tuned to show that their CO2 hypotheses can match the temperature trends and then use those same models to claim that no other hypotheses can explain the temperature trends. It is a meaningless comparison.

That's not what they're doing. Models were created based on what we know so far. Initial conditions were input. Models based only on natural causes didn't give results that matched historical records. Models based only on human causes didn't give results that matched historical records. But models that incorporated both gave relatively accurate matches. Based on this it's concluded that human causes are a factor in driving climate change. There's nothing circular about that.

Your argument only has merit if you think the models were unscientifically modified to give a specific result. But that argument can equally apply to models that show no human caused climate change. In other words you're assuming that the models are wrong in order to show that the models are wrong. Ironically this is circular logic.

Only for models that have been validated. These models have never been validated so they cannot be presumed to be right.

It is premature to assume that a model is valid until it has demonstrated the ability to predict outcomes over a statistically significant period of time.

If you'd read my post you'd know that I'm saying it's too early to tell if they're right or wrong. But you've already stated that you're making a judgment on the models without a statistically significant period of time passing.

You fail to see that tuning - even when scientifically justified - can hide a deficiencies in a model.

You failed to see where I admitted that's possible. Then failed to recognize that this criticism applies to all scientific models of all types including the models that claim no human caused climate change.

No disagreement from me. The only difference is I am saying models cannot be used to drive policy decisions until they have been validated against future data. The IPCC is not doing that. They are taking models that have only been verified with hindcasts and claiming that they are reliable enough to for policy decisions.

Policy decisions don't require 100% certain data otherwise we'd never make policy decisions. Risk assessment and cost benefit decisions can be made by multiplying the cost of something by the probability that it'll happen. The IPCC models contain uncertainty information and can be used as a valid basis for policy decisions - just like models not produced by the IPCC or models that have nothing to do with climate change.

From the Wegman report: bad method + correct result = bad science. I am appalled that so many people who are supposedly trained in science don't get that.

You said the graph was invalid and I pointed out that wasn't true so now you've switched back to your bad method point. But that doesn't change the fact that you can't claim the IPCC is unethical because they're using an invalid graph - since the graph hasn't been proven invalid.

I can't find spenser's rebuttle to the real climate post at the moment but the entire thing is a strawman that does not address the point he was making. The smear against him over the errors in the satillite data are extremely ironic give real climate's defence of the hockey stick. The contrast between Spencer and Mann is interesting. Both tried to develop new methods of extracting new insights from existing data. Spencer published his method and his data and allowed others to verify it. They found an error and Spencer acknowledged it, fixed it and moved on. Mann, on the other hand, refused to release his methods or data and refused to acknowledge his errors even when they were so obvious that they could be denied.

All scientists should be releasing all results - I'm not saying otherwise. You're just attempting to avoid the fact that Spencer manipulated variables in his more recent graphs to prove what he wanted to prove. This was something you were supposedly against and used this reason to discount the work of literally hundreds of other scientists.

Mann's actions don't invalidate all of the work done by others who think humans are causing climate change. If you're against that position because of Mann then you need to rethink your position since similar actions have been taken by those who don't think humans are causing climate change.

[Riverwind]

All I was saying was that your argument against tuning was flawed because most tuning is done with improved knowledge not randomly picking possibly justifiable values.

Then explain why different models choose radically different values for the same paramters?

This applies to all models - both models that find human caused climate change and models that find against human caused climate change. But you only argue against those models that show human caused climate change.

Sure. I am not saying that natural variability hypothesis is proven. All I am saying is the claim that the warming from 1960 was "most likely" (IPCC words) due to GHGs grossly over states the certainty of the science. All we can really say is the hypothesis that that the warming from 1960 was due to GHGs is plausible.

That's not what they're doing. Models were created based on what we know so far. Initial conditions were input. Models based only on natural causes didn't give results that matched historical records. Models based only on human causes didn't give results that matched historical records. But models that incorporated both gave relatively accurate matches. Based on this it's concluded that human causes are a factor in driving climate change. There's nothing circular about that.

They tune the models to match the historical record. They then used the same models but removed the anthropogenic forcings and - surprise - surprise - the natural forcing only runs don't match. The only way that kind of comparison would be valid is if they developed new models and applied the same kind of "scientifically justifiable" parameter selection in order to get the best fit possible for natural forcings. If they still did not match it would not conclusively prove that the models were right but it would no longer be a circular argument.

You failed to see where I admitted that's possible. Then failed to recognize that this criticism applies to all scientific models of all types including the models that claim no human caused climate change.

Agreed. I think we don't have a clue what is happening to the climate and pretending we do is quite arrogant. The CO2 hypothesis is plausible but not proven. However, the same applies to the natural forcings + land use + CO2 hypothesis.

Policy decisions don't require 100% certain data otherwise we'd never make policy decisions. Risk assessment and cost benefit decisions can be made by multiplying the cost of something by the probability that it'll happen.

The problem is created because people are using the outputs of unvalidated models to calculate the probability of adverse outcomes. This can lead to dangerously flawed decisions making. The was an op-ed in NYT a little while ago that claimed that there was a 5% chance of 10degC rise in temperature and 1% chance of a 20 degC rise. These numbers were apparently derived from the model outputs, however, you don't have to be climate scientist to understand that assigning a 1% chance of a 20 degC is patently absurd because the planet has never been that hot even when CO2 levels where 20 times what they are today.

IOW, waiting for 100% certainity is a bad idea but making decisions based on probability distribution from unvalidated models is an even worse idea.

You said the graph was invalid and I pointed out that wasn't true so now you've switched back to your bad method point. But that doesn't change the fact that you can't claim the IPCC is unethical because they're using an invalid graph - since the graph hasn't been proven invalid.

The graph is invalid because the methods used to create it were flawed. Whether is happens to match *some* other studies is irrevelent - the graph is still invalid and should have been completely omitted from subsequent reports.

You're just attempting to avoid the fact that Spencer manipulated variables in his more recent graphs to prove what he wanted to prove.

The critique of Mann was supported by Wegman and grudgingly acknowledged by the NAS panel. Even his defenders are reduced to saying the method was bad but the results are still correct. The critique of Spencer comes from Pierrehumbert on RC who consistently misrepresents the work of other scientists so he can create the illusion that he debunked the paper. Here is Spencer's reply:

http://climatesci.org/2008/05/22/a-respons...by-roy-spencer/

Ray is quite simply wrong — and the reviewers of our paper (Piers Forster and Isaac Held) agree with me. It matters a great deal whether radiative fluctuations are the result of feedback on surface temperature, versus the myriad other variables that control cloudiness. Piers Forster was honest enough to admit that their neglect of the internal variability term in Eq. 3 of “The Climate Sensitivity and its Components Diagnosed from Earth Radiation Budget Data” (Forster and Gregory, J. Climate, 2006) was incorrect, and that it indeed can not be neglected in feedback diagnosis efforts using observational data. He also stated that the climate modeling community needs to be made aware of this.

Mann's actions don't invalidate all of the work done by others who think humans are causing climate change.

Of course not. But the fact that many others seem to be willing to defend him, even now, tells me that many in the climate science community are not looking at the science objectively. This means I cannot rule out the opinions of dissenting scientists because their opinions are not shared by the consensus. I cannot know if the consensus is based on real objective science or an example of groupthink that infests many organizations.

BTW: Here is a good illustration of why I don't think natural factors have been properly accounted for: http://www.climate-skeptic.com/2008/08/global-warming.html

Look at the two graphs - virtually identical slopes yet one is mostly natural and one is mostly anthopogenic. We are talking 50 year periods too so you can't argue the timeframe is too short. We also know from that latest solar research that the sun was doing the same thing in 30s as it was in 90s so the sun can't explain the rise. I don't see how any objective scientist can look at those graphs and claim that they are 90% certain that the planet would have cooled from 1960 to 2008 without anthopogenic forcings.

Edited September 5, 2008 by Riverwind

[peter_puck]

And higher CO2 levels increase the amount of plant growth - particularily in the ocean - which results in a negative feedback. The carbon cycle is poorly understood. Scientists don't know why the planet is consistently absorbing 1/2 of human emissions even as these emissions are increasing. This increase in absorption is even more puzzling if one assumes that higher temperatures reduce the amount of CO2 that the oceans can absorb.

The carbon cycle is well understood in terms of the amount in the air. The rate of C02 has been increasing dramatically. There may be plenty of sinks, but they are not absorbing it nearly as fast as it is produced.

You also have to remember too that carbon sinks are not linear. The arctic has been a carbon sink, but may soon "flip" to become a carbon source because of melting permafrost.

The same thing with the ocean, right now it is a sink, but if the world gets much hotter, it could start to emit more than it absorbs as well.

You also run into the problem that there is a limit as to how much carbon a given ecosystem. Increased C02 may cause trees to grow taller in the rain forest (hypothetical), but eventually that tree will run out of nutrients, run into the law of physics, etc.

The same thing with the oceans. If you increase the number of aquatic plants, you will increase predators.

[Bryan]

You also run into the problem that there is a limit as to how much carbon a given ecosystem.

Carbon is not being added to the system though, it's just being redistributed. Ground, oil, air, oceans, plants, trees, etc, it's just moving from one place to another.

[marksman]

Then explain why different models choose radically different values for the same paramters?

Most models are based on different scenarios like what happens if we continue emitting as we've been emitting, what happens if we reduce by a certain amount etc. Even so different values doesn't prove that they're being chosen unscientifically.

Sure. I am not saying that natural variability hypothesis is proven. All I am saying is the claim that the warming from 1960 was "most likely" (IPCC words) due to GHGs grossly over states the certainty of the science. All we can really say is the hypothesis that that the warming from 1960 was due to GHGs is plausible.

I don't think "most likely" is so far fetched when several models created by lots of different scientists all point in that direction.

They tune the models to match the historical record.

Once again you've started from the position that the models were designed to match the historical record and not designed to reflect the theory. If this is your starting point then you'll always conclude that it's circular logic.

Agreed. I think we don't have a clue what is happening to the climate and pretending we do is quite arrogant. The CO2 hypothesis is plausible but not proven. However, the same applies to the natural forcings + land use + CO2 hypothesis.

Human caused climate change isn't proven but it's more than plausible. No science is perfect but the majority of the work points to likely human caused climate change. What I find arrogant is the fact that most scientists - even many of the people saying that climate change research is more uncertain than the IPCC says it is - believe that humans are changing the climate to some degree but people still insist on a do nothing approach because we don't "know" for certain. Instead of dealing with that situation we're ignoring it and arguing about the difference between likely and most likely. Many of the proposals for dealing with climate change are valid for other reasons as you've already suggested - energy efficiency, R&D - but even this doesn't seem to stop people from saying we should do nothing.

IOW, waiting for 100% certainity is a bad idea but making decisions based on probability distribution from unvalidated models is an even worse idea.

Or we could take the third option which is use models that while unvalidated by future predictions still have acceptable uncertainty. Despite criticisms most in the scientific community think the climate models have acceptable uncertainty.

The graph is invalid because the methods used to create it were flawed. Whether is happens to match *some* other studies is irrevelent - the graph is still invalid and should have been completely omitted from subsequent reports.

Other studies that used acceptable methods reached similar conclusions. You'll even notice the graph has changed slightly over the years.

Your reasoning would mean that if we're both working on the same problem and come up with similar results and you use 1 method and I use a more questionable method then your results are invalid because of my more questionable method. That makes no sense.

You can't say the graph is invalid because one of the scientists who came up with that result used questionable methods even though others reached the same results.

The critique of Spencer comes from Pierrehumbert on RC who consistently misrepresents the work of other scientists so he can create the illusion that he debunked the paper. Here is Spencer's reply:

http://climatesci.org/2008/05/22/a-respons...by-roy-spencer/

I'm sure he "consistently misrepresents the work of other scientists" but those arguing against the work that concludes humans are having an effect on climate don't do that.

I've read the reply. It did nothing to defend against the allegations that Spencer cherrypicked values to make his graph. It also didn't instill a lot of confidence in the robustness of his work.

And besides, the SOI/PDO example took me 1 hour on a weekend with a very simple single idea, internet access, and an Excel spreadsheet.

Why is the graph being promoted as showing no human caused climate change under these conditions?

BTW: Here is a good illustration of why I don't think natural factors have been properly accounted for: http://www.climate-skeptic.com/2008/08/global-warming.html

Look at the two graphs - virtually identical slopes yet one is mostly natural and one is mostly anthopogenic. We are talking 50 year periods too so you can't argue the timeframe is too short. We also know from that latest solar research that the sun was doing the same thing in 30s as it was in 90s so the sun can't explain the rise. I don't see how any objective scientist can look at those graphs and claim that they are 90% certain that the planet would have cooled from 1960 to 2008 without anthopogenic forcings.

If the blogger wanted to compare a 50 year natural period with a 50 year anthropogenic period why didn't they pick 1850 to 1900 for the natural period? Probably because the curve is flat and it wouldn't have "proven" their point. Chopping a graph in 2 doesn't really prove much and this graph isn't proof either way that natural causes have been accounted for or not. On the other hand scientists have been able to show that a combination of causes reasonably explains the entire century with the result that human caused climate change is likely responsible for the upward trend at the end of the century.

I was once shown two other graphs that looked remarkably similar - just like your link. One showed human birth rates increasing then decreasing. The other showed the stork population increasing then decreasing. Therefore storks bring babies.

I'm more convinced by actual scientific research then someone using Windows Paint on a graph.

[Riverwind]

On the other hand scientists have been able to show that a combination of causes reasonably explains the entire century with the result that human caused climate change is likely responsible for the upward trend at the end of the century.

Your argument is fallacious. Demonstrating that GHGs *could* explain the warming from 1960 does not show that the GHGs *did* cause the warming. The graph from 1890 to 1940 demonstrates something other than GHGs can cause a similar rise so any one who wishes to claim that GHGs have "most likely" caused the warming from 1960 must first 1) explain what caused the identical warming until 1940 and 2) demonstrate that the same cause has not contributed to the warming from 1960. The IPCC has not done that. They have simply asserted that GHGs can be the only possible explaination since they were able to find a combination of model parameters that allows the models to match the most recent record.

If you disagree then please explain why the temperature rise until 1940 happened and please provide evidence that the same cause is not at work over the last 50 years. Keep in mind you can't blame it on the sun because the most recent research by solar scientists like Leif Svalgaard suggests that the sun has remained constant for the last century or more. And you also shold note that the IPCC's own model runs referenced in AR4 Chapter 9 do not really match the record up until 1940 because they only show a rise of 0.2 degC when the actual record says the rise was 0.6 degC.

If climate scientists don't have a clue why the temperature rose from 1910 to 1940 then they cannot claim to be extremely certain that GHGs caused the rise from 1960.

P.S. One other landmine to watch out for. Some models do actually make use of the old solar data to explain the warming. This indicates that the theory behind those models is clearly wrong yet these models are incorporated into IPCC projections. So you will need to also explain why we should make policy decisions based on models when we already know that they need to be reworked to take into account the latest solar research.

P.P.S Roy Spencer's models were likely developed using the same process that was used for the IPCC models (i.e. tune them until they produce the desired result). However, Roy Spencer is *not* claiming that his model proves that the IPCC models are wrong. He is only demonstrating what I have been arguing all along: that the IPCC has not proven that natural factors do not exist. IOW - he (unlike the IPCC) is not over stating the significance of his results.

Edited September 7, 2008 by Riverwind

[marksman]

Your argument is fallacious. Demonstrating that GHGs *could* explain the warming from 1960 does not show that the GHGs *did* cause the warming.

Which is why I talk about certainty with the models and don't make statements like "GHGs 100% did cause warming". You've mischaracterized my argument. Unless you think the phrase "is likely responsible" is equivalent to "is responsible".

An unsound argument would be something like saying that you think all models are built to show a certain result which is why you can't trust them even though there's no proof that models are built to show any particular result.

The IPCC has done studies on temperature in the last century and to the best of their knowledge - which means the best knowledge of 100s of scientists - the factors that explain the 1890-1940 warming do not explain the subsequent warming unless you include GHGs. You ignore this by saying that

They have simply asserted that GHGs can be the only possible explaination since they were able to find a combination of model parameters that allows the models to match the most recent record.

without showing that the models were rigged in this way or explaining how different models have reached the same conclusion.

Keep in mind you can't blame it on the sun because the most recent research by solar scientists like Leif Svalgaard suggests that the sun has remained constant for the last century or more.

You ignore your own sources when you say things like this. Isn't one of your sources saying that changes with the sun allow more cosmic rays into the lower atmosphere causing cloud formation? Total solar irradiance may be more limited than we thought but that doesn't mean the sun's total effect on climate is necessarily negligible.

P.S. One other landmine to watch out for. Some models do actually make use of the old solar data to explain the warming. This indicates that the theory behind those models is clearly wrong yet these models are incorporated into IPCC projections.

So your objection here is that models created in 2004 should've taken into account scientific results from 2007? Nothing in that graph shows the model is "clearly wrong" since it says nothing about how the line for solar forcing was determined.

So you will need to also explain why we should make policy decisions based on models when we already know that they need to be reworked to take into account the latest solar research.

Research constantly moves forward. You're back to arguing that we need 100% certainty before doing anything.

P.P.S Roy Spencer's models were likely developed using the same process that was used for the IPCC models (i.e. tune them until they produce the desired result). However, Roy Spencer is *not* claiming that his model proves that the IPCC models are wrong. He is only demonstrating what I have been arguing all along: that the IPCC has not proven that natural factors do not exist. IOW - he (unlike the IPCC) is not over stating the significance of his results.

So you discount Mann's work because of his methods - and also discount anyone else who reached similar conclusions even if they used different methods - but you don't discount Spencer's work when he uses questionable methods. Unlike IPCC models it's been clearly demonstrated that Spencer's values were picked unrealistically.

The IPCC has never stated that natural factors don't exist so they wouldn't have any reason to "prove" this fact. When Spencer says that natural forces can explain global warming and his proof is a made up graph then he's overstating the significance of his results. Nothing here has shown that the IPCC is overstating the significance of their results.

[Riverwind]

Unless you think the phrase "is likely responsible" is equivalent to "is responsible".

There is a 0.6 degC warming from 1910 to 1940 that is not explained by the models. Therefore, it is not reasonable to claim that the GHGs are likely the cause of the warming from 1960. The only claim that can be made is that the warming from 1960 can be plausibly explained by GHGs. Any statement with more certainty than that is not supported by the evidence.

An unsound argument would be something like saying that you think all models are built to show a certain result which is why you can't trust them even though there's no proof that models are built to show any particular result.

There is no reason to believe the models were not tuned to produce a result that conformed to the hypotheses preferred by the modellers. There are simply too many unknown parameters for them to do anything else.

The IPCC has done studies on temperature in the last century and to the best of their knowledge - which means the best knowledge of 100s of scientists - the factors that explain the 1890-1940 warming do not explain the subsequent warming unless you include GHGs. You ignore this by saying that

And they have published those results and those results do not support their claims. The 0.6 degC rise is not explained by the IPCC models but they have simply ignored that inconsistency.

You ignore your own sources when you say things like this. Isn't one of your sources saying that changes with the sun allow more cosmic rays into the lower atmosphere causing cloud formation? Total solar irradiance may be more limited than we thought but that doesn't mean the sun's total effect on climate is necessarily negligible.

Skeptics have a number of hypotheses that could result in a much larger solar effect but these hypotheses have been vigously rejected by the climate modellers. Therefore these hypotheses cannot be used to explain the rise in order to defend the IPCC claims. If they are used to defend the IPCC claims then it will also be necessary to include the effect in the post 1960 rise as well.

So your objection here is that models created in 2004 should've taken into account scientific results from 2007? Nothing in that graph shows the model is "clearly wrong" since it says nothing about how the line for solar forcing was determined.

That model needs the solar irradiance changes in order to explain the rise until 1940. Without that change then it does not match which means no useful conclusions can be drawn from the model until it is updated. Obviously science moves forward but when it is rediculous to insist that we can make policy decisions based on models which can't even get hindcasts right. We don't need 100% certainty but unvalidated models that can't get the hindcasts right don't even give us the minimal certainty required for an sensible decision.

The IPCC has never stated that natural factors don't exist so they wouldn't have any reason to "prove" this fact.

The IPCC insists that only GHGs can explain the rise since 1960. IOW - the IPCC has chosen to ignore all factors other than GHGs. Pro-forma acknowledgements that 'natural factors exist' are meaningless if they have no impact on the ultimate conclusions.

In any case, it comes down to trust. The fiasco over the Mann hockey stick demonstrates that the IPCC and many influential climate scientists cannot be trusted. Nor am I absolutely certain that I can trust skeptical scientists like Spencer either. But I can look at the data and the arguments and the data is tells me that the actual warming has been consistently less than what the models predict and it clear that the IPCC has choosen to ignore a number of plausible explainations for the warming that would result in lower than expected warming. The trends may reverse in a few years and I will adjust my opinion accordingly. Until then, I think it is irresponsible to make policy decisions that would make no sense if the entire CO2 scare turns out to be nothing.

Edited September 9, 2008 by Riverwind

[Moonbox]

Does anyone remember in the 1970's how we were supposedly headed towards a new Ice Age? Remember all the scientists who were backing that theory up?

I don't have an opinion on whether or not global warming is happening and whether or not it's caused by man.

What I do know, however, is that both sides have come up with nothing but highly disputable evidence.

Anyone who says, "Global Warming is happening and there can be no doubt about it." is full of shit.

Anyone who says the opposite is also full of shit.

We don't know yet and we have no proof yet.