Global Warming backdown

[Pliny]

Yet such models and estimates are used to correctly predict a variety of other very complex phenomena, from combustion inside jet engines, to airflow over complex bodies, to the mechanics of the shockwaves in nuclear explosions, to the evolution of stars and galaxies.

Dismissing computer models off hand, when they now form the backbone of scientific and engineering work, is somewhat silly.

The models are only as good as the data put in. There are too many variables in climate to make accurate models. Computer models are not definitive but are excellent in probabilities. Something entirely unnecessary in a predetermined time continuum, if you know what I mean.

Edited July 8, 2010 by Pliny

[waldo]

squawk! Pope Gore! squawk! Pope Gore! squawk! Pope Gore!

What's the latest from the Pope?

I imagine like your poster child Pope Gore,

... Pope Al Gore and The Book of An Inconvenient Truth? Yuk! Yuk!

Al Gore still the infallible Pope?

...such as presented by Pope Gore

It will take some time before you get the word from Pope Gore.

...from your High priests and Pope Gore.

Edited July 8, 2010 by waldo

[waldo]

TimG - you are unfit for this forum - you make far too much sense....you're much too rational.

sycophants-R-Us, hey Simple?

[waldo]

most refreshing to see you acknowledge a longer-term timeframe for, as you say, “climatic significance”. Perhaps that might actually rub off on the usual cast of MLW numnuts who regularly trot out their 4-to-7 year cherry-picked short-term temperature trending nonsense….. or even those Shady practices that allow one particular MLW numnut to idiotically, to this day, quote the Phil Jones response concerning statistical significance and post1995 warming of CRU data, within his MLW signature.

Your usual snubbing of arguments without really addressing them.

The argument presented was that computer climate models from the late '90's predicted higher warming trends than what occurred in reality. According to those models the warming trend was slowing and obviously if warming is slowing down then that is often termed "cooling".

Politicians have ordered a slowing down of the global warming rhetoric and have decided that global cooling is the trend. Scientists will soon present the peer reviewed data to back them up. It will take some time before you get the word from Pope Gore.

huh! Pliny, if you had a point to make you certainly didn't need to reply to a quote that had no bearing on some "argument" you would presume to make - as you've attempted to make, in the past. That actual quote was in regards TimG's highlighting, in his view, that ~20 years of Greenland ice sheet melting data isn't sufficient to separate signal/noise. When you've made this same bone-head play in the past I've simply asked you to substantiate it... put it in context, state 'which models', what timeframe, etc. Even if we give you credit for parroting some denier blog, your 90's reference puts you smack into the numnut category that attempts to cherry-pick short-term trends. Of course, even if it were true (which it's not), the real howler is that you would presume to categorize "less warming"... still warming, but at a lesser warming rate... as, wait for it, wait for it...... "global cooling"!

It doesn't help your cause to not address the issue. This thread has many references that you ignore and think you are finished with it. If you can recall the thread at all, if it is not just a blur of conquests and put downs in your own mind then you must remember the references that were cited in regards to the above. My reciting or requoting old material on the thread is not really productive.

The only thing that you serve a purpose for is in reminding us to keep on track and seek better and cleaner energy sources, deal with our garbage and keep things clean. But let's forget being a disciple of, "The sky is falling!" alarmist political rhetoric from your High priests and Pope Gore.

the only issue I'm aware that you keep flogging is something to do with your nonsensical claims of global cooling... but let's play! I really haven't a clue as to what thread you're talking about... what references you state you offered. Here's a thought! Resurrect the thread/post you'd like me to ignore respond to. But I tells ya, Pliny... you make reference to "late 90's" model "predictions"... if the going in position you presume to leverage relies on model to temperature observations for the 2000-2009 period, you're into the numnut category of attempting to legitimize trending results over the shortest of intervals - regardless of what the results show.

[Bonam]

1) Engineering models are validated with thousands of real life tests in the lab before they are ever used to make predictions. There are no lab experiments that can validate climate models.

No lab experiments that can validate climate models? Is your contention that we cannot know anything about the climate until the specific thing we want to know about has happened in our full scale "experiment" of Earth? If so, I disagree. Individual effects can be tested in small scale experiments, in the lab. The interactions between these effects can similarly be observed in the lab. Additionally, historical data exists for how these effects have played out in the past on Earth. All of this information can be used to construct models, whose accuracy continues to improve as our understanding improves.

2) Engineering models are not used outside of the range where they have been validated. i.e. an engineering model 'trained' with data over the last 100 years cannot be used to predict changes if key parameters go outside their ranges for the last 100 years. Climate models are routinely used to extrapolate way outside their 'training range' which means their output is largely meaningless.

Of course engineering models are used outside of ranges where they have been validated. Engineers routinely design novel systems that have not previously been designed, and such systems often go outside of the ranges of previous models. Whether it is aircraft that fly at higher speeds and thus higher mach numbers, outside of the range of previous models, or a variety of other phenomena, computational models are used first, and lead to preliminary designs which can then be tested. Such prototypes can cost millions of dollars, and the results from live testing of such prototypes often agrees closely with the predictions of the models used to design them. That's the whole point of models, once you've got the physical effects nailed down and incorporated into the model, you can use it to make predictions about a variety of situations which have not necessarily been physically tested before.

3) Engineering problems such jet engines and nuclear explosions are actually much simpler than climate because they do not need to incorporate the effects of poorly understood biological systems.

Some of these systems are simpler, others are not. I agree that many biological systems are poorly understood and certainly add substantial complexity and uncertainty. On the other hand, climate models benefit from the fact that the system they are describing is an equilibrium system, where even strong perturbations in factors cause only relatively small shifts of the equilibrium position, rather than wild swings, bifurcations, or chaotic behavior. In comparison, many engineering problems deal with chaotic systems or non-equilibrium systems, and these are inherently far more difficult to work with.

4) Engineering models are put through a formal V&V process. This ensures that the software actually does what the designers claim it does. Climate models are ad hoc developments by people who are not software engineers and have no concept of proper software test procedures.

This point is hardly relevant. Are you trying to say that climate models are not to be trusted in comparison to commercially used models because they don't go through a software testing process? Do you know how much buggy sh*t gets through software testing? Do you know what kind of people work on software testing? Chew on this: I software tested critical marine navigation software systems for Canadian navy surface vessels and submarines as a summer job after the first year of my undergrad degree.

5) Engineering models are run 1000s of times to generate a true probability density field for each set of model parameterizations. Climate models are run only a couple of times with the same parameterizations. To compensate they combined the outputs of different models with different parameterizations to create the PDF. This has the effect of making virtually any outcome 'consistent with' the models which makes the models impossible to falsify with real data.

If this is true, this is a problem with how the models were used, not with the models themselves. There is nothing stopping the climate models from being run thousands of times, with different sets of parameters, except perhaps availability supercomputer time. But again, that is not an issue with the models themselves.

IOW - I am not dismissing climate models out of hand. I am dismissing them because they do not live up to the standards that any aircraft/nuclear engineer would require before they would use a model to make any decisions with serious economic implications.

Climate models may not be perfect, and may not yet be as reliable as the types of models that engineers usually prefer to work with, but they nonetheless represent our best current understanding of how the climate may evolve in the future. Ignoring our best understanding, despite its current limitations, is foolhardy.

[TimG]

No lab experiments that can validate climate models? Is your contention that we cannot know anything about the climate until the specific thing we want to know about has happened in our full scale "experiment" of Earth?

I am saying is a computer model has limited use as a predictive tool unless it is used to make predictions and then validate those predictions in real life experiments. Weather models actually meet this criteria because their predictions are constantly validated against reality. It simply takes too long to check climate models against data. That is a problem unique to climate science that cannot be solved.

That's the whole point of models, once you've got the physical effects nailed down and incorporated into the model, you can use it to make predictions about a variety of situations which have not necessarily been physically tested before.

There are two types of models: physical models and parametrized models. Physical models that directly compute their results from first principles can be extremely reliable and can be used in novel situations. Parametrized models depend on unphysical approximations for key processes which means the models are only valid within the range where the parametrization have been validated. Climate models are parametrized models and cannot be compared to physical models.

Aside: did you know that there is no weather in climate model outputs? i.e. even though they have a time resolution in hours/days they do not produce the atmospheric patterns that we would associate with weather like storm fronts, cyclones etc. This is because climate models are unphysical approximations and do not actually solve the fundamental physics equations (a numerically impossible task).

On the other hand, climate models benefit from the fact that the system they are describing is an equilibrium system

Climate is NOT an equilibrium system. Every 24 hours the earth rotates from light to dark causing huge swings in temperatures. The blanket of clouds varies chaotically and depends on everything from temperature to cosmic rays to vegetation. The is no reason to believe the statistical properties of climate measured today are going to be the same if the planet warms by a degree. The claim that climate can be analyzed as a boundary value problem is simply an assumption that is disputed in the peer reviewed literature.

This point is hardly relevant. Are you trying to say that climate models are not to be trusted in comparison to commercially used models because they don't go through a software testing process?

Yes. Companies working in high risk industries have a lot at stake and they must be able to show in court that they have verified the correctness of their models. There is a lot at stake with these climate models - a;lot more than a submarine that gets lost. There is no excuse for not using the same processes that a building or nuclear engineer would use.

If this is true, this is a problem with how the models were used, not with the models themselves. There is nothing stopping the climate models from being run thousands of times, with different sets of parameters, except perhaps availability supercomputer time. But again, that is not an issue with the models themselves.

We don't have the computing resources to run climate models many times. This problems is not going to go away because as computing power increases climate modellers increase resolution instead of increasing the number of runs.

This tendency to use model ensembles turns out to be a perfect way to impose group-think on the field. i.e. someone who comes up with a new model has no way to prove the correctness of their model so the model is instead validated by its ability to mimic existing models.

they nonetheless represent our best current understanding of how the climate may evolve in the future.

Climate models have their place but they are no substitute for real data and cannot be used to provide evidence.

To illustrate: I can use the gravity equations to predict the speed of a ball I drop off a building. My calculated results represent a hypothesis. Verifying my hypothesis requires that I actually drop the ball and measure its speed. In this case, the equations have been verified millions of times so my hypothesis is likely correct - but it is still a hypothesis - not a fact.

Scientific studies that use computer models to make claims which have never/can never be verified with real physical measurements are hypotheses - they are not evidence or proof of anything. That is why a study that claims they can estimate the mass balance changes in 1958 from a few measurements of melt water cannot be considered evidence. We simply do not have the data that would allow us to validate the model being used.

Edited July 8, 2010 by TimG

[WIP]

Sure and I would rather err on the side of caution and not trash the economy by imposing an expensive and largely useless regime to monitor and reduce CO2 emissions.

So, economy trumps ecology on your list! According to your logic, if a robber points a gun at you and says:"your money or your life," you say "I want my money," and get your brains blown out! That's what you are advocating on the global scale with this greedy attitude that the economy must be our highest concern!

Almost all of those subsidies are subsidies to consumers in developing countries. The relevant figures:

And that makes it okay? The large multinational oil companies benefit from these subsidies also. BP's no.1 source of income is not in the Gulf of Mexico, or in the Middle East. It's in Russia, which provides a quarter of BP's oil...and that doesn't include their stake in Russian natural gas developments. Russian consumer subsidies are going directly into BP's pockets. Same thing goes for the others who are providing cheap gas.

The greedy bastards at the Financial Times restrict the original article to subscribers. The bloggers who publish excerpts from the IEA Report do not provide the full list of 37 large, developing nations. I'd like to see who else is on there.

The subsidies in rich countries are non existant and the subsidies in developing countries go to consumers - not oil companies.

Do you just repeat the same things over and over regardless of evidence? Here it is again from yesterday's post. These are the direct subsidies: For instance, the U.S. government has generally propped the industry up with:

* Construction bonds at low interest rates or tax-free

* Research-and-development programs at low or no cost

* Assuming the legal risks of exploration and development in a company's stead

* Below-cost loans with lenient repayment conditions

* Income tax breaks, especially featuring obscure provisions in tax laws designed to receive little congressional oversight when they expire

* Sales tax breaks - taxes on petroleum products are lower than average sales tax rates for other goods

....and all of the indirect subsidies:

* Giving money to international financial institutions (the U.S. has given tens of billions of dollars to the World Bank and U.S. Export-Import Bank to encourage oil production internationally, according to Friends of the Earth)

* The U.S. Strategic Petroleum Reserve

* Construction and protection of the nation's highway system

* Allowing the industry to pollute - what would oil cost if the industry had to pay to protect its shipments, and clean up its spills? If the environmental impact of burning petroleum were considered a cost? Or if it were held responsible for the particulate matter in people's lungs, in liability similar to that being asserted in the tobacco industry?

* Relaxing the amount of royalties to be paid (more below)

And even I can think of one major item overlooked on that list -- U.S. farm bills that subsidize factory farms and huge corn and soybean operations, are indirect subsidies of the oil industry, since the oil-based fertilizers (which make it all possible) have to be purchased by the growers.....so much for non-existent subsidies! Oil companies don't hire lobbyists and make large campaign donations for nothing.

So? Warming is not necessarily bad and even if there are some bad effects people can adapt to changes.

Someone already pointed out several pages ago how stupid this reasoning is, since the coming adaptation will mean mass migrations from overpopulated equatorial regions of the world, which are suffering the greatest impacts of climate change at the moment.....which is a major reason why the CIA has added climate change to its list of threats to national security. Apparently they don't trust the soothing words of the so called skeptics either! http://www.npr.org/templates/story/story.php?storyId=121352495

BS. Almost every person on this planet is better off than they would have been 400 years ago. Even peasent farmers in Africa benefit from modern technologies.

Got anything in the way of evidence to back that up with? Here in Canada, or most other developed nations, where poverty means at least having enough for a marginal existence, it's worth noting that almost one billion people in the world live on less than a dollar a day, and 80% of the world's population have to make it on less than $10 per day. Over 9 million people die each year because of hunger and malnutrition; and 5 million of them are children under 5 years. Another shocking stat: In 2005, the wealthiest 20% of the world accounted for 76.6% of total private consumption. The poorest fifth just 1.5%. http://www.globalissues.org/article/26/poverty-facts-and-stats

The prevailing economic system of disaster capitalism, which dominates most of the world, has improved the lives of many people, but many others, like that poorest 20% are still scratching for enough to eat, and dealing with sickness on an almost daily basis because of having dirty, contaminated drinking water. Far from "almost every person" being better off than 4 centuries ago, a large chunk of the world is likely worse off. And they likely don't give a rat's ass about the concerns of the wealthiest 20% who fear losing cheap and dirty energy supplies!

Specialization is what brings wealth. Eliminate that specialization and standard of living goes down.

Many farmers (my father would have been one if he were still alive) could tell you that specialization, especially stock-raising of animals, did nothing to improve farming. The drive towards stockraising and specialization of crops came from the meat processors and food wholesalers who wanted fewer, larger suppliers for their products, rather than dealing with a bunch of small farmers who had a few cows, a few pigs and chickens etc. They wanted agriculture set up in a manner that benefited their operations.

Considering that specialized agriculture is highly oil dependent...it is the source of 18% of the world's greenhouse gas production.... it is certainly not cheap when all of the costs are factored in.

Every corporation gets CCA deductions. Even a the hot dog vendor on the street corner. The only difference is some types of expenditures can be deducted faster than others but even if the speed of deduction is decreased companies are still entitled to the deduction! IOW - there are no lost tax revenues with CCA deduction and no subsidies.

According to the source above, they get exceptionally large write-offs from CCA's, although they are hopefully not as insane as the oil development subsidies offered during the 70's. The worst example was Canada's 130% CCA offered back then, which made drilling dry wells profitable. Back then, Dome Petroleum was turning profits even though they didn't find a drop of oil.

Also, it's stupid to compare the hotdog vendor, or even the smaller corporation with the oil company, since the rewards are going to go to the biggest. Of the ten largest and most profitable corporations in the world, Walmart, Toyota and ING Bank are the only non-oil companies on the list. http://money.cnn.com/magazines/fortune/global500/2009/full_list/ That should tell you something about the power and influence of this industry. Tobacco companies didn't want annoying lung cancer reports to cut in to their business, and likewise, the oil companies don't want an irritant like wrecking the environment, to get in the way of making money.

Exactly. We have succussfully adapted to virtually every climate on the planet. One can argue that climate change will likely cause disruptions to a largely stationary population but over time people will adapt to these disruptions and humanity will continue. The premise that humanity might become exintict is simply absurd.

You ought to read that article! Biodiversity does not make a comeback after an extinction until the source of the extinction is removed. Either we become a species that works in harmony with our environment, or we go extinct so that other life can make a comeback.

Or it could be a virus or parasite. The problem is we do not know and we should be focusing on these kinds of problems instead of CO2.

Yeah, nothing to worry about....except for a collapse in North American food production.

Why? There are 14,000 wells in the Gulf of Mexico alone. There was one blow out. In 10-15 years the ecosystems will recover and the experience will teach oil companies how to ensure such a disaster never happens again.

Yeah, one out of 14,000 isn't bad...even if it is already polluting every Gulf state and killing off an unknown amount of sea creatures in the Gulf. The reason we don't know is because government lackeys of the oil companies, including the Coast Guard, won't allow an investigation of the extent of the damage so far. If one blown wellhead can do this much damage so far, then that is still one too many, and drilling deep underwater is too risky to continue -- shut down the rest of them, including the ones off our shores!

Finally, there are way more than 14,000 wells in the Gulf. You are likely only counting the active ones. It seems the old wells that have been abandoned as far back as the 1940's, may be leaking, and no one is checking:

More than 27,000 abandoned oil and gas wells lurk in the hard rock beneath the Gulf of Mexico, an environmental minefield that has been ignored for decades. No one - not industry, not government - is checking to see if they are leaking, an Associated Press investigation shows.

The oldest of these wells were abandoned in the late 1940s, raising the prospect that many deteriorating sealing jobs are already failing. http://www.sott.net/articles/show/211720-27-000-Abandoned-Gulf-Oil-Wells-May-Be-Leaking

[WIP]

ya WIP, Simple has rarely (if ever) substantiated anything he's blindly parroted... in this case he pompously ignores the salient point concerning the most significant rise in atmospheric CO2 concentration. No biggee, hey Simple? For what it's worth, not to presume for you, but there is the oft referenced figure that reflects how fast CO2 emissions have grown over the last decade... 3% per year, on average, during 2000-2009 (notwithstanding the slowing recessionary influence within the 'western world' during that period). But ya, Simple would rather presume to play gotcha - cause he's got no real game!

Thanks! That's probably where I heard the 3% figure. Math is not my specialty, so I get mixed up on the difference between mean annual increase, and a figure measuring the acceleration in the rate of increase:

The sharp acceleration in CO2 emissions since 2000 to more than a 3% increase per year (more than 2 ppm per year) from 1.1% per year during the 1990s is attributable to the lapse of formerly declining trends in carbon intensity of both developing and developed nations.

http://en.wikipedia.org/wiki/Greenhouse_gas

Like I said previously, anyone who was honestly trying to find the truth would not consider this more important than alarming fact that CO2 rates are rising at an accelerating pace. This is similar to the nobs who think that three unpleasant emails at the CRU means that global warming is a hoax.

[TimG]

So, economy trumps ecology on your list!

Climate change is NOT an ecological issue. The earth's biosystem is quite resiliant and will have no problem adapting to whatever changes we could impose on it. Climate change is a human economic problem so what we do about it has to be driven by pragmatic economic considerations.

And that makes it okay? The large multinational oil companies benefit from these subsidies also.

No they don't. The governments buy the oil at market prices from the oil companies and sell the gas at a loss. The oil companies get nothing more than they would get if there were no subsidies (other than a slightly higher price due to higher demand). It is poor consumers in these countries that benefit from the subsidies.

The greedy bastards at the Financial Times restrict the original article to subscribers. The bloggers who publish excerpts from the IEA Report do not provide the full list of 37 large, developing nations. I'd like to see who else is on there.

The key word is "developing" nations. Canada and the US are not on the list.

Do you just repeat the same things over and over regardless of evidence? Here it is again from yesterday's post.

Unless you can quantify them they don't mean much (i.e. one few million in low cost loans is a drop in the bucket). All of the quantative analyses I have seen claim the majority of 'subsidies' in rich countries comes from tax credits which are not really subsidies.

Someone already pointed out several pages ago how stupid this reasoning is, since the coming adaptation will mean mass migrations from overpopulated equatorial regions of the world, which are suffering the greatest impacts of climate change at the moment.

Actually, there is no evidence of anyone suffering 'impacts' of climate change. What we have is a scientific culture that simply assumes that any bad thing must be due to climate change even though they have zero evidence. Here is a peer reviewed paper that analyzes patterns of drought in the world. The conclusions are there is zero evidence of a trend but droughts tend to affect different regions at different times. i.e. droughts in Africa are part of a natural cycle where droughts shift around the world. There is no justification for attributing recent droughts in Africa to climate change. Similar issues exist with virtually every other claim 'climate impact'.

Got anything in the way of evidence to back that up with?

400 years ago people simply died if there was a famine or natural disaster. There was no global infrastrature that allowed food and supplies to be quickly moved to areas where they are needed.

Over 9 million people die each year because of hunger and malnutrition; and 5 million of them are children under 5 years.

In percentage terms those numbers are lower than they ever have been in the history of man.

Another shocking stat: In 2005, the wealthiest 20% of the world accounted for 76.6% of total private consumption. The poorest fifth just 1.5%.

The disparity was even worse back in the days when most economies depended on slave labour.

Far from "almost every person" being better off than 4 centuries ago, a large chunk of the world is likely worse off.

So in your world no one died of starvation or disease 400 years ago? Forgive me that is quiet implausible.

According to the source above, they get exceptionally large write-offs from CCA's, although they are hopefully not as insane as the oil development subsidies offered during the 70's. The worst example was Canada's 130% CCA offered back then, which made drilling dry wells profitable. Back then, Dome Petroleum was turning profits even though they didn't find a drop of oil.

A 130% CCA is a subsidy. But that is not what is happening today. CCA < 100% - no matter how fast the write off occurs - does not change the total amount tax collected from companies. It simply encourages companies to spend money on capital instead of paying it in dividends.

Also, it's stupid to compare the hotdog vendor, or even the smaller corporation with the oil company, since the rewards are going to go to the biggest.

BS. CCA is available to all companies and cannot be singled out as 'fossil fuel' subsidy

Yeah, nothing to worry about....except for a collapse in North American food production.

Exactly. The honeybee is a tangible verifiable problem that should be addressed directly. We need to find out what is causing it and find some way to deal with it. Spending money on CO2 reduction takes money away that could be better used on real problems like honeybees.

Yeah, one out of 14,000 isn't bad...even if it is already polluting every Gulf state and killing off an unknown amount of sea creatures in the Gulf.

A problem made worse by incompentent governments who prevented deployment of foreign clean up equipment.

If one blown wellhead can do this much damage so far, then that is still one too many, and drilling deep underwater is too risky to continue -- shut down the rest of them, including the ones off our shores!

So you must be a person that believes that cars must be banned because of thousands of a people killed every year driving in them? Life is full of risk. We need the oil. There are no plausible alternatives. The US can drill in their own waters or outsource the pollution/risk to other countries.

[waldo]

in actuality, you’re referencing back to the TAR era report where a sensitivity of 1.5°C was within the ‘likely’ 1.5°C-to-4.5°C range… but, again, at the low-end. With advances between TAR and AR4, the ‘likely’ range is now expressed as 2.0°C-to-4.5°C. So, your choosing a sensitivity of 1.5°C is quite telling. Notwithstanding, of course, that you’re also prepared to reference the IPCC report, yet ignore it’s quoted best estimate value of a sensitivity of ~3°C. As for the assigned probabilities, IPCC guidelines advise that, ‘Likelihood may be based on quantitative analysis or an elicitation of expert views’.

As I said. The probabilities assigned in the IPCC documents are subjective guesses that may or may not have an connection with reality. What is more important is the range of plausible values. From the AR4:

So 1.5 is still within the range of plausible values even though they did not assign a subjective probability to the 1.5-2.0 range.

yes, the AR4 quote reflects directly upon the exact figures I've been stating over these latest posts... highlighting that your selection of the lowest sensitivity figure that exists (1.5°C) doesn't fit within an assigned probability range (i.e. the ‘likely’ range is now expressed as 2.0°C-to-4.5°C)... highlighting that you are accepting to reference the IPCC sensitivity values, yet not willing to at the same time accept that sensitivity value described as the "best estimate value" (~3°C). Again, your selective choice is most telling. As for your emphasis on 'plausible' values, the only basis I recall for your substantiating your selection of the most optimistic, lowest sensitivity value of 1.5°C, is your reference to realizing a, as you stated, "maximum temperature rise of 1.5-2.0 degC under the business as usual scenario". I asked you just what scenario that was... presuming to move us off the discussion of equilibrium sensitivity (as reflects upon sustained forcing and the doubling of CO2), and into a discussion of actual temperature projections. However, you did not answer my request for you to qualify your reference to a "business as usual" scenario.....

and, again, the probabilities have an element of subjectivity; however, to overtly label them as, as you say, "guesses", does not reflect upon the specific IPCC guideline concerning assigning probabilities... the guideline that states, as I related, "Likelihood (probabilities) may be based on quantitative analysis and/or an elicitation of expert views".

You also need to remember that the 2degC limit was plucked out of hat so it is not a 'hard' limit.

As for the 2°C limit, you most certainly know it wasn’t, as you suggestively denigrate it, “plucked out of a hat”… you may not like or agree in how it was arrived at, but it most certainly has a foundation in terms of policy, vis-a vis 2000-to-2050 CO2 emission reduction, vis-à-vis comparisons to pre-industrial times.

This topic came up when you asked if I had any scientific basis for believing that no action is required. I demonstrated that basis which depends on where reality falls within the uncertainties set out by the IPCC. The big question is how much weight does one put on purely subjective estimates of probability by the IPCC. I personally put little weight on them because there is no objective scientific justification for the numbers.

again, assigned likelihood (probabilities), per IPCC guideline, may be based on quantitative analysis and/or an elicitation of expert views. In any case, yes, the IPCC did assign probabilities to CO2ppm levels in regards to realizing the 2°C target limit… based on the analysis/probabilities offered within many related studies. Equally, separate from the IPCC proper, meta-studies have been undertaken to analyze across the breadth of available studies; these, similarly, have brought forward CO2ppm related probabilities in realizing the 2°C target limit. In both these instances, IPCC and separate meta-studies, the probability assignments were not, as you suggested, “purely subjective estimates”.

In terms of the post-AR4 study I linked to to (‘Mass balance of the Greenland ice sheet from 1958 to 2007’), when you say, “ignored that data as well”… well, of course, that post-AR4 study wasn’t included (so couldn’t have been ignored) – it occurred post-AR4! It’s a study that meets your >30 year climatic significance. You simply choose to ignore/discount it.

Because there is no reliable data on ice sheet mass prior to the GRACE satellite. If the paper is like others I am seen it probably uses a computer model to estimate the mass balance prior to GRACE. I do not consider estimates with computer models to be evidence of anything.

yes – the 1958-2007 period surface mass balance study was based on meteorological models… separate independent sets of time series were created to place the remotely sensed results, (the ~20 years of data the IPCC acknowledges… and that you suggest isn’t a sufficient amount of time within a climatic reference time frame), into a longer, multi-decadal, climatic view. So you don’t accept models; well - bully for you! But now you also want to discount that ~20 years worth of airborne/satellite laser-altimetry and satellite radar interferometry data analyzes? Well – bully for you!

so… we have that ~20 years of observational laser-altimetry & radar interferometry data, since 2001 we have the GRACE satellite targeting mass balance analysis, we have/had the ICESat laser-altimetry results (2003-2010), we have, since 2002, the GRACE satellite now focused on, from a climate change perspective, mass calculations based on gravitational measurements… and in Greenland GRACE coupled with the land-based GPS network to provide precise mapping capabilities. All of this… collectively showing and reinforcing Greenland ice-sheet melting… Greenland mass loss – accelerated in recent years. But, in your apparent, rather cavalier, laissez-faire approach, you would prefer to wait for... what... another 15 years of data? Well - bully for you!

here… just another recent study to discount - hey? British Antarctic Survey – Lasers from space show thinning of Greenland and Antarctic ice sheets

The most comprehensive picture of the rapidly thinning glaciers along the coastline of both the Antarctic and Greenland ice sheets has been created using satellite lasers. The findings are an important step forward in the quest to make more accurate predictions for future sea level rise.

Reporting this week in the journal Nature researchers from British Antarctic Survey and the University of Bristol describe how analysis of millions of NASA satellite measurements* from both of these vast ice sheets shows that the most profound ice loss is a result of glaciers speeding up where they flow into the sea.

The authors conclude that this ‘dynamic thinning’ of glaciers now reaches all latitudes in Greenland, has intensified on key Antarctic coastlines, is penetrating far into the ice sheets’ interior and is spreading as ice shelves thin by ocean-driven melt. Ice shelf collapse has triggered particularly strong thinning that has endured for decades.

Extensive dynamic thinning on the margins of the Greenland and Antarctic ice sheets

[TimG]

the probabilities have an element of subjectivity; however, to overtly label them as, as you say, "guesses", does not reflect upon the specific IPCC guideline concerning assigning probabilities. the guideline that states, as I related, "Likelihood (probabilities) may be based on quantitative analysis and/or an elicitation of expert views".

Guesses by experts are still guesses. That said, guesses by experts are not necessarily useless - as long as one remembers they are guesses and subject to the biases of the experts making the guess. When it comes to the IPCC, the authors were quite keen on having a 'tidy story' which promoted the Kyoto II political agenda so we pretty much have to assume the guesses are biased in a way that supports that agenda.

yes – the 1958-2007 period surface mass balance study was based on meteorological models.

The starting point of 1958 is an interesting coincidence because it is the minimum melt between the peaks of 1930-1950 and 1990-2010. It looks like they cherry picked their starting point to give them a maximally scarey trend.

It also appears that the melt rate is already slowing down:

So much for Greenland ice’s Armageddon. “It has come to an end,” glaciologist Tavi Murray of Swansea University in the United Kingdom said during a session at the meeting. “There seems to have been a synchronous switch-off” of the speed-up, she said. Nearly everywhere around southeast Greenland, outlet glacier flows have returned to the levels of 2000. An increasingly warmer climate will no doubt eat away at the Greenland ice sheet for centuries, glaciologists say, but no one should be extrapolating the ice’s recent wild behavior into the future

[waldo]

and, again, the probabilities have an element of subjectivity; however, to overtly label them as, as you say, "guesses", does not reflect upon the specific IPCC guideline concerning assigning probabilities... the guideline that states, as I related, "Likelihood (probabilities) may be based on quantitative analysis and/or an elicitation of expert views".

Guesses by experts are still guesses. That said, guesses by experts are not necessarily useless - as long as one remembers they are guesses and subject to the biases of the experts making the guess. When it comes to the IPCC, the authors were quite keen on having a 'tidy story' which promoted the Kyoto II political agenda so we pretty much have to assume the guesses are biased in a way that supports that agenda.

ya, ya... putting aside your conspiratorial musings, even if you discount so-called "expert views" (guesses, in your determination), why do you totally discount quantitative analysis, as the other IPCC guideline aspect to be used in assigning likelihood (probability)?

yes – the 1958-2007 period surface mass balance study was based on meteorological models… separate independent sets of time series were created to place the remotely sensed results, (the ~20 years of data the IPCC acknowledges… and that you suggest isn’t a sufficient amount of time within a climatic reference time frame), into a longer, multi-decadal, climatic view. So you don’t accept models; well - bully for you! But now you also want to discount that ~20 years worth of airborne/satellite laser-altimetry and satellite radar interferometry data analyzes? Well – bully for you!

The starting point of 1958 is an interesting coincidence because it is the minimum melt between the peaks of 1930-1950 and 1990-2010. It looks like they cherry picked their starting point to give them a maximally scarey trend.

It also appears that the melt rate is already slowing down

interesting... so far you've linked to the skeptic/denier blogs, climatefraudit and, at least a couple of times now (including this your latest reply), the Cato Institute Pat Michaels', WCR site. And you now offer up a "presentation" from Knappenberger/Michaels. By the way, I always got a slight chuckle reading the signature closing of "Chip" Knappenberger's RC commentary... the one written as "Chip Knappenberger - to some degree, funded by the fossil fuels industry since 1992". At least in that specific case, he is forthright, if nothing else! But, uhhh... by the by... I've linked you to several related peer-review study articles from respected journals, with at least one that post-dates your linked Knappenberger/Michaels "presentation". That "presentation" analysis would have been around now for close to a couple of years, or so... one wonders why they haven't bothered to (or been able to) get it actually published - hey?

[bloodyminded]

[ That "presentation" analysis would have been around now for close to a couple of years, or so... one wonders why they haven't bothered to (or been able to) get it actually published - hey?

My understanding is that peer-reviewed publishing (aside from the infallibility of right-wing economists) is in and of itself proof of perfidy and dishonesty. The tautology is perfect, and wondrous.

[TimG]

why do you totally discount quantitative analysis, as the other IPCC guideline aspect to be used in assigning likelihood (probability)?

Because the are not any 'quantitative analysis' techniques that are not also extremely subjective. Bayesian analysis is a common technique but its results depend entirely on the choice of priors which is a subjective decision. I really get the impression that you know how to quote the lingo but have little understanding of what it means.

And you now offer up a "presentation" from Knappenberger/Michaels.

I gave to a link to a presentation at an AGU meeting and a quote from report on meeting that was printed in Science Magazine. The sources are not peer reviewed but the onus is really on you to show that there is something factually wrong about claims. For example, the comment about the minimum is something I have seen in other sources. Using 1958 at a start date is definitely a cherry pick just like choosing 1998 as start date is a cherry pick. Edited July 8, 2010 by TimG

[waldo]

ya, ya... putting aside your conspiratorial musings, even if you discount so-called "expert views" (guesses, in your determination), why do you totally discount quantitative analysis, as the other IPCC guideline aspect to be used in assigning likelihood (probability)?

Because the are not any 'quantitative analysis' techniques that are not also extremely subjective. Bayesian analysis is a common technique but its results depend entirely on the choice of priors which is a subjective decision. I really get the impression that you know how to quote the lingo but have little understanding of what it means.

interesting… you started out belittling IPCC likelihood (probability) assignments by labelling them as “guesses”. In terms of ‘expert views’, one of the IPCC guideline associations available to assist in assigning a likelihood (probability), you maintained your position by stating, as you said, “Guesses by experts are still guesses”. You then shifted slightly by speaking to bias and subjectivity. In your assessment, you’re not willing to accept the expert view likelihood (probability) assignment as an informed knowledge/choice (biased, or not)… you simply choose to denigrate it all as “guess work”!

in terms of ‘quantitative analysis’, the other of the IPCC guideline associations available to assist in assigning a likelihood (probability), you make a blanket statement that categorizes all quantitative analysis as, “extremely subjective”… you then proceed to speak of Bayesian analysis and prior distribution choice… and throw in a sprinkling of negative personalization (was it something I said?). Now, in my somewhat dated statistical training, I seem to recollect a strong emphasis on attempting to exclude (or at least limit) said bias/subjectivity in prior distributions… if you adhere to the subjectivist school of Bayesian analysis. On the other hand, the issue of prior distribution bias/subjectivity doesn’t come forward… if you adhere to the objectivist school of Bayesian analysis (and objective prior distributions)… a view of Bayesian analysis you seem unfamiliar with. As you said, “I really get the impression that you know how to quote the lingo but have little understanding of what it means”.

now, since our emphasis is, of course, climate change discussion, an easy search pops up a nice example study reference of Bayesian analysis and objective prior distribution… Objective Probabilistic Forecasts of Future Climate Based on Jeffreys' Prior: the Case of Correlated Observables

To include parameter uncertainty into probabilistic climate forecasts one must first specify a prior. We advocate the use of objective priors, and, in particular, the Jeffreys' Prior. In previous work we have derived expressions for the Jeffreys' Prior for the case in which the observations are independent and normally distributed. These expressions make the calculation of the prior much simpler than evaluation directly from the definition. In this paper, we now relax the independence assumption and derive expressions for the Jeffreys' Prior for the case in which the observations are distributed with a multivariate normal distribution with constant covariances. Again, these expressions simplify the calculation of the prior: in this case they reduce it to the calculation of the differences between the ensemble means of climate model ensembles based on different parameter settings. These calculations are simple enough to be applied to even the most complex climate models.

in any case, the real point of discussion was your belittling categorization of IPCC likelihood (probability) assignments… as guesses (guess work), reflected upon the informed knowledge/choice of “expert views”, or as subjective/bias, reflected upon “quantitative analysis” (regardless of whatever attempts may exist to eliminate/reduce subjectivity and bias). Your Bayesian analysis example, and my decision to belabour the distinction between the two schools/views of Bayesian analysis, was simply a point of departure… of diversion. Was it good… for you?

interesting... so far you've linked to the skeptic/denier blogs, climatefraudit and, at least a couple of times now (including this your latest reply), the Cato Institute Pat Michaels', WCR site. And you now offer up a "presentation" from Knappenberger/Michaels. By the way, I always got a slight chuckle reading the signature closing of "Chip" Knappenberger's RC commentary... the one written as "Chip Knappenberger - to some degree, funded by the fossil fuels industry since 1992". At least in that specific case, he is forthright, if nothing else! But, uhhh... by the by... I've linked you to several related peer-review study articles from respected journals, with at least one that post-dates your linked Knappenberger/Michaels "presentation". That "presentation" analysis would have been around now for close to a couple of years, or so... one wonders why they haven't bothered to (or been able to) get it actually published - hey?

I gave to a link to a presentation at an AGU meeting and a quote from report on meeting that was printed in Science Magazine. The sources are not peer reviewed but the onus is really on you to show that there is something factually wrong about claims. For example, the comment about the minimum is something I have seen in other sources. Using 1958 at a start date is definitely a cherry pick just like choosing 1998 as start date is a cherry pick.

that’s rich! At what point do skeptic/deniers take onus in actually substantiating their claims… in your case, you simply cough up a link to a “presentation” and a single paragraph description of it… your actual link is embedded within the words, “minimum melt”, and your linked one paragraph description has nothing to say about, “minimum melt”. Without substantiation, you proceed to speak of cherry-picking intended to emphasize, as you stated, “a maximally scary trend”.

you admit your sources are not peer-reviewed, yet you state the onus is on “me” to wrestle with your one paragraph description of a “presentation” and, as you say, “show that there is something factually wrong about claims”. Oh really!

of course, you fail to offer comment on my questioning why the Knappenberger/Michaels presentation (their reconstruction, analysis and findings) wasn’t published… not even in E&E! It would seem the emphasis placed on it, the emphasis you place on it, would certainly indicate it has much to offer. Really now, one wonders why such dramatic and significant findings were never published… is there a problem with them? You speak of onus – is there any onus on you to answer the questions on why their findings weren’t published… on why you should expect someone to scurry about to challenge the unsubstantiated/unpublished claims of avowed skeptic/deniers. Of course, you could also choose to counter the peer-review studies I’ve linked to… while you bring forward something/anything peer-reviewed, that can attest to the denier blog statements you directly quoted… the ones that indicate, “the recent speed-up of Greenland’s glaciers has even more recently slowed down”… that “Nearly everywhere around southeast Greenland, outlet glacier flows have returned to the levels of 2000.”… that, “but no one should be extrapolating the ice’s recent wild behavior into the future”. You could do that… you could do all that… since the onus is on you.

[waldo]

My understanding is that peer-reviewed publishing (aside from the infallibility of right-wing economists) is in and of itself proof of perfidy and dishonesty. The tautology is perfect, and wondrous.

oh snap!

[WIP]

Climate change is NOT an ecological issue. The earth's biosystem is quite resiliant and will have no problem adapting to whatever changes we could impose on it.

Now that's got to be the stupidest statement I've read in some time! If the biosphere has to adapt to climate change, as mentioned in sentence two, that in itself, makes climate change an ecological issue. Do you ever read your broad, sweeping, unsourced statements after you write them down?

And that 2nd one:

"will have no problem adapting to whatever changes we could impose on it."

What !@#$^&*( evidence do you have to back up that kind of claim? That belief that no matter what kind of mess we make of the environment, everything will be okay -- is a faith position, presented by Dominionists, who insist that the Earth is an infinite, unbounded resource, in spite of the obvious evidence for the opposite -- the Earth is a finite resource that cannot accommodate every mess created by an exponentially growing human population.

Climate change is a human economic problem so what we do about it has to be driven by pragmatic economic considerations.

The economic problems from changing climate are a secondary problem. You must think that we are separate and above the plants and animals of this world -- but we're not! Loss of biodiversity will have a determining effect on how many humans this planet can support.

No they don't. The governments buy the oil at market prices from the oil companies and sell the gas at a loss. The oil companies get nothing more than they would get if there were no subsidies (other than a slightly higher price due to higher demand). It is poor consumers in these countries that benefit from the subsidies.

How do these "poor consumers" benefit from those subsidies? They buy gasoline or other oil products. And yet you think the oil company selling the product, which has their product subsidized by a government, is not gaining any benefit! Tell me you don't actually believe this! Do you work for an oil company, or are you heavily invested in oil stocks? Or do you get some kind of kick by supplying propaganda free of charge to the greediest, most ruthless and dishonest businessmen in the world.

Unless you can quantify them they don't mean much (i.e. one few million in low cost loans is a drop in the bucket). All of the quantative analyses I have seen claim the majority of 'subsidies' in rich countries comes from tax credits which are not really subsidies.

If it means they get to keep money they would have otherwise payed in taxes, then it is the same damn thing!

Actually, there is no evidence of anyone suffering 'impacts' of climate change. What we have is a scientific culture that simply assumes that any bad thing must be due to climate change even though they have zero evidence. Here is a peer reviewed paper that analyzes patterns of drought in the world.

I think I am starting to understand why Waldo blows his top every time he looks at some crap that you refer to as evidence!

1. First, the generalized unverifiable statement: "there is no evidence of anyone suffering 'impacts' of climate change."

2. Then it's the counter-evidence that comes from a dubious source: "World Climate Report." And who produces this high traffic climate blog that is presented as serious, unbiased climate research? Why it's the Western Fuels Association, which supplies coal for coal-fired power stations across the western states and the mid-west....yes, we can trust a coal supplier to take an honest, unbiased look at anthropogenic climate change! This is the environment version of creationist blogs like Evolution News and Views, which tries to fool readers into thinking it is studying evolution.

3. The dodgy evidence from the dubious source: your "peer reviewed paper" doesn't even claim it is from a peer reviewed source, and it only analyzes conditions up till the year 2000. That's ten years ago buddy! A lot has changed since then. Just go outside and take a look at the weather for christs sake! And they were only analyzing one aspect of climate change -- drought activity. What about severe storm activity, extreme heat, extreme cold etc.

400 years ago people simply died if there was a famine or natural disaster. There was no global infrastrature that allowed food and supplies to be quickly moved to areas where they are needed.

The global infrastructure you refer to doesn't even do an adequate job in Haiti, let alone places like Western Sudan. It just doesn't sink in that the poorest quarter of the population is not benefiting from the wonders of modern technology!

Blah Blah Blah and......

So you must be a person that believes that cars must be banned because of thousands of a people killed every year driving in them? Life is full of risk. We need the oil. There are no plausible alternatives. The US can drill in their own waters or outsource the pollution/risk to other countries.

Put it this way -- we did without oil until about two hundred years ago, you act like it's the elixir of life. If it's a matter of survival, we do without oil again! That's what we have to do anyway. In case you weren't aware, oil is a limited resource, it does not magically replenish itself underground. The reason why our environmental situation is getting so dire, is because rather than start the transition to the inevitable post-oil age, the enormously powerful and influential oil industry is using all of its clout to double down, and extract the dirtiest and most dangerous sources of black gold. That's why there is a growing ecological disaster in the Gulf of Mexico. Instead of bypassing high risk deep sea sources of oil, they are being developed to create ecological disasters and make the AGW problem even worse.

[TimG]

In your assessment, you’re not willing to accept the expert view likelihood (probability) assignment as an informed knowledge/choice (biased, or not)… you simply choose to denigrate it all as “guess work”!

Because that is what it is.

You want a concrete example look at climate sensitivity. We have no idea what the actual sensitivity is nor do we even know if it is a constant (i.e. sensitivity could be higher coming out of an ice age than it is today). Also sensitivity is something that can never be measured directly - it is something which can only be inferred by using a model that assigns values to many different unknown quantities (aerosols, cloud cover, et. al.). So what the IPCC did is assume that all estimates of CO2 sensitivity represented independent measurements and that CO2 sensitivity is a constant. It then constructed a PDF based on the range of values for these various methods. The trouble is this PDF could be completely wrong if any of their assumption are wrong. In this case the assumption that each estimates independent measurements is likely wrong because researchers estimates for unknown values like aerosols are affected by the expected result and this introduces a bias into the estimate. This bias is likely why the estimates of CO2 sensitivity have not changed for 30 years (1.5-4.5).

In fact there is a historical precedent for the kind of bias I describe with the estimates of electron charge which were originally conducted by Millikan. His value was wrong but it took a long time to discover this because scientists that replicated the experiment assumed that Millikan was right and adjusted their numbers to ensure a better match with Millikan. In an field were scientists that suggest a low CO2 sensitivity are immediately attacked at stooges of 'fossil fuel companies' it is not reasonable to claim that the exact same bias is not at work.

What this all means is the the 'quantitative' probabilities don't mean much. What matters is the range of plausible values. That is why I say the planet will warm due to CO2 and the only question is by how much.

Edited July 9, 2010 by TimG

[Bonam]

I am saying is a computer model has limited use as a predictive tool unless it is used to make predictions and then validate those predictions in real life experiments. Weather models actually meet this criteria because their predictions are constantly validated against reality. It simply takes too long to check climate models against data. That is a problem unique to climate science that cannot be solved.

We have thousands of years of climate data that we can check models against. While such data is less detailed and reliable than recent data when direct climate records have been kept, we nevertheless a good idea of the Earth's temperatures, atmospheric gas concentrations, etc for a long enough time period to test models on.

There are two types of models: physical models and parametrized models. Physical models that directly compute their results from first principles can be extremely reliable and can be used in novel situations. Parametrized models depend on unphysical approximations for key processes which means the models are only valid within the range where the parametrization have been validated. Climate models are parametrized models and cannot be compared to physical models.

Different models use approximations of differing physical validity. No models (except those modeling the interactions of just a few individual atoms) directly solve everything from first physical principles, because it is numerically impossible. As long as the approximations used are in good agreement with data and are applied over their range of validity, they are appropriate for use in these models. Using models appropriately is fundamental to doing good science, and any scientist that published a paper where models are used far outside their regions of validity without noting this and mentioning that it introduces large uncertainties on the results would be shot down in peer review or in opposing papers very quickly.

Aside: did you know that there is no weather in climate model outputs? i.e. even though they have a time resolution in hours/days they do not produce the atmospheric patterns that we would associate with weather like storm fronts, cyclones etc. This is because climate models are unphysical approximations and do not actually solve the fundamental physics equations (a numerically impossible task).

Yes I am aware that most climate models do not include weather. This is because weather is a chaotic phenomenon, whereas as I mentioned before, climate is an equilibrium one. You cannot model weather with these kinds of models.

Climate is NOT an equilibrium system. Every 24 hours the earth rotates from light to dark causing huge swings in temperatures. The blanket of clouds varies chaotically and depends on everything from temperature to cosmic rays to vegetation. The is no reason to believe the statistical properties of climate measured today are going to be the same if the planet warms by a degree. The claim that climate can be analyzed as a boundary value problem is simply an assumption that is disputed in the peer reviewed literature.

Of course climate is an equilibrium system. Day to day temperature variation is not climate. Average temperature over longer time periods is climate. Blankets of clouds behave vary chaotically, and yet the Earth's total % cloud cover averaged over long periods is not chaotic. The movement of particles in a gas is extremely chaotic when you look at them one by one, but the overall effect is very orderly, to produce a uniform pressure on the vessel in which the gas is contained. As for whether climate can be analyzed as a boundary value problem, well, it has boundaries, and the various parameters in question have values defined along those boundaries, and if the phenomena taking place within the volume defined by those boundaries can be described by a closed system of solvable equations, then it can be treated as a boundary value problem. Since you used the term "boundary value problem" I'll assume I can use some mathematical terminology here... Given that the Earth's climate exists and is unique, and the boundaries are physical and are defined, in itself shows that one can define such a boundary value problem, and that its solution exists and is unique. The challenge is simply to define the boundaries, parameters, and equations describing it all.

It is not necessary to show well-posedness, existence, and uniqueness in the same manner that a mathematician would like to do. For example, mathematicians still have been unable to prove the existence and uniqueness of solutions of the Navier-Stokes equations, and yet these are solved routinely and very accurately using numerical methods.

Yes. Companies working in high risk industries have a lot at stake and they must be able to show in court that they have verified the correctness of their models. There is a lot at stake with these climate models - a;lot more than a submarine that gets lost. There is no excuse for not using the same processes that a building or nuclear engineer would use.

And yet there are some tasks that cannot be tested in advance with such high standards of certainty, when nevertheless people must sometimes make decisions. We do not always have the luxury of perfect information.

We don't have the computing resources to run climate models many times. This problems is not going to go away because as computing power increases climate modellers increase resolution instead of increasing the number of runs.

Of course it will go away, computational abilities continue to increase at exponentially accelerating rates. Numerical simulations thought impossible 10-20 years ago can now be easily carried out. Direct numerical simulation of the Earth's entire ocean and atmosphere as one system resolved at the Kolmogorov timescales and lengthscales will be possible within a decade or two if the current trends in computational advance continue.

This tendency to use model ensembles turns out to be a perfect way to impose group-think on the field. i.e. someone who comes up with a new model has no way to prove the correctness of their model so the model is instead validated by its ability to mimic existing models.

As I said before, climate scientists also have about a century of careful climate data and thousands of years of less detailed data to check their models against.

Climate models have their place but they are no substitute for real data and cannot be used to provide evidence.

Given that we have no direct and incontrovertible knowledge of the future until it happens, climate models are our best tool to understand and predict what the nature of that future will be. Regardless of their imperfections, they are the best evidence we have.

To illustrate: I can use the gravity equations to predict the speed of a ball I drop off a building. My calculated results represent a hypothesis. Verifying my hypothesis requires that I actually drop the ball and measure its speed. In this case, the equations have been verified millions of times so my hypothesis is likely correct - but it is still a hypothesis - not a fact.

At a certain point it's not really a hypothesis anymore, but a "law", Newton's Law of Universal Gravitation in this case. A semantic point perhaps, since opposing evidence can still of course falsify it, but it is important to be able to make a distinction between theories that have been tested and hold true beyond all reasonable doubt (in their regions of validity) and hypotheses that have yet to be put through rigorous testing.

Scientific studies that use computer models to make claims which have never/can never be verified with real physical measurements are hypotheses - they are not evidence or proof of anything. That is why a study that claims they can estimate the mass balance changes in 1958 from a few measurements of melt water cannot be considered evidence. We simply do not have the data that would allow us to validate the model being used.

They are not proof, no, but they are evidence, and the better the models the stronger such evidence is. We have no evidence of the future besides the predictions made by such models.

Edited July 9, 2010 by Bonam

[TimG]

We have thousands of years of climate data that we can check models against. While such data is less detailed and reliable than recent data when direct climate records have been kept, we nevertheless a good idea of the Earth's temperatures, atmospheric gas concentrations, etc for a long enough time period to test models on.

So where is the data on the distribution/type of clouds during the last ice age? You won't find yet it is impossible to reproduce the climate without this data. Climate modellers just make it up. More importantly there are a large number of similar knobs which can be fiddled which means their models could be completely bogus yet they still would be able to fit past temperatures because of these adjustable data points.

Different models use approximations of differing physical validity. No models (except those modeling the interactions of just a few individual atoms) directly solve everything from first physical principles, because it is numerically impossible.

Wrong. A model based on newton's gravitational equations can be solved numerically. There are many similar problems where the mathematical equations can be solved which means the models are 'physics based'. These models are only limited by the range of validity for the equations used. Climate models are not like that. They depend on approximations for complex processes which means the models are only valid for the range of values where these approximations have been validated with real data.

Yes I am aware that most climate models do not include weather. This is because weather is a chaotic phenomenon, whereas as I mentioned before, climate is an equilibrium one. You cannot model weather with these kinds of models.

There is a huge difference between predicting weather and replicating its behavoir. Climate is the running average of weather so any 'physics based' computer model must reproduce weather like phenomena. If it does not it is a non-physical model which is subject to the limits imposed by the training data. The lack of weather related phenomena proves that climate models are non-physical no matter what claims are made by their makers.

Of course climate is an equilibrium system. Day to day temperature variation is not climate. Average temperature over longer time periods is climate. Blankets of clouds behave vary chaotically, and yet the Earth's total % cloud cover averaged over long periods is not chaotic.

Again completely wrong. The average cloud cover does vary over decadal timescales and there is no evidence that it is constant or predictable. We do not have enough data to properly characterize its behavoir.

It is not necessary to show well-posedness, existence, and uniqueness in the same manner that a mathematician would like to do. For example, mathematicians still have been unable to prove the existence and uniqueness of solutions of the Navier-Stokes equations, and yet these are solved routinely and very accurately using numerical methods.

Here is a short commentary by a respected expert in the field numerical modelling. He explains why climate modellers are misrepresenting the reliability of their models:

Convergent numerical solutions have shown that the estimates of this scale are extremely accurate. If the numerical model does not resolve the correct number of waves indicated by the estimate, the model blows up. If the model resolves the number of waves indicated by the estimate, the numerical method will converge to the continuum solution for long periods of time. Thus, if a numerical model is unable to resolve the spectrum of the continuum solution, the model is forced to artificially increase the viscosity coefficient or use a numerical method that has nonphysical viscosity built into the method.

...

The updating discussed above is not possible in a climate model and because climate models use even a coarser mesh than a large-scale weather prediction model, they must use an effectively larger viscosity than a global weather prediction model. Recently (BAMS, 2004), it has been shown that a climate model also deviates from reality in a matter of hours because of the errors in the parameterizations (not unexpected based on result above) and over longer periods of time the effectively larger viscosity causes the numerical solution to produce a spectrum quite different than the real atmosphere unless forced in a nonphysical manner.

Given that we have no direct and incontrovertible knowledge of the future until it happens, climate models are our best tool to understand and predict what the nature of that future will be. Regardless of their imperfections, they are the best evidence we have.

They are NOT evidence. They a hypothesis that needs to be confirmed by real data before it can be treated as a fact. The 'only tool we have' argument does not cut it either. Tools have limitations it is wrong to ignore these limitations.

Perhaps the biggest problem is climate science is too many scientists are not willing to simple admit we don't know the answers.

Edited July 10, 2010 by TimG

[Bonam]

They are NOT evidence. They a hypothesis that needs to be confirmed by real data before it can be treated as a fact. The 'only tool we have' argument does not cut it either. Tools have limitations it is wrong to ignore these limitations.

Perhaps the biggest problem is climate science is too many scientists are not willing to simple admit we don't know the answers.

I understand the limitations of the models, and actually I appreciate being able to debate with someone who can speak on a more scientific level as you have in our last few exchanges. I do not think the models are useless, and I also am convinced that they will only continue to become ever more accurate and reliable. What would you propose to do while these models have not yet fully matured? Just throw up our hands and declare that we give up? I maintain that our actions should be based on what knowledge we have, even if that knowledge is imperfect. I see no other rational course of action.

[TimG]

I do not think the models are useless, and I also am convinced that they will only continue to become ever more accurate and reliable.

I also believe models are useful tools as long as one is diligent about reporting the limitations with the results. But there are limits to what can be done with computer models and chaotic systems. For example, the earth's orbit is chaotic over long timescales which means the position of earth is impossible to predict for more than 2 million years into the future. This tidbit of information floored me when I first learned because I thought the gravity equations were so well defined that the solution would be deterministic. That is why I feel climate models will never be able to predict the future climate with any useful certainty.

I maintain that our actions should be based on what knowledge we have, even if that knowledge is imperfect.

What I said earlier: models can define the range of physically plausible outcomes. i.e. the models can tell us that adding CO2 is not going to cool the planet. That is useful information. What models cannot do is tell us how much warming we will see nor can they tell how the macroscale changes in temperature will effect life because life evolves and adapts.

Models also cannot tell us what we should do about the risks. The decision about what to do is a question of politics, economics, technology and values. Values are probably the most important issue because different people have different values. For example, China when faced with evidence that overpopulation was a threat adopted a one child policy. Enforcing this policy required that authorities engage many abhorrent acts such as forced abortions but it was something that the Chinese were willing to do. No amount of scientific evidence would ever convince a western country to adopt a similar policy.

Different values is why the climate debate has split along the left-right political axis. People on the left believe (even celebrate) the idea of collective action for collective good. This means they are willing to embrace policies that would give governments complete control of energy production in society. People on the right believe in the rights of the individual. They do not trust government and are not interested in giving government more power. This means many of the mitigation policies are categorically rejected by people on the right just like most people in Canada would categorically reject a policy of forced abortions.

These differences in values put constraints on what policies can be considered. So far, the left has relied on strategy which exaggerates the scientific certainty in order to bully people into accepting their policies. They have also engaged in vicious personal attacks on anyone who dares to question their policies. There will be no way forward until people accept that value differences exist and they are not going to go away.

Edited July 10, 2010 by TimG

[Bonam]

I also believe models are useful tools as long as one is diligent about reporting the limitations with the results. But there are limits to what can be done with computer models and chaotic systems. For example, the earth's orbit is chaotic over long timescales which means the position of earth is impossible to predict for more than 10,000 years into the future. This tidbit of information floored me when I first learned because I thought the gravity equations were so well defined that the solution would be deterministic. That is why I feel climate models will never be able to predict the future climate with any useful certainty.

This is an interesting example. One can talk at length about the complex reasons for why the Earth's orbit is not truly predictable in the very long term. For example, the effect of bodies besides the Sun and Earth (such multibody problems have no closed form solutions), the asphericity of these bodies and the asymmetry of their mass distributions, the varying interplanetary medium through which they travel, non-gravitational interactions, etc. And yet, despite these complexities, on a wide range of interesting timescales and levels of accuracy, the behavior of the system can be accurately predicted by drastically simplified models.

What I said earlier: models can define the range of physically plausible outcomes. i.e. the models can tell us that adding CO2 is not going to cool the planet. That is useful information. What models cannot do is tell us how much warming we will see nor can they tell how the macroscale changes in temperature will effect life because life evolves and adapts.

I think we'll have to disagree here. I think it is entirely within the realm of possibility for models to be able to tell us what is a reasonable amount of warming to expect over a given timeframe as a result of a given increase in CO2. If they are not there yet, then they will be in the near future.

Models also cannot tell us what we should do about the risks. The decision about what to do is a question of politics, economics, technology and values. Values are probably the most important issue because different people have different values. For example, China when faced with evidence that overpopulation was a threat adopted a one child policy. Enforcing this policy required that authorities engage many abhorrent acts such as forced abortions but it was something that the Chinese were willing to do. No amount of scientific evidence would ever convince a western country to adopt a similar policy.

On this part, I agree wholeheartedly.

Different values is why the climate debate has split along the left-right political axis. People on the left believe (even celebrate) the idea of collective action for collective good. This means they are willing to embrace policies that would give governments complete control of energy production in society. People on the right believe in the rights of the individual. They do not trust government and are not interested in giving government more power. This means many of the mitigation policies are categorically rejected by people on the right just like most people in Canada would categorically reject a policy of forced abortions.

I agree with your criticism of those on the left here, as well. In fact, despite being relatively convinced of the science, I still maintain what you call the "right" view on this topic, I oppose the implementation of draconian government action, besides additional funding for research on the issue and for the use of nuclear technology. At the same time, however, I don't think most people on the right really believe in the "rights of the individual" anymore. Certainly the policies of the Conservatives in Canada or of the Republicans in the US have been no less collectivist in nature than those of their Liberal/Democratic counterparts. Followers of the philosophies associated with the supremacy of individual rights (libertarians, objectivists, etc) form only a tiny minority, and their views diverge increasingly from mainstream "right" parties. But that is of course a wholly different topic.

I believe much of the resistance on the right is not a true defense of individual rights, but simply backlash against some of the ridiculous and inconvenient proposals made by proponents of rapid action in response to global warming proposed by some on the left, often from very hypocritical positions (people like suzuki, gore, etc, come to mind). Some "greens" would like to see us effectively return to the stone age, others call on others to save energy and reduce personal luxuries while they enjoy far more luxuries themselves, others idiotically bash investments in the technologies that are our only real means of reducing emissions (nuclear), so of course this will meet with significant resistance from those outside their movement.

These differences in values put constraints on what policies can be considered. So far, the left has relied on strategy which exaggerates the scientific certainty in order to bully people into accepting their policies. They have also engaged in vicious personal attacks on anyone who dares to question their policies. There will be no way forward until people accept that value differences exist and they are not going to go away.

I agree with this criticism, and if you have the patience to wade through the history of this thread you would see my criticisms of waldo/wyly precisely for this on several occasions.

[TimG]

on a wide range of interesting timescales and levels of accuracy, the behavior of the system can be accurately predicted by drastically simplified models.

Yes. But the key message here is all computer models have a limited range of validity and it is mistake to used them without understanding their range of validity. For example, weather models are reasonably good for 7 days and extremely useful within that timeframe.But anyone who tries to use a weather model beyond that range is peddling nonsense.

So the question becomes: on what timescales are climate models valid? I gave you the opinion of one expert that claims a few hours at most. Where is the evidence that shows climate models are valid on timescales of a decade or more? I am talking evidence - not theoretical discussions because we both agree they could be useful. I am saying that we don't really know if they are and there is some evidence that they are not.

I think we'll have to disagree here. I think it is entirely within the realm of possibility for models to be able to tell us what is a reasonable amount of warming to expect over a given timeframe as a result of a given increase in CO2.

We will have a good idea of CO2 sensitivity in 30 years. This will because we will have real data that tells us what it is. This real data will either validate or repudiate climate models.

Certainly the policies of the Conservatives in Canada or of the Republicans in the US have been no less collectivist in nature than those of their Liberal/Democratic counterparts.

The republicans are a coalition of social conservatives, security hawks and libertarians. This leads to a completely incoherent ideological view when one looks at all issues. The opposition to climate policies is driven by the libertarian wing. Edited July 10, 2010 by TimG

[Keepitsimple]

I also believe models are useful tools as long as one is diligent about reporting the limitations with the results. But there are limits to what can be done with computer models and chaotic systems. For example, the earth's orbit is chaotic over long timescales which means the position of earth is impossible to predict for more than 2 million years into the future. This tidbit of information floored me when I first learned because I thought the gravity equations were so well defined that the solution would be deterministic. That is why I feel climate models will never be able to predict the future climate with any useful certainty.

What I said earlier: models can define the range of physically plausible outcomes. i.e. the models can tell us that adding CO2 is not going to cool the planet. That is useful information. What models cannot do is tell us how much warming we will see nor can they tell how the macroscale changes in temperature will effect life because life evolves and adapts.

Models also cannot tell us what we should do about the risks. The decision about what to do is a question of politics, economics, technology and values. Values are probably the most important issue because different people have different values. For example, China when faced with evidence that overpopulation was a threat adopted a one child policy. Enforcing this policy required that authorities engage many abhorrent acts such as forced abortions but it was something that the Chinese were willing to do. No amount of scientific evidence would ever convince a western country to adopt a similar policy.

Different values is why the climate debate has split along the left-right political axis. People on the left believe (even celebrate) the idea of collective action for collective good. This means they are willing to embrace policies that would give governments complete control of energy production in society. People on the right believe in the rights of the individual. They do not trust government and are not interested in giving government more power. This means many of the mitigation policies are categorically rejected by people on the right just like most people in Canada would categorically reject a policy of forced abortions.

These differences in values put constraints on what policies can be considered. So far, the left has relied on strategy which exaggerates the scientific certainty in order to bully people into accepting their policies. They have also engaged in vicious personal attacks on anyone who dares to question their policies. There will be no way forward until people accept that value differences exist and they are not going to go away.

Very well put. Thanks.