Where did all the heat go?

EBD....it's melting the ice-caps on Mars. Oh wait. Sorry. That's Bush's fault.

Silly things like accuracy of the models doesn't matter, it's the end-game that counts.

Big Oil is bad and must be punished!
America is bad and must be punished!

Seriously though, models for very well-defined and well-known simple systems can still result in surprising errors.

Models for complex systems (and let's face it, there are few systems more complex than planetary climate) are more an expression of the modeller's mind (or rather their take on that system) than the system itself.

Making gov't policy based on model predictions for 100 years in the future is about as assinine as it gets.

It's about the same accuracy as shoving your thumb up your as& to check for brain cancer!

But, then again, one never knows where a leftie's brain is!

Well, this study surely is the final nail on the AGW coffin! Human-induced climate change is hereby confirmed a fraud! The science is settled!

http://www.weather.com/common/onlinepoll/results/lap_undeclared.html?generic_poll172

Very close.. 50.4% don't care...

QE - your sarcastic ability (entertaining as it is) simply doesn't match your continued ability to MISS THE POINT (entertaining as that is too).

Are you sure you don't have brain cancer, better check - safety first you know!

I'm cold!

But but, Canadians still want to waste billions if not trillions on helping the alzheimers infected Suzuker and his ilk buy more mansions in beautiful coastal places so they can flush their raw sewage into the Pacific. Still the moronic press run to this little demented man for a "supposedly lucid" quote about how he will save the planet from a necessary gas, co2. and stop the evil Harper from singlehandedly killing said planet. Hard to believe where we are headed when there are people stupid enough to believe this great scam, global warming, started by one man basically, the MO-ron.

QE- the correct conclusion, which you are missing, is that AGW, or human-induced global warming, is considered UNPROVEN.

The reasons for this conclusion are both an inadequate data base and inadequate modeling techniques.

bartinsky - what many/most Canadians aren't aware of, because the MSM press doesn't report it, but the Harper govt has put many millions into Clean Water projects across Canada. That includes not only the massive Great Lakes cleanup project but the many city/municipal water projects across the land.

Hmmm....

That 'reality' thing seems to be derailing the 'watermelon express'.... The 'greens' themselves are becoming concerned that they simply cannot defend their positions against 'AGW deniers' (oooh, you do have to admire the subtle manipulation, associating AGW sceptics with neonazis - their venom is much better honed than their science).

The Reference Frame (I came across it recently and am amazed by it - good stuff and real science here) writes that the pro-AGW crowd is beginning to wake up to the fact that they are loosing the debate (you know, the one that no longer exists because they won it years ago) BECAUSE THEY MIGHT BE WRONG!!!

http://motls.blogspot.com/2008/05/why-are-green-advocates-failing-in.html

Quite understandable ET the msm of this country and the US is so blinded and filled with pure hatred of Bush and Harper, they will never report anything constructive about either of them. What horrible places the journalism schools must be to turn out profs like Layton and the legions of losers that fill the 1600 dollar a month positions to write bs about constructive people.

I'm 49 years old so I remember how cold the 70's were...One summer was so cold we did not use the backyard pool. I also remember my grandfather telling me about the 1930's and how hot it had been...One winter was so warm and no snow that lumber could not be taken out of the woods because the ground was too wet and not frozen.

Several weeks ago, I predicted (Based on the past cold winter Canada and others around the world experienced and my personal conclusion that the cycle is again turning towards a colder period) to the landscaping crew I'm currently working with that this summer would be colder with lot's of cloody days and rain. They all looked at me, puzzled as they had all learned from medias that this would be a record breaking scorching hot summer again.
We'll eventually see who's right but so far here in southern Ontario, this spring has been below average temp wise.

"The science is settled!"
QE

Nice attempt at sarcasm.
What you seem to miss is that the Global Warming Hysterics have failed to make their case that the STATUS QUO needs radical change.

The facts are making the status quo is safer from your junk science revolution.

For anyone staying current on the epistemology of science, and the application of statistics to that endeavor specifically, Koutsoyiannis' work continues to be at the forefront.
The fact that some, e.g., QE do not recognize this only serves to indicate what useless twits they, in fact, are and remain.

Great post, Kate. More evidence.

IMO, a Gore fraud guilty verdict would work wonders in bringing the world to it's senses.

And a hot clamy feeling - for the fanaticals in the environmental movement. And those at the DSF, in public school administration, Arts Faculty, UN, Hollywood. But especially in media news rooms.

Garden path, comes to mind.

I think we should force people who believe in GW to wear a green star. These fools will not come out in public and spew their sh*t for fear of looking foolish. They just post anonymous fliers on bulletin boards and chat amongst themselves about how enlightened they are. Okay so you can't force people to wear the green star but us "deniers" could voluntarily where CO2 pins or something.

Back in the Dark Ages before everything in Canada was made safe by our loving government, we used to have huge fun exploding my mother's tulips with firecrackers on Firecracker Day, AKA Victoria Day.

I hadn't seen a tulip still in bloom on the 24th of May in a long time. Until this year. I've got a bunch of tulips still going strong out here and its 54 F outside.

Mr. Dion sir, bring on the carbon tax. I'm going to have a really good giggle watching the Liberals trying to sell that this year with the furnace still going in June and the fuel costs topping $1.30 a quart. (I hate metric, too.)

If you really stand back and look at the whole GW thing I have to ask myself, where are the scientists? If this is the real deal ,should there not be a high profile presence of dedicated climatologists convincing the public,working the talk shows etc, there are none! Anyone would be hard pressed to name 10, OK we have the fruitfly but I don't think he counts. Its all political.

Say, I just read Ron's comment about Algore and it suddenly hit me: in Canada we have a law against "spreading False News".

Given the mounting evidence of fraud regarding AGW, Algore and Dr. Fruitfly have about as much credibility as Ernsty Zundel and that Keegstra jerk. Those guys wen to JAIL for spreading their lies, should the Gore/Fly bunch be allowed a walk? Inquiring minds wanna know!

Lets try to bring the fruitfly up on charges.

Models are awesome. Based on empiical data, measuring the growth of my 3-year-old son, and extrapolating in the same way, I'm assured that he will be 37 feet tall by the time he's 46.

Yes, it is becoming increasingly evident that garbage in = garbage out when it comes to climate modeling, especially when it comes to "predictions" (and yes, I use the term loosely) of the contributions of CO2 to AGW.

The data from the Nasa Aquos satellite is already proving devastating to the AGW alarmists cause, and as more objective data and measurements become available, more and more people will become denialists of a problem that never existed in the first place.

ZZ- It seems strange to me that so many scientists are relying on extrapolation to predict climate trends. Anyone with a head for math, or even basic geometry knows that extropolation is the least accurate form of measurement. Especially inaccurate when the observed distance is shorter than the projected distance, as in your 3 year olds predicted height.

Again, it so amuses me how instantaneously skeptical SDA'ers are of any piece of research that supports the AGW thesis, yet utterly uncritical of any findings that purport to refute it.

Why might one be skeptical of this latest research by Koutsoyiannis and colleagues? At least two reasons come to mind:

1) Albany, Florida does not exist (Albany, GA, however, does).

That the researchers missed this basic fact, one can chalk up to unfamiliarity with US geography. As for Kate, well, with such juicy, damning material, is anyone really surprised that she could manage only a surface reading of the docs before parroting--sorry, blogging about--it on SDA?

Anyway, this minor gaff is slightly embarrassing, but it hardly throws the study results into question.

2) On the other hand, the study's methodology DOES throw its results into question. As the researchers themselves notes, GCMs are better at predicting temp trends at the macro-scale (continental and above) than at the regional level. Every modeller worth her salt--and, indeed, even the IPCC authors--will note that great caution is needed when downscaling GCM outputs for use at the regional level (entire chapters of the AR4 WGI report are devoted to this).

So what does Koutsoyiannis and Co. do? As far as I can tell from their presentation, they went ahead and downscaled the models right to the NOAA station level (that is, they took data from four nearby grid points, squished them together, and called it 'close enough'). They say this is based on methodology developed by Georgakakos (2003), but in that earlier study, Georgakakos only downscaled from his chosen GCM's grid square level to the US climate division level (which is in the magnitude of 10,000km2, and contain multiple stations).

But wait, they say (in slide #3), this "did not prevent IPCC to give regional projections"! But slide #3 is disingenuous (and not only because it printed the wrong IPCC slides and/or cited the wrong IPCC authors), because the IPCC never downscaled GCMs to the sub-regional level, and they were in any case quite open in admitting that their regional-level climate projections are imperfect.

So, bottom-line: Koutsoyiannis et al. took models and did what the models were never designed to do, and then blamed them for not doing it. And that, apparently, is what constitutes good, hard science for the typical SDA reader.

Again, it so amuses me how instantaneously skeptical SDA'ers are of any piece of research that supports the AGW thesis, yet utterly uncritical of any findings that purport to refute it.

Why might one be skeptical of this latest research by Koutsoyiannis and colleagues? At least two reasons come to mind:

1) Albany, Florida does not exist (Albany, GA, however, does). But this is a minor point.

2) Less, minor, however, is the study's questionable methodology. As the researchers themselves note, GCMs are better at predicting temp trends at the macro-scale (continental and above) than at the regional level. So what do they do? As far as I can tell from their presentation, they went ahead and downscaled the models right to the NOAA station level (actually, they took data from four nearby grid points, squished them together, and called it 'close enough'). They say this is based on methodology developed by Georgakakos (2003), but even that study only went down to the US climate division level (which is in the magnitude of 10,000km2, and contain multiple stations).

So, bottom-line: Koutsoyiannis et al. took models and did what the models were never designed to do, and then blamed them for not doing it. And that, apparently, is what constitutes good, hard science for the typical SDA reader.

There is, however, a New Albany, Florida.

I also notice the 8 sites chosen are from across the globe, not four 'nearby' grid point.

But hey, why pay attention to what they are saying, the science is settled, right?

"There is, however, a New Albany, Florida. I also notice the 8 sites chosen are from across the globe, not four 'nearby' grid point."

Wow. Chiding me for not "paying attention" and yet...

1) There is in fact no New Albany, Florida, at least not according to the US Census: factfinder.census.gov/servlet/SAFFPopulation?_event=Search&_name=new+albany&_state=&_county=new+albany&_cityTown=new+albany&_zip=&_sse=on&_lang=en&pctxt=fph

2) There is no New Albany, Florida, NOAA climate station: 3w.ncdc.noaa.gov/oa/climate/stationlocator.html

3) Within each of the eight global sites, the researchers took data from the four GCM grid points closest to the selected station in order to create their station-level model time series.

You got that, otter? I'll say it again, just so you're clear: four grid points, for each of the eight sites.

Again, wow. Some real bright lights your site attracts, Kate.

"There is, however, a New Albany, Florida. I also notice the 8 sites chosen are from across the globe, not four 'nearby' grid point."

Wow. Chiding me for not "paying attention" and yet...

1) There is, in fact, no New Albany, FL, at least not according to the US Census.

2) There is, in any case, no New Albany, FL, NOAA climate station.

3) Within each of the eight global sites, the researchers took data from the four GCM grid points closest to the selected station in order to create their station-level model time series.

You got that, otter? I'll say it again, just so you're clear: four grid points, for each of the eight sites.

Again, wow. Some real bright lights your site attracts, Kate.

The Albany referred to is in Georgia, apparently, not Florida.

www.climateaudit.org/?p=3086 has a long commentary discussion of this presentation, mostly in support of the thesis, on balance, it appears. Not for the faint of technical heart...

QE: Good catch on Albany. Otter, I can't find a record of either an Albany or a New Albany in Florida and now it appears that those at ClimateAudit can't either.

Regards,
John

Uh John Cross, QE I looked at the wording "from the Albany, Florida inputs". I don't think Albany is a place let alone in Florida. It could be "the Albany" is a publication covering the climactic inputs from Florida.

"I don't think Albany is a place let alone in Florida. It could be "the Albany" is a publication covering the climactic inputs from Florida."

Follow the links from Kate's post to Dr. Koutsoyiannis et al.'s actual poster presentation (especially slide #5). Albany is indeed a place (and it can be found in Georgia).

Albany, sure it isn't Albuquerque, maybe we should ask Bugs Bunny to provide some input here.

As this crucial piece of evidence is apparently necessary to debunk the debunkers we must make sure it is correct!

Seriously QE and JC, perhaps you should spend as much time analysing the deficiencies in AGW theory as you do in finding holes in the "debunker science".

Actually boys, I didn't even have time to read this article, too busy working. However, from my extensive use/knowledge of models my original statement is bang on.

The fact that AGW supporters are continuously updating/changing their models only make their predictions even more ridiculous.

After all, if the model was accurate 10 years ago why change it now? Accuracy doesn't change with time, only bias does!

As I said, when so much credibility is given to some models, while ignoring others, that it leads to the formation of radical gov't policy then we are truly entering the twilight zone.

Our economic models are much, much more accurate than our climate models - and yet look at what happens when politicians get a hold of them for their own purposes. The UK prime minister just a few years ago said that there was no longer any boom/bust cycle because we understood economics so well.

Tah dah - and now we are supposed to believe that climate models will lead to good policy- give your head a shake!!!!

Hi QE,

Thanks for your critique. You have found a number of errors that we were unaware of. Albany is, indeed, Albany, Georgia. And, yes, the figures of slide 3 are from chapter 10 of Climate Change 2007, although we incorrectly attribute them to chapter 11. In addition, slide 3 says that "this did not prevent IPCC to give regional projections". Our intended meaning was that "this did not prevent IPCC to give geographically distributed projections". These corrections would make slide 3 be consistent, I think.

Now, concerning the essence of your objection, which concerns the downscaling level (regional/subregional/station, whatever this means). Let's take Albany, for an example. We picked up the four grid points that were nearest to Albany, and we used that linear combination of the four which better correlates to the real data. That is, we did not merely use the nearest point or so, but the best estimate of the nearest four points. These four points have distances between them which depend on the model, but they are of the order of 400 km.

Let me understand your objection: do you mean that four points define an area that is too small? Or do you mean that the real Albany data is sensitive to the microclimate of the area and therefore might not correlate well with the "average" climate of the area? Is it, in other words, possible for the temperature of the area to change and for Albany's to change differently?

If the answer is yes, would it not be strange the correlation is bad not for one, or two, or three, but for eight out of eight stations around the globe?

Antonis: Thanks for clearing up those issues. If I may ask a question (perhaps your paper makes it clear but I could not find it in your slide presentation), what criteria did you use to select your 8 stations beyond the fact they represent different climates and have records over 100 years?

Also, I liked the presentation, but do you have a link to your paper?

regards,
John

Antonis:

Thank you for your attention. Is there a forum where either you or your colleagues discuss your work or interests that might be accessible to a math minor or the like?

Dr. Christofides,

Thank you for your reply. My concern is less with your study's methodology, than with the conclusions that result from them.

For instance, the study concludes that "future climate projections [are] not credible" because:
(1)"model outputs at annual and climatic (30‐year) scales are irrelevant with reality," and
(2) "model predictions are much poorer that an elementary prediction based on the time average."

However, the model time series predictions used in your comparisons were themselves derived from a novel technique that your team developed. The poor correlations between historical readings and model outputs at all eight station sites may therefore be due to inadequacies in the chosen downscaling technique rather than any inherent fatal flaw in the GCMs.

Thus, the conclusion that "future climate projections [are] not credible" is overstated. Rather, a more accurate conclusion may be that "small regional scale future climate projections, based on time series estimates produced by our downscaling technique, are not credible."

John: I don't know exactly how Dr Mamassis chose the stations, but I'll be surprised if he did anything else than search the web for suitable stations and pick up the first he came across with in each area.

Right now there is only the presentation, there is no paper, but we are quite busy preparing a paper as well.

Gary: The short answer is no, we don't have a forum. But if you explain what exactly you are interested in, we might look into it. Am I right in understanding that you'd like to understand more about the global warming debate but you are not an expert? Join the club, I'm not either. It is in our plans to make this accessible to more people, though we don't have a concrete schedule.

Until we get there, I'd suggest the following two works by Koutsoyiannis: "nonstationarity vs scaling", which you can download from the journal if you have a subscription; otherwise you can read the pre-print version at http://www.itia.ntua.gr/en/docinfo/673; it doesn't have the fancy format of the journal version but it is ok. For a person with background in science or math, it's like reading a fairy tale. The second suggestion is a work also presented in Vienna together with the presentation we've been discussing: http://www.itia.ntua.gr/en/docinfo/849. I don't know how easy it is to follow it. I found it quite easy, but I've heard Koutsoyiannis explain similar things lots of time in the corridors, which could make a difference. If you find these interesting, you could try to browse http://www.itia.ntua.gr/en/documents/ to see if there is anything else you like.

QE: I'm not a Dr, and please call me Antonis. In fact, I am the programmer of the team. I truly appreciate the manner in which you are questioning everything.

Koutsoyiannis explained to me that I'm making incorrect use of the term "downscaling". He says that what we did is not downscaling. Downscaling is another thing: you accept that the output of a climate model is OK, and then you try to make predictions for a region. But because the local output for the region sucks, you make some transformations in order to adapt the output to the reality of the region; then you use the transformed model output in order to make local predictions. This is what I understood that downscaling is.

What we did is nothing like that. We merely compared local measurements to, say, the nearest grid point. But in order to make it easier for the model, we made the trick with the four points as I explained above. I wouldn't name this a novel technique that we developed, but the obvious thing someone would do for testing. Therefore, I disagree with the "based on time series estimates produced by our downscaling technique", that you write. I'd rather conclude that regional projections suck, period.

But, yes, we've gone further than that and we've removed the word "regional" from that conclusion, and that is, indeed, bolder. I argued with Koutsoyiannis that you are right, and that maybe we are concluding the things that we want to conclude and not the ones that our results show. Koutsoyiannis disagrees, however, and he has important points. I will explain one of them in my own words. Let me point out that in all this post I speak for myself only and not for the other authors, but even more so in the paragraphs below.

Suppose there is a building whose structure has been analyzed using the Finite Element Method. Someone goes and picks up eight individual elements and installs measurement devices, and it turns out that the FEM model output is substantially wrong in all these eight elements. Then the designer tells you, yes, I can't get the elements correctly, but I get the building correctly. Would you enter the building after that? I mean, what kind of reasoning is that? Mr. designer, you claim that the elements show noise around a trend, and you model the ELEMENTS including their noise, you get them wrong, but somehow you get the trend right? Until you provide me with some serious evidence for this strange claim, I maintain that disproval at the local level is full disproval. In other words, if I let an apple fall down and the way it falls down disproves Newton, then, in the absence of convincing evidence to the contrary, Newton is disproved at the planetary level as well.

There is more to it than that. There is much background behind the claim that models work at large scales but not at small scales. But I won't go into any more details, if not anything else because I am unqualified. I hope that these questions will be answered when the paper is out.

But I'd also like to point out that, even with "regional" in, this result is already important, it can't be dismissed on the grounds that it is known that regional projections are "imperfect". In my opinion, we have proven that the projections are considerably worse than imperfect.

Antonis:

Well a simple query that comes to mind:

We often hear AGW modellers complain that 'red noise' in the data obscure the effects their models seek to characterize.

I have difficulty with the concept of 'noisy data', as though the result is known but the data are imperfect.

If an effect is too small to be seen against 'white', or 'red', or 'grey' noise, whatever, doesn't that confound its consequence?

Is this 'noise' a misdirection or a real problem of measurement?

A second thought: Roy Spencer of UAH has a paper in process on the CO2 curve, an upward trendline with a seasonal sine-like signal riding on top. Using the F test, he showed that the 12C and 13C variation in the two, the trend and the seasonal signal, were identical. My interpretation is that this result indicates the CO2 fluence responsible for both is one and the same, e.g., Southern Ocean heating.

Any critical thoughts?