Models: some are good, some are not so good, some are useless.
To conclude, climate models can and have been verified against observations in a property that is most important for many users: the regional trends. This verification shows that many large-scale features of climate change are being simulated correctly, but smaller-scale observed trends are in the tails of the ensemble more often than predicted by chance fluctuations. The CMIP5 multi-model ensemble can therefore not be used as a probability forecast for future climate. We have to present the useful climate information in climate model ensembles in other ways until these problems have been resolved.
From http://www.realclimate.org/index.php/archives/2013/04/verification-of-regional-model-trends/
h/t to Steve Mosher
See also: http://www.gfdl.noaa.gov/cms-filesystem-action/user_files/tk/knutson_et_al_regional_trends_jan2013.pdf
Shelama, did you have anything to say, except insults?
Maybe the realclimate authors just got their annual Sports Illustrated Swimsuit issue, that’s they only way they will ever come close to some “good models”.
“climate model ensembles”
Are a bit like my gambling ensemble model: I just put money on every square and I’m never wrong! Which is as successful as Lewansky’s “chat up ensemble model” …he just goes through the list of chat up lines until he finds one that will get him into bed with the model.
There’s a joke … how do you know when climate researcher’s band is playing? You hear their ensemble playing every note in the book … in the hope that one fits the tune.
As of now, the only comment on the RealClimate thread is by me. It reads:
Geert Jan van Oldenborgh’s video abstract of the “Reliability of regional climate model trends” paper is very well done:
http://iopscience.iop.org/1748-9326/8/1/014055
Thanks, Geert Jan.
The actual quote from George Box was “essentially all models are wrong, some are useful”
Friends:
Climate model ensembles are nonsense.
Average wrong is wrong.
If one of the models were right then one would not know which.
But if its outputs are averaged in an ensemble then all the others would be wrong.
Average wrong is wrong, and there is no way to determine how wrong.
Richard
Credit to Mosher for the hat tip. That couldn’t have been easy for him.
lsvalgaard says:
April 15, 2013 at 10:16 am
“There is no evidence for long-term changes in spectral composition. There may be such change within each solar cycle, although this is not firmly established yet and some models do incorporate that [and find only minimal impact – less than a tenth of a degree or so],”
Such a model would still suffer from the known shortcomings of all GCM’s so what it does find for the variation in the solar spectrum is as realistic as the rest of its results – not much. At least that must be the default assumption as long as all the state’s climatologists and all the state’s supercomputers do not show ANY predictive skill.
Biggest admission at RealClimate is in a post about the tar sands. They evaluate the impact of using shale with a climate model that has been tuned to yield a 3C response.
Along with your “quote of the week”, you could also have several “Quotes of the Weak”, and even just plain ole “Weak Quotes”
DirkH says:
April 15, 2013 at 1:25 pm
“There is no evidence for long-term changes in spectral composition. There may be such change within each solar cycle, although this is not firmly established yet and some models do incorporate that [and find only minimal impact – less than a tenth of a degree or so],”
Such a model would still suffer from the known shortcomings of all GCM’s so what it does find for the variation in the solar spectrum is as realistic as the rest of its results – not much.
The model is not the issue. The point is that there is no evidence for any long-term changes in spectral composition.
Bob Tisdale says:
April 15, 2013 at 12:13 pm
———————————-
Thank you for the link Bob, very useful. It says it more clearly:
“However, when the global mean climate response is factored out, the ensemble is overconfident: the observed trend is outside the range of modelled trends in many more regions than would be expected by the model estimate of natural variability and model spread. Precipitation trends are overconfident for all trend definitions. This implies that for near-term local climate forecasts the CMIP5 ensemble cannot simply be used as a reliable probabilistic forecast.”
They couldn’t quite bring themselves to say “the sun”.
But this is funny:-
Finally, the model response to the changes in greenhouse gases, aerosols and other forcings may be incorrect.
About time.
How very trenchant Barry.
Enough monkeys at enough typewriters will produce all the works of Shakespeare. http://en.wikipedia.org/wiki/Infinite_monkey_theorem_in_popular_culture
A stopped clock is right twice a day. http://en.wiktionary.org/wiki/a_stopped_clock_is_right_twice_a_day
Given enough different climate models, one of them will eventually produce a plausible result.
Given enough commenters firing potshots from the trenches into the foggy air, some comments will eventually be trenchant.
Given enough time and posts, Barry will be absolutely correct.
LOL. The problem is, being able to recognize the above remarkable occurrences when they happen.
I still say get rid of them.
Methinks GCMs are having a Humpty Dumpty moment.
There’s a Josh cartoon here somewhere.
Kurt in Switzerland
lsvalgaard says:
April 15, 2013 at 10:16 am
milodonharlani says:
April 15, 2013 at 9:58 am
GC models were known GIGO even before major discoveries in climatology of the past 16 years such as the PDO, AMO & fact that TSI may not change much but its spectral composition does
There is no evidence for long-term changes in spectral composition. There may be such change within each solar cycle, although this is not firmly established yet and some models do incorporate that [and find only minimal impact – less than a tenth of a degree or so], e.g. http://lasp.colorado.edu/sorce/news/2011ScienceMeeting/docs/presentations/6b_Cahalan_Sedona_9-15-2011.pdf
————————————————————————-
Thanks for 2011 link. I’d feel better about its conclusions were the models not run by NASA GISS.
Gavin Schmidt himself used one of his GIGO climate models to predict milder NH winters as a result of the greenhouse effect. Back in 2004 Gavin and Michael Mann stated that:
These scientist should leave their computer games at home and carry out some measurements and do some actual hard work instead of playing. Play time is over.
References
http://dx.doi.org/10.1038/20905
http://www.meteo.psu.edu/~mann/shared/articles/Schmidtetal-QSR04.pdf
When the MSM heralds this as important news that should significantly change government policy, I’ll rejoice. Until then, it’s “meh”.
James Ard says: “Credit to Mosher for the hat tip. That couldn’t have been easy for him.”
Mosh is dedicated to truth and the correction of falseness as he sees it.
“…climate models can and have been verified against observations in a property that is most important for many users: the regional trends.
I have a three-sided coin that will correctly predict regional trends 33% of the time.
Willis Eschenbach says: April 15, 2013 at 10:05 am
———————
Indeed, here’s another snippet of great scientific insight:
” [ … ] climate models can and have been verified against observations in a property that is most important for many users: the regional trends. This verification shows that many large-scale features of climate change are being simulated correctly, but smaller-scale observed trends are in the tails of the ensemble more often than predicted by chance fluctuations. [ … ] ”
Funny that their models are not capable of forecasting awesome effect of ‘El Nino’ and ‘La Nina’ events … are these what they term “smaller scale observed trends” ? They’ve also failed at forecasting the lack of increase in temperature relative to CO2 gains … yeah, the cornerstone of their entire CAGW premise must equally be one of the unfortunate “smaller scale observed trends”.
Their models are so weak that they couldn’t pull the skin off a custard, so to speak.
Based on mileage yielded thus far, models could help mitigate global warming.
milodonharlani says: “GC models were known GIGO even before major discoveries in climatology of the past 16 years such as the PDO, AMO & fact that TSI may not change much but its spectral composition does”
lsvalgaard says: “There is no evidence for long-term changes in spectral composition. There may be such change within each solar cycle, although this is not firmly established yet”
From: Solar forcing of winter climate variability in the Northern Hemisphere
Sarah Ineson, Adam A. Scaife, Jeff R. Knight, James C. Manners, Nick J. Dunstone, Lesley J. Gray & Joanna D. Haigh:
“An influence of solar irradiance variations on Earth’s surface climate has been repeatedly suggested, based on correlations between solar variability and meteorological variables [1]. Specifically, weaker westerly winds have been observed in winters with a less active sun, for example at the minimum phase of the 11-year sunspot cycle [2, 3, 4]. With some possible exceptions [5, 6], it has proved difficult for climate models to consistently reproduce this signal [7, 8]. Spectral Irradiance Monitor satellite measurements indicate that variations in solar ultraviolet irradiance may be larger than previously thought [9]….
“…If the updated measurements of solar ultraviolet irradiance are correct, low solar activity, as observed during recent years, drives cold winters in northern Europe and the United States, and mild winters over southern Europe and Canada, with little direct change in globally averaged temperature….”
http://www.nature.com/ngeo/journal/v4/n11/full/ngeo1282.html
IF UV affects weather (through raising the black body temperature of the sky, decreasing atmospheric ozone, altered winds, or whatever mechanism), it is not necessary for there to be long term changes in spectral composition in order to contribute to the 20-odd year apparent warming trend. It’s only necessary for peak UV to fall within that time frame more than it usually does.
But based on the uncertainty cited in the paper above, we probably simply don’t know if there is or isn’t a long term change. The reason there is “no evidence for long-term changes in spectral composition” may be just that we haven’t been looking long enough.
“Between 2004 and 2007, the Solar Irradiance Monitor (blue line) measured a decrease in ultraviolet radiation (less than 400 nanometers) that was a factor of four to six larger than expected.”
http://www.nasa.gov/topics/solarsystem/features/solarcycle-sorce.html
“We may have a lot more to learn about how solar variability works, and how the sun might influence our climate,” Robert Cahalan, NASA, ibid.
They know full well they are incorrect but they are under peer pressure, political pressure, financial pressure to conform.. Your only other choice is unemployment as the deniers are not exactly hiring right now..
Think of how many people or businesses have a stake in the green revolution.. They have been promised tax funded profit outside the normal capitalist system.. All they have to do is support the green agenda of doom and watch their bank accounts grow..
Ten cents on the dollar is not a return.. its a disaster.
jorgekafkazar says:
April 15, 2013 at 5:13 pm
But based on the uncertainty cited in the paper above, we probably simply don’t know if there is or isn’t a long term change. The reason there is “no evidence for long-term changes in spectral composition” may be just that we haven’t been looking long enough.
That is what it means: no evidence. The original poster stated as if such changes were already a fact. Futhermore the Harder et al. suggestions are not established yet. Many think they are spurious, due to uncertain calibration.
There is some evidence for no changes on century time scales: http://www.leif.org/research/GC31B-0351-F2007.pdf