From the University of Gothenburg
Climate models are not good enough
Only a few climate models were able to reproduce the observed changes in extreme precipitation in China over the last 50 years. This is the finding of a doctoral thesis from the University of Gothenburg, Sweden.
Climate models are the only means to predict future changes in climate and weather.
“It is therefore extremely important that we investigate global climate models’ own performances in simulating extremes with respect to observations, in order to improve our opportunities to predict future weather changes,” says Tinghai Ou from the University of Gothenburg’s Department of Earth Sciences.
Tinghai has analysed the model simulated extreme precipitation in China over the last 50 years.
“The results show that climate models give a poor reflection of the actual changes in extreme precipitation events that took place in China between 1961 and 2000,” he says. “Only half of the 21 analysed climate models analysed were able to reproduce the changes in some regions of China. Few models can well reproduce the nationwide change.”
China is often affected by extreme climate events. Such as, the flooding of 1998 in southern and north-eastern China caused billions of dollars worth of financial losses, and killed more than 3,000 people. And the drought of 2010-11 in southern China affected 35 million people and also caused billions of dollars worth of financial losses.
“Our research findings show that extreme precipitation events have increased in most areas of China since 1961, while the number of dry days – days on which there is less than one millimetre of precipitation – has increase in eastern China but decreased in the western China.”
Cold surges in south-eastern China often cause severe snow, leading to significant devastation. Snow, ice and storms in January and February 2008 resulted in hundreds of deaths. Studies show that the occurrence of cold surges in southeast China significantly decreased from 1961 to 1980, but the levels have remained stable since 1980 despite global warming.
” Paul Westhaver says:
March 25, 2013 at 10:56 pm
7) change the model, go to 3)”
That’s what a modeller would say. A physicist would say
‘7) identify where the physics of the model is incorrect, change the model and then go to 3).’
That way the use of fudge factors is eliminated.
But a mathematician would ask ‘Is the system in question intrinsically chaotic and if so is it realistic to use the model to try to support my hypothesis? If no then go to the pub, if yes then correct model and go to 3).
Climate models will always be wrong if the wrong basics are factored in. It is also difficult/impossible to model a chaotic system.
What I find amazing is that for the past 15-17 years, “natural factors” have managed to cancel out, with exquisite precision, the increase in global temperature purported to result from increased CO2 in the atmosphere. In fact these “natural factors” have managed to increase their effect, year by year, to exactly match the increased radiative effect of this CO2.
What do you think the odds of this are (assuming independence and no-autocorrelation- which is what climate psientists do in their reconstructions)?
A not unreasonable assumption is to say that the probability that for any one year the chances of “natural factors” exactly matching the radiative effects of increased CO2 is 50%, then the binomial outcome culmulative probability = (0.5)^15 = 0.00003, or 1 chance in 33,333
If we are really generous and say the probability is 70% then (0.70)^15 = 0.0047, or 1 in 212.
Compare the “mighty” GCMs that can’t forecast or hindcast, with Scafetta’s “lowly” model that can both forecast and hindcast.
Scafetta N., 2012. Testing an astronomically based decadal-scale empirical harmonic climate model versus the IPCC (2007) general circulation climate models. Journal of Atmospheric and Solar-Terrestrial Physics 80, 124-137. DOI: 10.1016/j.jastp.2011.12.005. PDF – Supplement
Harmonic model for solar and climate cyclicalvariation throughout the Holocene based on Jupiter–Saturn tidal frequencies plus the 11-year solar dynamo cycle. Presented at 2012 Fall Meeting, AGU, San Francisco, Calif., 3-7 Dec, 2012.
Press release @ur momisugly Eurekalert
Regional Climate Group – University of Gothenberg Regional Climate Group Website
PhD defense 2013-02-25
Thank you Dr. Leif.
Indur M. Goklany says:
March 25, 2013 at 5:54 pm
“Climate model simulation of precipitation has improved over time but is still problematic. Correlation between models and observations is 50 to 60% for seasonal means on scales of a few hundred kilometers.” (CCSP 2008:3).
I know this is not a quote from you Indur but correlations of 50 to 60% on precipitation? What is the the standard deviation? Use it alone and you will beat the hell out of the models 9 times out of 10.
Would a link URL to Mr. Ou’s paper not be appropriate? It is available from his university’s dissertation bank.
Leo Geiger says:
March 25, 2013 at 7:19 pm
Just so no one gets the impression that they are saying climate models are useless, or anything remotely along those lines.
When you arrive back on this planet give us a call !!
The key design aspect of all successful fishing lures is that they are designed first and foremost to catch fishermen. When the fisherman is looking over a series of lures on display on the wall, the successful fishing lure is the one that catches the fisherman’s eye and ends up being purchased.
The same process is at work in climate models. Whether a climate model can successfully model the future is only a secondary design requirement, because it will be many years at best before a model will be proven to actually be fit for purpose.
The primary purpose of all successful climate models is to attract scientists, and through their activities to attract funding. Whether the model is any good at predicting the future is largely irrelevant to the success of the model, because it lags the funding.
One thing to keep in mind is that models are generally created from weather models. As such they do have a certain existing weather knowledge within them. For example, the Sahara will not get vast amounts of rain. It should not be surprising then that they will some answers correct. It will happen automatically. That doesn’t mean they have any usefulness relative to climate. After all, a stopped clock is right twice a day and even a blind squirrel finds a nut once in awhile.
David L. Hagen says:
March 26, 2013 at 4:50 am
Compare the “mighty” GCMs that can’t forecast or hindcast, with Scafetta’s “lowly” model that can both forecast and hindcast.
=============
Early humans learned to predict the future in just this fashion, by observing the location of the planets and stars in the sky. We learned to predict the seasons long before we understood the cause.
We accurately predict the tides in much the same fashion, by observing the position of the sun, moon and major planets in the sky. We understand the cause of the tides rather well, but if you try and model the tides using first principles as we do in climate models, the predictions fail horribly.
Loosely, the future defies all attempts at prediction except in very simple cases. For example, we can pedict the orbits of 2 planets around each other, but if you introduce a third the future becomes murky. The problem is that while first principles work in principle, they do not work in practice becuase the computational problem size grows exponentiallly. This exponential growth in problem size makes first principles impossible to solve using computers.
Thus, until we solve the problem of computational problem size, we need a different method of predicting the future. While Astrology is a dirty word in science, we do in fact use the position of the moon, planets and stars in the sky that accurately predicts the seasons and the tides.
Given that climate is computationally impossible to solve from fist principles on any modern computer, it should not be a surprise to anyone that the same techniques that led to accurate predictions of the seasons and tides show greater accuracy than computers when applied to cliamte prediction.
Thank you all for the interesting discussion and some of the good suggestions and comments. It is a new start for me and my research work. Please refer to https://gupea.ub.gu.se/handle/2077/31816 or the home page of our group (http://rcg.gvc.gu.se/) if you want to find more detail of the work.
We can’t 100% accurately predict the future for both theoretical and practical reason. We just hope the models can be better and better which is true from the developing history of the global climate models from 1D to 4D. It is very important for the model is the only means to project the future.
Paul Westhaver said: on the 25th at 10:56 pm
First thing I want all you new budding scientists to know is the scientific method.
Paul, you forgot 3 steps:
2A) you get a $5 million grant from the gove’mnt to pay for your existence for a year (and 4 post-docs)
6A) you write and have published in a “science” mag a report with a title such as “Global Warming to Accelerate do to Human Existence!” which at the end calls for more study
6B) you get another $5 million grant from the gove’mnt for (same reason above) “more study”…
opps, that’s “due” in 6A above…
‘Climate models aren’t good enough to hindcast, says new study’
Tinghai just kissed any idea of a career idea in climate science good bye , ‘the Team nether forgives nor forgets .
Tinghai says:
March 26, 2013 at 6:33 am
“We just hope the models can be better and better which is true from the developing history of the global climate models from 1D to 4D. It is very important for the model is the only means to project the future.”
As someone who has been practicing CFD for over 20 years, I wish you the best in your research. Please continue to strive for accuracy and robustness, while clearly identifying and documenting your methods and formulations, something which is clearly lost on low quality/accuracy codes such as Model E from the controversial NASA/GISS. (I urge everyone who has an interest in improving the quality of climate prediction software to visit the NASA/GISS Model E site to see how NOT to program in FORTRAN and how NOT to document you methods clearly).
Thank you for the encouraging and suggestions. I need to learn more so as to prepare for the future.
Welcome Tinghai Ou.
I have been influenced by N. N. Taleb (Black Swan, and Antifragile) on forecasting in finance and meteorology and the demarcation problem.
Dr. Tinghai Ou
Congratulations. Thanks for the links to your thesis: Observed and simulated changes in extreme precipitation and cold surges in China: 1961–2005
6 Conclusions
fred berple
Good observation. Re ” Astrology is a dirty word” – try “Astronomy” & “Planetary Physics”.
Cold surges may be increasing again.
Flagship Daily DIE WELT Stuns Germany: “Scientists Warn Of Ice Age”, Cites New Peer-Reviewed Russian Study
“Latitude says:
March 25, 2013 at 7:01 pm
Few models can well reproduce the nationwide change
=====
what are the odds when you flip a coin………….”
50/50 odds,which is better then the models,which are 100% wrong.
Thank you for your interests with this topic. Precipitation is one of the variables which is very difficult to be reproduced by the models. East Asian summer rainfall is affected by the East Asian Summer Monsoon, the precipitation pattern is poorly reproduced even if the model can well simulate the atmosphere circulation. There are many reasons for this, such as the model physics to simulate the precipitation. And some factors like Pacific Decadal Oscillation (PDO) is not well resolved in the global climate models. There is hope that the models can be better. The models are improving.
This is not like flipping a coin, which is a random number, and this can also be 100% wrong if you are not lucky. Hope you will not loose you confidence on the models based on my work which is not my initial opinion related to this.
Garbage assumptions In : Garbage Gospel Out.
When will we replace the current politicians, who are bedazzled by the shiny, cannot comprehend the limitations of assumptions dressed up as computed models?
As many here have pointed out, the IPCC grade projections are rubbish, were rubbish yet were considered good enough for government, government mandated robbery that is.
Too much incompetence and avoidance of responsibility here to be accidental.
There is clearly something wrong with the past! Historic data should be corrected in line with climate models.
“Whoever watches the wind will not plant; whoever looks at the clouds will not reap.”
-You can’t predict the weather that precisely. Well, at least according to the Good Book.