RealClimate Raises the Bar AGAINST Climate Models
With the death of Ed Lorenz and a world apparently taking a hiatus on the way to unstoppable anthropogenic global warming, It has taken a group effort at RealClimate to try to deal with the issue of chaotic weather vs. climate modelling: “Butterflies, tornadoes and climate modelling“.
Rather unfortunately for the authors, the conclusions contain a remarkable amount of unintended irony.
[…] But how can climate be predictable if weather is chaotic? The trick lies in the statistics. In those same models that demonstrate the extreme sensitivity to initial conditions, it turns out that the long term means and other moments are stable. […] Climate change then is equivalent seeing how the structure changes, while not being too concerned about the specific trajectory you are on
In other words, “climate change” is an entity that can only become observable in the long, long term. And since there is little concern for the “specific trajectory”, there literally exists NO possible short-term sets of observations that can falsify the climate models.
Another way of saying it is that for the climate problem, the weather (or the individual trajectory) is the noise. If you are trying to find the common signal that is a signature of a particular forcing then averaging over a number of simulations with different weather works rather well […]
In other words, since each and every atmospheric event can be obviously described as “weather”, there is no single observation that can falsify the climate models.
Their work doesn’t have to deal with any single observation, no short-term sets of observations…do they realize what they are saying???
Real climate is in their own words almost perfectly insulated from the real world. Nothing that can ever happen will be able to disprove the work of the climate modellers, apart from multi-decadal averages that are so poorly defined, they can easily be used to demonstrate anything.
Is this “science”? Looks more like long-term guaranteed employment to me… No wonder Anthropogenic Climate Change has important detractors in the metereological community.
In further irony, the above pairs up perfectly well with RC’s “comments policy” that can be summarized more or less into “we will censor everything we do not like“.
RealClimate: the insulated web site, where insulated researchers post insulated content. Now I understand why poor Gavin Schmidt had such a hard time dealing with an open debate…
The Unbearable Nakedness of CLIMATE CHANGE 
“Something that cannot be falsified cannot really be proven true either”
Evolution for a long time was a non-falsifiable theory and yet I would say that there was high confidence that it was true. I would say the more correct relationship is not falsifiability and truth but falsifiability and predictive power. Theories which have strong predictive power are also easily falsifiable. Theories which don’t have very little power. Evolution for a long time until the revolution in molecular genetics had extremely poor predictive power.
“The observation of a laboratory fluid flow that was inconsistent with the Navier Stokes equations would also falsify the climate models. ”
As would falsification of thermodynamics, energy conservation, mass conservation and Newton’s laws. The real question isn’t whether climate models confirm what we already know (Navier Stoke’s equations) but whether they tell us anything new. Therefore falsifiability is only in the are of long term climatic averages.
My concern isn’t so much falsification. Its the opposite. Its reliability. The climate models are supposedly making reliable predictions about 30 year climatic trends. Therefore we need about 30 years to get one data point. Even if climatic models are exactly correct, I will not be impressed. I want at least 100 successful predictions before I think we can have confidence in the model and that will take what 30*100 = 3000 years.
The observation of a laboratory fluid flow that was inconsistent with the Navier Stokes equations would also falsify the climate models. There are tons of ways in which the climate models can be “falsified” on the basis of short term obersvations. Unfortunately, most of them involve falsifying well-established laws of physics.
The laboratory observation of CO2 absorbing infrared radiation and reemitting in the UV would constitute a single short term observation that falsified the climate models.
To streamtracker:
I have explained why models in the “RC world” cannot be falsified by a single observation (“weather”, that is “noise”), or by a set of short-term observations (a “specific trajectory” about which they are “not too concerned”).
As for awards for climate models, I am all for that as long as they are used appropriately (in words of the RC’ers, as “scenarios” rather than “predictions” or “forecasts”). And “scenarios”, of course, cannot be tested against the real world either, as they contain no “prediction”. At most, they can be compared to past data.
Nonsense, that blog post in no way implies that the models can not be tested. They can and have been tested.
Here’s an article reporting on this year’s prestigious International Meteorological Organization winner. What did he win it for? Testing climate models.
Something that cannot be falsified cannot really be proven true either. Any theory that is not falsifiable is neither science nor useful from any perspective. — John M Reynolds
This appears to resemble theology rather than science; arguments about the number of angels on pinheads spring to mind. RealClimate’s line of reasoning also reminds me of Descartes’ proof of the existence of God.
Descartes:
1) Necessarily, God is perfect.
2) If God didn’t exist he would be less than perfect.
3) Therefore God exists.
RealClimate:
1) Necessarily the climate models cannot be falsified.
2) If the climate models didn’t represent reality, they could be falsified.
3) Therefore the climate models represent reality.
Methinks they have too much insulation — But someone needs to explain why we are freezing.
Paging Kuhn! Meet the new science, same as the old science.