No, they do a HORRIBLE job of hindcasting temperatures of both the atmosphere (with a modeled warming rate that is 1.6X the observed rate) and the ocean (with a modeled warming rate which is 2X the observed warming rate):GRA wrote:Only loosely related to the discussion, but one of my favorite quotes by a well-known climate scientist (whose name escapes me for the moment) is "All models are wrong; some are useful". Or in other words, all models (climate in this case) simplify complex and often only partially understood phenomena, and while they are inherently incomplete and 'inaccurate', they can nevertheless provide useful insight as to what is likely to occur, if they do a good job of reflecting what has actually occurred.
The above graphs showed the FAR predictions (2.8X the observed warming), here is what AR5 shows (at 1.67K/century):
Here are the sea-surface predictions versus measurements:
They do not have the ability to model cloud effects beyond basic parameterization, so they do not have the ability to even model the impact of the clouds on temperature. As such, they are not fit for purpose.
BTW, scientists are just now starting to have an ability to model some of the impacts of clouds. They are finding that clouds have a dominant effect on the climate.
But ignoring massive failures of the models to predict (or even hindcast) what is going on does not lead to furthered understanding. Science only advances when investigators acknowledge the failings of the current beliefs and look for corrections. In fact, that is what science is all about.GRA wrote:Expecting 100% accuracy is a pipe dream.
Fortunately, there is work being done to understand the errors in the current models in terms of their mathematics, structure, and more structure. When a corrected model is constructed without these significant errors, you find that the impact of CO2 on the temperature of the planet is not at all alarming. Specifically, rather than finding that the water vapor emissions layer ascends with increasing CO2 concentration, we learn that it actually descends. That is why there is no "hotspot." (It was never "missing", the models were simply incorrect.) The conclusion is that the impact of a doubling of CO2 on the Earth's temperature is extremely low:
This is all summarized in a PDF here.Dr. David Evans wrote:- The ECS might be almost zero, is likely less than 0.25C, and most likely less than 0.5C.
- The fraction of global warming caused by increasing CO2 in recent decades, mu, is likely less than 20%.
- The CO2 sensitivity, lambdaC, is likely less than 0.15C W^-1 m^2 (less than a third of the solar sensitivity).
Note that this work does not argue ANY against any of the basic tenets of modern climate science about the physics of the system. It simply corrects the failings of the current model to determine how it has gone wrong.
We've known for a long time that the climate models in widespread use today predict WAY more warming than actually occurs. Now we know, in detail, why they are wrong. It's time to move on and actually try to predict what will happen rather than simply making alarming claims about the future.