Glad to see you've got your chart up ImranCan.
It's a little fun to see people twisting in the wind, trying to explain away the dreadful model preformance that these GCMs have.
Too many people simply don't understand modelling, and think of it as a tool to explain the unknown. Models are worth nothing if you don't understand the underlying processes. Everybody should repeat that 10 times.
These are highly parameterized models, which means they have turned enough knobs, far enough, that they were able to replicate past climate (this is commonly known as the "calibration phase"). It doesn't say anything about the model's ability to predict future climate, just that the model can replicate past climate well. What most (all?) GCMS are missing, is the verification phase of modelling. This is where, prior to calibration, the modellers set aside a portion of the time period,which they don't even consider, or look at, during calibration. Once they are satisfied with their calibration, they test, or verify, the models ability to replicate climate in this time period that was set aside. For example, the entire modelling period of 1850-present, you'd calibrate with the time period of 1850-1980 or so....and then verify from 1980-2000. If the model was able to accurately predict the climate in the 1980-2000 using the only the parameters that you calibrated to in the 1850-1980, you'd say the model predictions were verified.
This is the only way you can test the model's ability to replicate climate. But did they do this? Of course not. They twisted enough knobs, introduced enough fudge factors (google "flux adjustment") until the GCMs matched the observed temp record from 1850-present.
And it worked..... the gullible, uneducated media and public ate it up. Wow, look how good the model fit is! The model's must be right!! Until that is, enough time had passed where you could actually go back and see how well they preformed against recent temperature data. As ImranCan is showing..........it's pathetic.
By the way, now that the difference between anomalies and trends have been explained, I guess I can show the following chart. This is global avg temp from the RSS satellites, the temperatures for May have just been released. All 5 months in 2008, have been pretty much right at average. Doesn't look like those GCMs are getting any better at predicting future climate.
I just got Office 2007.....you can make charts look real pretty now
