Model to data temperature comparisons – how well did they do?

In the last few months, we’ve heard the comment “Notwithstanding the dramatic increases in man-made CO2 emissions over the last decade, the world’s warming has stopped” pushed around with alarming regularity. We discussed this in a previous posting (‘Abbott’s climate change policy is “bullshit”‘) which included a great analysis by Tamino, who demonstrated pretty well that the temperature this decade did exactly as expected – it’s getting warmer (see also “WMO finds 2000–2009 the warmest decade; so much for that “global warming pause” meme“)

Another one of those re-occurring memes that’s been posted here on the comments section of Climate Shifts is that the “computer climate models (are) known to not be able to predict future climate very well” and “observational evidence does not match the models to date“. Considering that it’s almost the end of the first decade of the 21st century, it seems like a good time to be checking our predictions and future projections of climate change. Enter Gavin from Real Climate, who points out the vital importance of “updat(ing) all of the graphs of annual means with another single datapoint”:

Above is the annual mean anomalies from the IPCC AR4 models (black line) and their 95% envelope (grey shading) and the surface temperature records from HadCRUT3v (red line) and GISTEMP (blue lines). As Gavin points out:

As you can see, now that we have come out of the recent La Niña-induced slump, temperatures are back in the middle of the model estimates. If the current El Niño event continues into the spring, we can expect 2010 to be warmer still.

So far, the model predictions are looking solid. We can pretty much debunk the meme “observational evidence does not match the models to date“.

Next up, Gavin deals with a pretty key question: how does the oldest of all climate models hold up? Below are three scenarios from Hansen’s 1988 paper on climate projections:

How about that “computer climate models (are) known to not be able to predict future climate very well” meme? It seems like even the first GCM model did a pretty good job of predicting the increases in temperature pretty well, although the B & C scenarios are a little warm compared to the actual surface temperature records.  To conclude:

…despite the fact these are relatively crude metrics against which to judge the models, and there is a substantial degree of unforced variability, the matches to observations are still pretty good, and we are getting to the point where a better winnowing of models dependent on their skill may soon be possible. (Read more over @ Real Climate)

Leave a Reply

Your email address will not be published. Required fields are marked *