Interesting article. I disagree in methodology with the article on several fronts however. It basically boils down to them selecting IPCC predictions after the

Interesting article. I disagree in methodology with the article on several fronts however. It basically boils down to them selecting IPCC predictions after the fact, and even after cherry picking what they compare with, they still adjust the data. Then they also cherry-pick what dataset to compare against. Even that is only enough to raise the observed level just above the lower bounds. Even then, I find that the data presented in their graphs does not look at all like a "remarkably good" prediction. But I disagree strongly with this way of evaluating the data. When evaluating a prediction made in the past you do not get to cherry pick out of a set of predictions, you ONLY get to pick the one you found most likely in the past. Second you most certainly do not get to adjust the prediction made in the past in any way. The article violates both. That means if you want to convince me, please unadjusted business-as-usual scenario only. From the article: > Figure 2 accounts for the lower observed GHG emissions than in the IPCC BAU projection, and compares its 'Best' adjusted projection with the observed global surface warming since 1990. In case you're wondering, adjusted means adjusted by the author of the article (in december 2012). That's just unfair. Also "Best" does NOT mean the best scenario as set forth in the IPCC first assessment report, but the one he finds agrees with the data. This is bullshit. Cherry-picking predictions. Adjusting the results of you prediction AFTER the results are known. Sorry, but let me summarize my attitude here : fuck off. Furthermore he cheats by omission : there are 2 major temperature records, and guess what, he uses the warmest measurement. Not mentioned anywhere in the article. From the IPCC first assessment report (policymaker's section): > under the IPCC Business-as-Usual (Scenario A) emissions of greenhouse gases, a rate of increase of global mean temperature during the next century of about 0 3°C per decade (with an uncertainty range of 0 2°C to 0 5°C per decade), this is greater than that seen over the past 10,000 years This will result in a likely increase in global mean temperature of about 1°C above the present value by 2025 and VC before the end of the next century The rise will not be steady because of the influence of other factors. If you simply take the recommendation from the policymaker's section, and evaluate the temperatures against that, an entirely different picture emerges. I don't intend to make a graph here, but let's summarize : IPCC prediction in BAU scenario : 0.2 to 0.5 degrees rise per decade from 1990 to 2010 Now obviously there's various ways to calculate the actual temperature difference, but none of them see 0.2 degrees rise: 1990 vs 2000 : 0.03 degrees warmer in 2000 (and no, 1990 was not particularly warm, although 2000 appears to have been a relatively cold year. Taking 1999 or 2001 or even 2002 still doesn't fix their prediction though) Average of the 90s versus average of 2000-2010: 0.13 degrees rise (this I find the "fairest" way of measuring, but it doesn't allow us to evaluate 2000-2010 yet) Average of the first 2 years of the decade (meaning you can check accuracy for 2000-2010 as well) 1990 -> 2000 : 0.15 degrees warmer in 2000 2000 -> 2010 : 0.05 degrees warmer in 2010 I have made my own graphs of the situation (Ipython rocks !), and overlayed them on the actual pictures from the IPCC publications, and the simple observation is : we are currently below the 95% certainty interval of the BAU scenario in the IPCC FAR. And it is a matter of 0.9 degrees in 2010 (that is, the real observed temperature was 0.9 degrees below the lower bound set forth in the IPCC's report), which is more than the predicted rise during that period (in other words : the error is bigger than the prediction). So sorry to say, but this article fixes up the data by cherrypicking from the FAR, adding in "adjustments" and yes, it seems to be able to come to a reasonable picture. I call bullshit on that methodology though, for obvious reasons. If you're defending your prediction made in 1990 you do not get to revise said prediction in any way. You also don't get to make 5 predictions in 1990, then pick the "average" from 2 of them and show it matches the data. So perhaps I should be more explicit : comparing the current satellite temperature record to the IPCC FAR report's "Business as usual" scenario, without adjusting it, the temperature has been below the lower bound of the 95% certainty interval since 1994. After 1996 there hasn't been a single temperature peak in the IPCC's predicted bounds. They were wrong, and not just a little bit. A similar problem occurs between my own evaluation of the IPCC's second third and fourth report (and the article doesn't look at the fifth, probably because there's not that much data to evaluate it). I find this article woefully unconvincing.