I just love when scientists absolutely pwn other scientists during peer review. This happened most spectacularly with the cold fusion debacle. Oh, by the way, that crank is still a physicist in Utah, and is one of the biggest 9/11 truthers out there. Then there was the Korean cloner, who was caught just making stuff up, and so badly was his reputation tarnished, that anyone who used to work for him gets doubly reviewed, as happened earlier this year with another Korean team proving that they did clone some wolves. Now we have the Mann millenial temperature reconstruction, also known as the Hockey Stick, which is why the climate modelers over at RealClimate are pejoratively known as The Team. The reconstruction was discredited last year, but is still being used by many groups to justify alarming global warming predictions. So, I was curious how the 4th Assessment (AR4) would differ from the 3rd Assessment (TAR) in presenting the millenial temperature reconstruction. Climate Audit got the goods.
The new global millenial temperature reconstruction (the so-called "spaghetti graph") includes the downward temperature trends that were missing in the TAR. Here's the new Spaghetti Graph in AR4, and below is the Mann reconstruction featured in the TAR:
The AR4 graph doesn't look as convincing as the older graph, now does it? Especially when we see the cooling trend happening in Northern Russia and Greenland. Maybe we should be treating climate impacts much more regionally than some of these international bodies propose.
Dude - I expect more spaghetti to be made in the coming months as more data becomes available.
ReplyDeleteGeh, who can see any trend at all in these charts? Plus, the data from the older proxies become less and less reliable as newer data sources get added to the models. Time to add more fudgings, I mean, more feedback forcings, to the models, right?
Dude:
ReplyDeleteClimate Science had a good post today on why the climate models always show an increase in temperature after a few decades: they're designed that way. They only give a good answer depending on the initial variables for about twenty years. But "good" doesn't mean that much, because the regional feedback forcings haven't been programmed in. There's also good explanations for why weather predictions are good only for about 10-15 days.