You're basically saying that we have do a lot of weird things (cooking, processing) the data before we can test GW. That's not what I like to hear when evaluating a hypothesis. It is a bad sign, and worthy of raising suspicion.
No, I am saying that compiling and evaluating this massive a data set is more than a full time job. The quality of the data needs to be checked and rechecked. There are spatial and temporal issues to be addressed. Even the satellite data is very hard to interpret, no single (solar looking) satellite has been up for more than ten years, there are calibration issues and intercalibration issues.
I have looked at the pnas article, and it is a good one. Of course it doesnt say anything like the outrageous claim I called you on earlier. Ive punished you enough on that one, Ill stop harping on it now.
PNAS is not peer reviewed in the traditional sense, the articles are called for by the editor. But it is well respected, and the editors are top scientists in their respective fields. So it is archived, I fully accept it, and it would have a decent ISI rating.
It is a very long timescale modeling exercise and fully agrees with the forcings I posted from the Hansen article earlier. Everyone agrees that the sun is important, Milankovitch cycles are at the heart of climate research. A good climate model must include solar forcing, along with all the rest.
I even posted a number of other articles that talk about a solar link beyond the direct forcing (secondary effects through solar proton events). Good articles like the one you link to above. I am far from believing that human understanding of climate is complete.
You tone of debate has improved, I appreciate that. If you go back through the posts you will notice that I tend to respond in kind. I have read your papers, when I could access them. I didnt get to the pnas article because I was busy. It is a good article.
I totally agree that observations take precedent. Ive tried to bring them into this debate (my forcings stuff, and my continued appeal to start debating individual process models).
The thing is that CO2 isnt the only thing affecting climate, a large part of the correlation in the graph I posted above is due to other factors. Remember that correlation does not imply causation, one needs to posit a mechanism and then test that mechanisms independent variables in a controlled way. That is what is done with process models.
I would be interested to see the AAPG article, cant you find a PDF to post?
Edit: My favorite example of correlation and causation is that crime rates are strongly correlated with the density of churches.
Of course this is because there are lots of churches in inner cities where crime rates are also high, but it shows that you need to look at mechanisms and not data in isolation.
Lies, danm lies, and statistics...