But what you don't see are radical changes in the in the projections as a result.
------------
We recently found out that climate models vastly underestimate precipation increases wrt temp increases...
If you are referring to "Cloud and radiation budget changes associated with tropical intraseasonal oscillations"
http://www.agu.org/pubs/crossref/2007.../2007GL029698.shtml
I doubt the authors would agree that current models are "getting feed-backs like clouds completely wrong". Climate modelers know (and acknowledge) that cloud interactions are one of the less well understood inputs, but also that unrealistically large changes to current understandings would be needed to significantly affect their results - IMO this s another example of a case where critics of the models are just not being realistic about the likely results of improved modeling. ---------
"Meanwhile, methane in atmosphere has *levelled off*
See:
http://www.noaanews.noaa.gov/stories2006/s2709.htm
“Sure, the models keep getting refined.
But what you don’t see are radical changes in the in the projections as a result. “
Disagree. The recent Schwartz paper model wrt heat sink effect of oceans cuts AGW temperature sensitivity of CO2 forcings by two-thirds. Just by changing one variable, the time constant of ocean heat content. And there are dozens of other knobs to tweak that could similar double or half AGW impact. I’d say it is a significant departure model-wise. Even a “radical change”. see:
http://climatesci.colorado.edu/2007/08/20/new-paper-on-the-diagnosis-and-significance-of-ocean-heat-content-changes/
it’s been criticized by James Annan on his blog for the time constant assumption. Probably Annan is right that there is no time constant that fully represents the system, a truer model (like electronic circuit model) would have more than one moment; otoh, the paper opens up the question about assumptions of ocean thermal content in GCMs. Its given as 15 years? Why?
There are model claims that defy credulity. There are uncertainties, like cloud feedback, that are bigger than CO2 direct impact as a forcing. And the error bars on models and estimates are bigger than the AGW alarmists and IPCC let on. The science is far from settled on this.
The comment on precipitation was not referred to that paper specifically, but another paper that measured precipitation changes. GCMs are off by a factor of 3 in modelling precipitation sensitivities. This has been reported here on this blog as well:
http://climatesci.colorado.edu/
I believe 50 years from now, we will be seeing a similar chart for CO2.