We discovered an error in the figure I made for the IPCC report today. Turns out, I had accidentally overplotted summer data on top of winter data for two different methods instead of winter on winter. I know exactly what must have happened, too: I have to change a value in two different files to switch from summer to winter (stupid, but necessary because of the code structure), and I changed it in both places but didn't save both files. I almost did it again today, and I immediately knew that's what the problem was.
I feel really dumb about it, but it doesn't change any of the discussion or the conclusions, just some minor details on a plot. I suppose that's why things like this take so long to go to press, so there's time to find and correct errors like that. (Apparently, somebody else discovered that they had accidentally submitted data with winds going the wrong way around Antarctica. D'oh!)
Lots of interesting philosophizing going on over in dpolicar's journal today. Only some of it is my blather.