Dear Jim and Seth I agree that option 1 of sect 7.4 is the right one for this case i.e. time: mean within years time: mean within years. That describes what you have done. Within each year, you make a mean over (for instance) Jan 1 0000 to Jan 2 0000. Then you have a mean over years of these daily means.
I agree you could encode time with the noleap/365_day calendar if you don't want to provide a 29 Feb climatological mean; if you do want to provide one, you could code the time with the all_leap/366_day calendar. It's just an encoding really, and doesn't imply anything about the original data; when forming a climatological mean for the real world you are inevitably making some compromises with the precision of the calendar. Cheers Jonathan _______________________________________________ CF-metadata mailing list [email protected] http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
