[CF-metadata] Recording "day of year on which something happens"

Jonathan Gregory j.m.gregory at reading.ac.uk
Thu Mar 16 09:31:51 MDT 2017

Dear Jon et al.

>    Is cell_methods necessary in this case
>    (for the time axis bounds) – probably not since this isn’t a
>    statistical quantity like an average, but a value that’s
>    “representative” of the year.

It's always recommended to include cell_methods, because the default depends
on whether the quantity is intensive or extensive in the dimension concerned,
and that may not be obvious. If it's not "point", it should be "sum". That
method (Appendix E) means it applies to the cell as a whole.

It would be great if someone had the energy and interest to summarise this
discussion as a FAQ to be addded to http://cfconventions.org/faq.html,
because it's likely it'll be asked again, since it's been asked before!

Best wishes


More information about the CF-metadata mailing list