Hi,
Furthermore from the CMIP6 metadata requirements
https://docs.google.com/document/d/1os9rZ11U0ajY7F8FWtgU4B49KcB59aFlBVGfLC4ahXs/edit
Section 17.c.
ote that if a model has a gregorian calendar and the base time is defined as a date prior to 1582, then the user should specify 'proleptic_gregorian', rather than 'gregorian' for the calendar attribute.
That's a pretty good reason for proleptic_gregorian to be
treated correctly.
Russ
On 22/10/18 2:02 pm, Russ Fiedler
wrote:
Hi,
I've got a time series with the time definition
double time(time) ;
time:long_name = "time" ;
time:cartesian_axis = "T" ;
time:calendar_type = "PROLEPTIC_GREGORIAN" ;
time:calendar = "PROLEPTIC_GREGORIAN" ;
time:bounds = "time_bounds" ;
time:units = "days since 0001-01-01 00:00:01" ;
Now, Ferret is a bit odd in its behaviour with respect to
standard/gregorian calendars in that it uses the Proleptic
Gregorian calendar EXCEPT when the starting date
is 0001-jan-01 when it and times are after 1582 when reverts to
the cf compliant mixed gregorian/julian.
https://www.pmel.noaa.gov/maillists/tmap/ferret_users/fu_2005/msg00625.html
https://ferret.pmel.noaa.gov/Ferret/documentation/users-guide/Grids-Regions/GRIDS#_VPINDEXENTRY_634
http://cfconventions.org/Data/cf-conventions/cf-conventions-1.7/cf-conventions.html#calendar
I can handle the odd behaviour for legacy reasons but, surely,
if I specify PROLEPTIC_GREGORIAN Ferret should respect that
attribute and not muck about with the calendar.
Al the least I'd like a way to turn off this behaviour.
Maybe something like
SET AXIS/CF_COMPLIANT_CALENDAR taxis
to get proper axis specifications no matter what (or if the
"Conventions" attribute "CF-X.X" is found then do it according
to that convention).
Cheers,
Russ