Re: Err, time problems

[Steve Hankin <hankin@xxxxxxxxxxxxx>]

Hi Steve,

Ferret reads the COARDS (Unidata) standard time encoding wherein time is in
true binary (not yyyyddd) with units and offset determined by the units
attribute -- e.g. units = "seconds since 1-1-1900".  See COARDS
documentation at
http://www.unidata.ucar.edu/packages/netcdf/conventions.html

Ferret also has a hack buried in it so that it understands
units="yyyymmddhhmmss" (The units must be that text exactly and the time,
itself, must be typed as double.) With this as a model it would be easy to
hack in support for YYYYDDD. (It could even be made general by someone who
would take the trouble to do so. :-} )

The Ferret source code is available.  Contact us at ferret@pmel.noaa.gov if
you want to pursue this.

    - steve

---------------------------------------

Steve Cocks wrote:

> Hello All;
>
> I have been concatenating numerous files (multiple time steps) into a
> single large file (of only one day time steps).  In the new netcdf file,
> I have the time variable listed in YYYYDDD format (DDD is julian date).
> Err, seems ferret doesn't like this choice for a time variable.  Does
> ferret only read certain types of time variable formats?  Is there a way
> to get ferret to recognize YYYYDDD format at all?  Thanks in advance.
>
> V/R
>
> STEPHEN B. COCKS, Capt, USAF
> AFIT PhD Student
> Texas A&M University

--
Steve Hankin
NOAA/PMEL, 7600 Sand Point Way NE, Seattle, WA 98115-0070
ph. (206) 526-6080 -- FAX (206) 526-6744