[gradsusr] variable tdef in the netcdf file

[Jeff Duda]

For netCDF, GrADS doesn't read the time from the file itself if you use a
control file. So I would write a control file and then rename the files to
some arbitrary interval that is close to the actual valid times of the
files.

If you need exact and non-constant timing, then you have two other options
I can see
1) Use CHSUB templating (%ch) to manually type each time stamp into a
single control file and then use the time index to loop over those.
2) If the times change frequently or writing them all out is too tedious,
then you can use control *files* (plural) by using an external scripting
source such as Linux shell scripting to loop over individual files, making
a separate control file (or using sdfopen) to read/plot from each file in
the loop.

Jeff Duda

On Wed, Apr 29, 2020 at 12:23 AM Gevorgyan Artur <agm86 at yandex.ru> wrote:

> Dear Colleges,
>
> I have observed data with variable (not constant) observational time
> intervals. The data are in netcdf format, and the observational times are
> prescribed defined the time variable, and time:units = "seconds since
> 1970-01-01"? How can I plot my data using those observational times?
>
> Kind regards,
> Artur
>
> _______________________________________________
> gradsusr mailing list
> gradsusr at gradsusr.org
> http://gradsusr.org/mailman/listinfo/gradsusr
>


-- 
Jeff Duda, Research Scientist
University of Colorado Boulder
Cooperative Institute for Research in Environmental Sciences
NOAA/OAR/ESRL/Global Systems Laboratory
Boulder, CO
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://gradsusr.org/pipermail/gradsusr/attachments/20200429/a4106e0d/attachment.html>