problem with date in netCDF file

Don Hooper hoop at COLORADO.EDU
Tue May 31 14:56:04 EDT 2005


The sdfopen command deduces the increment from the first two time
steps.  It reads none of the other time values.  It presumes it is
a continuous time series.  Thus, it will think that the time value
following any given time step is one interval size later only.  Gaps
will not be handled correctly.  I don't know what to suggest for you,
other than splitting the file up into multiple files at the gaps in time.

-Hoop
> From owner-gradsusr at LIST.CINECA.IT  Tue May 31 12:27:13 2005
>
> Hi all
>
> I have extracted a part of daily data of ERA-40 using NCO utilities. Then
> when I load the netCDF file with sdfopen, everything is OK except that the
> time increment is worng for the second year. The first date is 1st May 1958,
> . the 184th date is 31st October 1958, then the 185th should be 1st may 1959
> but in Grads it is the 1st November 1958. When I open the netCDF file in
> Matlab, everything is OK and the the 185th date is well the 1st May 1959.
> So, the data seems to be OK but the time increment in Grads is wrong. Any
> idea about this apparent bug ? I am using Grads 1.8 (windows version)
>
> Ciao
>
> Vincent



More information about the gradsusr mailing list