Re: Time Unit with Descriptor Files

[Joseph Mclean <mclean@xxxxxxxxxxxxx>]

Hi Robin,
comments at the bottom.

Robin Kovach wrote:

> Hi again,
>
> I am trying to us a descriptor file to link two data files together under one
> data set heading. I'm pretty sure that my problem lies in the way I define the
> time axis in either the netCDF files, the descriptor file, and/or the XML file.
>
> The error is:
>
> Adding an acceptable error string: "*** NOTE: ".
>         NOAA/PMEL TMAP
>         Program FERRET
>         Version 5.22 - 07/27/00
>          7-Nov-01 10:14
>
> yes? cancel mode verify
> yes? cancel mode interp
> yes? canc data/all
> yes? go std_initialize "/obs/NSIPP/MODEL/POSEIDON/GLOBAL/2D/link" "1" "1"
> "surfaceheight"
> yes? set region/x="280":"380"/y="-30.0":"30.0"/k=1/t="16-Jan-1996":"16-Jan-1996"
> yes? set win/size=.5
> yes? set win/asp=0.9
> yes? GO "/tmp/lasgo270355527954.jnl"
>  **TMAP ERR: 1st timestep in datafile different from descriptor
>              Expected: 15 Found: 1.5625
>              Data set: /obs/NSIPP/MODEL/POSEIDON/GLOBAL/2D/link.des
>              Data file: /obs/NSIPP/MODEL/POSEIDON/GLOBAL/2D/1997_2d.nc
>  **TMAP ERR: 1st timestep in datafile different from descriptor
>              Expected: 15 Found: 1.5625
>              Data set: /obs/NSIPP/MODEL/POSEIDON/GLOBAL/2D/link.des
>              Data file: /obs/NSIPP/MODEL/POSEIDON/GLOBAL/2D/1997_2d.nc
>
> LAS site: http://mohawk.gsfc.nasa.gov/las/
>
> So I have two files (1996_2d.nc and 1997_2d.nc) that have the same variables and
> the same LON and LAT grid. The TIME variable for both files is:
>
> TIME = 366, 1096.48498535156, 1826.96997070312, 2557.455078125,
>     3287.93994140625, 4018.42504882812, 4748.91015625, 5479.39501953125,
>     6209.8798828125, 6940.365234375, 7670.85009765625, 8401.3349609375 ;
>
> and for 1996_2d.nc     double TIME(TIME) ;
>                        TIME:TIME_origin = "01-JAN-1996 00:00:00" ;
>                        TIME:units = "hours since 1996-01-01 00:00:00" ;
>
> and for 1997_2d.nc     double TIME(TIME) ;
>                        TIME:TIME_origin = "01-JAN-1997 00:00:00" ;
>                        TIME:units = "hours since 1997-01-01 00:00:00" ;
>
> I also have copies of the descriptor file (link.des) and las.xml and
> POSEIDON_Global_2d.xml which you can access from
> http://mohawk.gsfc.nasa.gov/~kovach/LAS_Test_Pages.html
>
> I've have looked at some example .des files and the archive help on "ui for
> single variable stored in multiple files"...
>
> I am still a bit confused on the D_TOTIME adn D_TIME_UNIT used in the .des files
> and how they correspond with the S_START and S_END there and the TIME info for
> the grids in the .XML file.
>

D_T0TIME is the starting point of the descriptor timesteps.  D_TIME_UNIT is the unit
of each time step in seconds.
Your NetCDF files use MONTH_REG as the time axis and this defines one month as
365.2425/12. This equals 2629746 seconds. Try changing your descriptor file to read:

$BACKGROUND_RECORD
   D_TITLE              = ' ',
   D_T0TIME             = '15-dec-1995 00:00:00',
   D_TIME_UNIT          = 2629746,
   D_TIME_MODULO        = .false.,

Then TIMESTEP 1 will be one reg_month from 15-dec-1995.  I think this will work for
you.  Let me know.

- Joe


--
Joseph McLean
NOAA/PMEL/TMAP
(206)526-4286