Tetsuya,
>Date: Tue, 13 Jul 1999 14:34:06 -0500 (CDT)
>From: tetsuya sakashita <tsakashi@xxxxxxxxxxxxxxxxx>
>Subject: java3D
In the above message, you wrote:
> These are the errors i receive when trying to run a 3D application on
> VisAD.ss.SpreadSheet. The files were taken from the VisAD spreadSheet web
> page.
>
> rad.nc (this finally worked with these errors):
>
...
>
> Couldn't decode attribute rpt_date:units="YYYYDDMM UTC": Unit not in
> database
> Couldn't decode attribute rpt_time:units="HHMM UTC": Unit not in database
> Couldn't decode attribute opstat:units="ennumerated in rarep.h": Unit not
> in database
> Couldn't decode attribute object:units="ennumerated in rarep.h": Unit not
> in database
> Couldn't decode attribute precip_type:units="ennumerated in rarep.h": Unit
> not in database
>
> Couldn't decode attribute intensity:units="ennumerated in rarep.h": Unit
> not in database
> Couldn't decode attribute trend:units="ennumerated in rarep.h": Unit not
> in database
> Couldn't decode attribute sunSensorIntensity:units="watts/meter^2": Unit
> not in database
> Couldn't decode attribute timeHHMMSS:units="hour.minute.second": Unit not
> in database
> Couldn't decode attribute
> atmosphericRadiance:units="milliwatt/meter^2/steradian/centimete
> r^-1": Unit not in database
I can speak authoritatively on the above "errors". Basically,
they are warning messages about the units in the netCDF file not
adhering to the netCDF/UDUNITS standard for the unit attribute.
That standard is documented in the netCDF User's Guide and closely
follows the US standard for unit specifications, as promulgated by
the NIST, which, in turn, closely follows the ISO standard for unit
specifications. Further information on unit specifications can be found
at <http://physics.nist.gov/cuu/index.html>.
Here are the specific problems with the above unit specifications:
YYYYDDMM UTC not a unit specification
HHMM UTC not a unit specification
ennumerated in rarep.h not a unit specification
watts/meter^2 should be "W/m2" or "watts/meter2"
hour.minute.second not a unit specification
milliwatt/meter^2/steradian/centimeter^-1
should be "mW m-2 sr-1 cm" or
"milliwat meter-2 steradian-1 centimeter"
--------
Steve Emmerson <http://www.unidata.ucar.edu>
