[ncl-talk] NCL to replace CDO

[David Brown]

Hi Marston,

Thanks for the files.
I see the experimental variables in only one of the files you sent me:
ICMGGECE3+199001
This file contains unknown variables with parameters numbers ranging
from 90 - 100. You can see them easily with the following command:

ncl_filedump ICMGGECE3+199001.grb |grep VAR_

warning:NclGRIB: Unknown grib parameter number detected (90, center
98, table version 128 grib record 1), using default variable name
(VAR_90)
....
      float VAR_90_GDS4_SFC ( initial_time0_hours, g4_lat_1, g4_lon_2 )
      float VAR_91_GDS4_SFC ( initial_time0_hours, g4_lat_1, g4_lon_2 )
      float VAR_92_GDS4_SFC ( initial_time0_hours, g4_lat_1, g4_lon_2 )
      float VAR_93_GDS4_SFC ( initial_time0_hours, g4_lat_1, g4_lon_2 )
      float VAR_94_GDS4_SFC ( initial_time0_hours, g4_lat_1, g4_lon_2 )
      float VAR_95_GDS4_SFC ( initial_time0_hours, g4_lat_1, g4_lon_2 )
      float VAR_96_GDS4_SFC ( initial_time0_hours, g4_lat_1, g4_lon_2 )
      float VAR_97_GDS4_SFC ( initial_time0_hours, g4_lat_1, g4_lon_2 )
      float VAR_98_GDS4_SFC ( initial_time0_hours, g4_lat_1, g4_lon_2 )
      float VAR_99_GDS4_SFC ( initial_time0_hours, g4_lat_1, g4_lon_2 )
      float VAR_100_GDS4_HYBL ( initial_time0_hours, lv_HYBL3,
g4_lat_1, g4_lon_2 )

One of these variables,
VAR_100_GDS4_HYBL, is a 4d variable with dimensions (
initial_time0_hours, lv_HYBL3, g4_lat_1, g4_lon_2 ) sized [ 247, 91,
256, 512 ].
This variable alone is composed of 22,477 GRIB records (247 * 91)
There are 18 other surface-only variables in the file, each composed
of 247 records (all the time steps).
This adds up to 26,923 records. The record-oriented GRIB tool 'wgrib'
confirms that the file contains 26,923 records, meaning that NCL has
accounted for every record in the file.

ICMSHglx2+200701 is the spectral file with 2 variables at Isobaric
level 1000 (hPa).

ICMGGglx2+200701 contains 8 variables at 8 timesteps (64 GRIB
records), all surface only. There are no unknown experimental
variables. It contains 64 records, (8 * 8), also confirmed by wgrib.

You should certainly be able to regrid these files to lat/lon using
the ESMF facilities provided by NCL.
 -dave







On Tue, Jun 27, 2017 at 6:05 PM, David Brown <dbrown at ucar.edu> wrote:
> Thanks Marston.
>  -dave
>
> On Tue, Jun 27, 2017 at 4:03 PM, Marston Johnston <shejo284 at gmail.com> wrote:
>> Hi,
>>
>> ifs_grib_types.zip is uploaded to your ftp: incoming.
>>
>> /M
>>
>> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>> Marston S. Johnston, PhD
>> Department of Earth Sciences
>> University of Gothenburg, Sweden
>> Email: marston.johnston at gu.se
>> SkypeID: marston.johnston
>> Phone: +46-31-7864901
>> Only the fruitful thing is true!
>> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>>
>>
>> On 2017-06-27, 22:34, "David Brown" <dbrown at ucar.edu> wrote:
>>
>>     ftp.cgd.ucar.edu
>>
>>