help reading GRIB1 dataset

[Dan Kokron]

I have a few datasets in GRIB1 format that I would like to use in GMT.
grdimage is having some trouble.  I've attached the GRIB file.  It
contains global soil texture data in 9 bins labeled 1-9 (plus zero for
water tiles so make it ten bins).

Is it a grib file and which version?

degrib/bin/degrib global_soiltype.1x1.grb -I
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
MsgNum, Byte, GRIB-Version, elem, level, reference(UTC), valid(UTC), Proj(hr)
1.0, 0, 1, SNOEV="Snow sublimation heat flux [W/m^2]", 0-SFC, 01/01/2000 00:00, 01/01/2000 00:00, 0.00
        ^
        |
     Version 1


What does grdinfo think about this file?

grdinfo global_soiltype.1x1.grb=gd
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
global_soiltype.1x1.grb: Title: Grid imported trhough GDAL
global_soiltype.1x1.grb: Command:
global_soiltype.1x1.grb: Remark:
global_soiltype.1x1.grb: Gridline node registration used
global_soiltype.1x1.grb: Unrecognized grid file format! Probably not a GMT grid
global_soiltype.1x1.grb: x_min: -179 x_max: 180 x_inc: 1 name: x nx: 360
global_soiltype.1x1.grb: y_min: -90 y_max: 89 y_inc: 1 name: y ny: 180
global_soiltype.1x1.grb: z_min: 0 z_max: 9 name: z
global_soiltype.1x1.grb: scale_factor: 1 add_offset: 0

Lets make a color table

grd2cpt -Cwysiwyg -E9 global_soiltype.1x1.grb=gd
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
Warning! : Can't figure out time unit of 4
#       cpt file created by: grd2cpt -Cwysiwyg -E9 global_soiltype.1x1.grb=gd
#COLOR_MODEL = RGB
#
0       64      0       192     1.125   64      0       192
1.125   0       128     255     2.25    0       128     255
2.25    64      224     255     3.375   64      224     255
3.375   64      255     192     4.5     64      255     192
4.5     192     255     64      5.625   192     255     64
5.625   255     224     64      6.75    255     224     64
6.75    255     32      64      7.875   255     32      64
7.875   255     160     255     9       255     160     255
B       0       0       0
F       255     255     255
N       128     128     128

Looks good, lets make an image.

grdimage global_soiltype.1x1.grb=gd -JM15c -R-179/180/-89/89 -Cglobal_vegfrac.0.144.decpercent.cpt -V > global_soiltype.1x1.ps
grdimage: Allocates memory and read data file
GMT_grd_is_global: yes, longitude cells span exactly 360
gdalread: Unsupported data type
grdimage: project grid files
grdimage: Given domain implies x_inc = 0.0164499
grdimage: Grid projection from size 360x180 to 360x180
GMT_grd_is_global: yes, longitude cells span exactly 360
GMT_boundcond_param_prep determined edgeinfo: gn = 0, gs = 1, nxp = 360, nyp = 0
grdimage: Evaluate pixel colors
grdimage: Creating PostScript image 24-bit [colorimage]
pslib: Colormap of 14 colors created
pslib: Image depth reduced to 4 bits
pslib: LZW compressed 32400 to 369 bytes

The trouble is that the postscript image is empty.  I have confirmed
with other programs that the file does contain the expected data.

I am using GMT-4.5.1 with GDAL-1.6.2 on a 32-bit Linux box.  GDAL was
configured with internal support for GRIB format.  The table of
supported format located at http://www.gdal.org/formats_list.html
indicates it supports both GRIB1 and GRIB2.

Any ideas?

--
Dan Kokron
Global Modeling and Assimilation Office
NASA Goddard Space Flight Center
Greenbelt, MD 20771
Daniel.S.Kokron@...
Phone: [phone]
Fax:   [phone]

To unsubscribe, send the message "signoff gmt-help" to listserv@...

[Joaquim Luis]

Dan Kokron wrote:
> I have a few datasets in GRIB1 format that I would like to use in GMT.
> grdimage is having some trouble.  I've attached the GRIB file.  It
> contains global soil texture data in 9 bins labeled 1-9 (plus zero for
> water tiles so make it ten bins).

Dan,

For one side there is a bad GDAL behavior. Using your commands, I got something in
the PS file. And part of it were the "Warning! : Can't figure out time unit of 4"
that are printed by GDAL. So it means that GDAL is issuing warnings into stdout
instead of stderr. I can try fill another ticket in GDAL, but honestly, if it is
to be left there ignored for another couple of years ...

Now, aside from that (serious) trouble, you can make it work by converting the
grib file into netCDF with

grdreformat global_soiltype.1x1.grb=gd gribed.grd

Than grdimage works ... under the condition that you change either the projection or
the map limits. The following worked for me.
... -JM15c -R-179/180/-87/87 ...
or
... -JX15cd -R-179/180/-89/89 ...

I don't know why there is no image with lat > +/- 87 but you shouldn't use Mercator
to plot at such high latitudes.

Joaquim

>
> Is it a grib file and which version?
>
> degrib/bin/degrib global_soiltype.1x1.grb -I
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> MsgNum, Byte, GRIB-Version, elem, level, reference(UTC), valid(UTC), Proj(hr)
> 1.0, 0, 1, SNOEV="Snow sublimation heat flux [W/m^2]", 0-SFC, 01/01/2000 00:00, 01/01/2000 00:00, 0.00
>         ^
>         |
>      Version 1
>
>
> What does grdinfo think about this file?
>
> grdinfo global_soiltype.1x1.grb=gd
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> global_soiltype.1x1.grb: Title: Grid imported trhough GDAL
> global_soiltype.1x1.grb: Command:
> global_soiltype.1x1.grb: Remark:
> global_soiltype.1x1.grb: Gridline node registration used
> global_soiltype.1x1.grb: Unrecognized grid file format! Probably not a GMT grid
> global_soiltype.1x1.grb: x_min: -179 x_max: 180 x_inc: 1 name: x nx: 360
> global_soiltype.1x1.grb: y_min: -90 y_max: 89 y_inc: 1 name: y ny: 180
> global_soiltype.1x1.grb: z_min: 0 z_max: 9 name: z
> global_soiltype.1x1.grb: scale_factor: 1 add_offset: 0
>
> Lets make a color table
>
> grd2cpt -Cwysiwyg -E9 global_soiltype.1x1.grb=gd
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> Warning! : Can't figure out time unit of 4
> #       cpt file created by: grd2cpt -Cwysiwyg -E9 global_soiltype.1x1.grb=gd
> #COLOR_MODEL = RGB
> #
> 0       64      0       192     1.125   64      0       192
> 1.125   0       128     255     2.25    0       128     255
> 2.25    64      224     255     3.375   64      224     255
> 3.375   64      255     192     4.5     64      255     192
> 4.5     192     255     64      5.625   192     255     64
> 5.625   255     224     64      6.75    255     224     64
> 6.75    255     32      64      7.875   255     32      64
> 7.875   255     160     255     9       255     160     255
> B       0       0       0
> F       255     255     255
> N       128     128     128
>
> Looks good, lets make an image.
>
> grdimage global_soiltype.1x1.grb=gd -JM15c -R-179/180/-89/89 -Cglobal_vegfrac.0.144.decpercent.cpt -V > global_soiltype.1x1.ps
> grdimage: Allocates memory and read data file
> GMT_grd_is_global: yes, longitude cells span exactly 360
> gdalread: Unsupported data type
> grdimage: project grid files
> grdimage: Given domain implies x_inc = 0.0164499
> grdimage: Grid projection from size 360x180 to 360x180
> GMT_grd_is_global: yes, longitude cells span exactly 360
> GMT_boundcond_param_prep determined edgeinfo: gn = 0, gs = 1, nxp = 360, nyp = 0
> grdimage: Evaluate pixel colors
> grdimage: Creating PostScript image 24-bit [colorimage]
> pslib: Colormap of 14 colors created
> pslib: Image depth reduced to 4 bits
> pslib: LZW compressed 32400 to 369 bytes
>
> The trouble is that the postscript image is empty.  I have confirmed
> with other programs that the file does contain the expected data.
>
> I am using GMT-4.5.1 with GDAL-1.6.2 on a 32-bit Linux box.  GDAL was
> configured with internal support for GRIB format.  The table of
> supported format located at http://www.gdal.org/formats_list.html
> indicates it supports both GRIB1 and GRIB2.
>
> Any ideas?
>

To unsubscribe, send the message "signoff gmt-help" to listserv@...

[Dan Kokron]

Thanks for the help.  Unfortunately, we climate folks use Mercator for
almost everything.

Dan
On Mon, 2009-10-19 at 11:52 -0500, Joaquim Luis wrote:

> Dan Kokron wrote:
> > I have a few datasets in GRIB1 format that I would like to use in GMT.
> > grdimage is having some trouble.  I've attached the GRIB file.  It
> > contains global soil texture data in 9 bins labeled 1-9 (plus zero for
> > water tiles so make it ten bins).
>
> Dan,
>
> For one side there is a bad GDAL behavior. Using your commands, I got something in
> the PS file. And part of it were the "Warning! : Can't figure out time unit of 4"
> that are printed by GDAL. So it means that GDAL is issuing warnings into stdout
> instead of stderr. I can try fill another ticket in GDAL, but honestly, if it is
> to be left there ignored for another couple of years ...
>
> Now, aside from that (serious) trouble, you can make it work by converting the
> grib file into netCDF with
>
> grdreformat global_soiltype.1x1.grb=gd gribed.grd
>
> Than grdimage works ... under the condition that you change either the projection or
> the map limits. The following worked for me.
> ... -JM15c -R-179/180/-87/87 ...
> or
> ... -JX15cd -R-179/180/-89/89 ...
>
> I don't know why there is no image with lat > +/- 87 but you shouldn't use Mercator
> to plot at such high latitudes.
>
> Joaquim
>
> >
> > Is it a grib file and which version?
> >
> > degrib/bin/degrib global_soiltype.1x1.grb -I
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > MsgNum, Byte, GRIB-Version, elem, level, reference(UTC), valid(UTC), Proj(hr)
> > 1.0, 0, 1, SNOEV="Snow sublimation heat flux [W/m^2]", 0-SFC, 01/01/2000 00:00, 01/01/2000 00:00, 0.00
> >         ^
> >         |
> >      Version 1
> >
> >
> > What does grdinfo think about this file?
> >
> > grdinfo global_soiltype.1x1.grb=gd
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > global_soiltype.1x1.grb: Title: Grid imported trhough GDAL
> > global_soiltype.1x1.grb: Command:
> > global_soiltype.1x1.grb: Remark:
> > global_soiltype.1x1.grb: Gridline node registration used
> > global_soiltype.1x1.grb: Unrecognized grid file format! Probably not a GMT grid
> > global_soiltype.1x1.grb: x_min: -179 x_max: 180 x_inc: 1 name: x nx: 360
> > global_soiltype.1x1.grb: y_min: -90 y_max: 89 y_inc: 1 name: y ny: 180
> > global_soiltype.1x1.grb: z_min: 0 z_max: 9 name: z
> > global_soiltype.1x1.grb: scale_factor: 1 add_offset: 0
> >
> > Lets make a color table
> >
> > grd2cpt -Cwysiwyg -E9 global_soiltype.1x1.grb=gd
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > Warning! : Can't figure out time unit of 4
> > #       cpt file created by: grd2cpt -Cwysiwyg -E9 global_soiltype.1x1.grb=gd
> > #COLOR_MODEL = RGB
> > #
> > 0       64      0       192     1.125   64      0       192
> > 1.125   0       128     255     2.25    0       128     255
> > 2.25    64      224     255     3.375   64      224     255
> > 3.375   64      255     192     4.5     64      255     192
> > 4.5     192     255     64      5.625   192     255     64
> > 5.625   255     224     64      6.75    255     224     64
> > 6.75    255     32      64      7.875   255     32      64
> > 7.875   255     160     255     9       255     160     255
> > B       0       0       0
> > F       255     255     255
> > N       128     128     128
> >
> > Looks good, lets make an image.
> >
> > grdimage global_soiltype.1x1.grb=gd -JM15c -R-179/180/-89/89 -Cglobal_vegfrac.0.144.decpercent.cpt -V > global_soiltype.1x1.ps
> > grdimage: Allocates memory and read data file
> > GMT_grd_is_global: yes, longitude cells span exactly 360
> > gdalread: Unsupported data type
> > grdimage: project grid files
> > grdimage: Given domain implies x_inc = 0.0164499
> > grdimage: Grid projection from size 360x180 to 360x180
> > GMT_grd_is_global: yes, longitude cells span exactly 360
> > GMT_boundcond_param_prep determined edgeinfo: gn = 0, gs = 1, nxp = 360, nyp = 0
> > grdimage: Evaluate pixel colors
> > grdimage: Creating PostScript image 24-bit [colorimage]
> > pslib: Colormap of 14 colors created
> > pslib: Image depth reduced to 4 bits
> > pslib: LZW compressed 32400 to 369 bytes
> >
> > The trouble is that the postscript image is empty.  I have confirmed
> > with other programs that the file does contain the expected data.
> >
> > I am using GMT-4.5.1 with GDAL-1.6.2 on a 32-bit Linux box.  GDAL was
> > configured with internal support for GRIB format.  The table of
> > supported format located at http://www.gdal.org/formats_list.html
> > indicates it supports both GRIB1 and GRIB2.
> >
> > Any ideas?
> >
>
> To unsubscribe, send the message "signoff gmt-help" to listserv@...

--
Dan Kokron
Global Modeling and Assimilation Office
NASA Goddard Space Flight Center
Greenbelt, MD 20771
Daniel.S.Kokron@...
Phone: [phone]
Fax:   [phone]

To unsubscribe, send the message "signoff gmt-help" to listserv@...