Problems with new ECMWF monthly forecast files

[Christian Marquardt]

If I remember correctly, there used to be a -e option to gribmap which takes
care of some filler bits ECMWF tends to add to their grib records. It won't
work without that option.

BTW, you did use gribmap, did you?

Hope this helps,

  Christian.



On Wed, Mar 19, 2008 at 10:39 AM, Daniele Gandini <
Daniele.Gandini at arpa.piemonte.it> wrote:

>  Dear GrADS users,
>
> I run GrADS 1.8sl11 on Sun Solaris Blade-1000 5.9
>
> Last week there was a change in ECMWF monthly forecast files
> and now I get the message "Entire grid undefined" when I try
> to display a variable.
>
> Here is an example of wgrib output for an old file which successfully
> works:
>
> wgrib -V EMEMO1500_20080306_011_032.grb | more
> rec 1:0:date 2008030600 Z kpds5=129 kpds6=100 kpds7=500 levels=(1,244)
> grid=255
> 500 mb ECext=23 class=operations type=Ensemble means stream=1206
> Version=0001 *5-11d ave:
> *  Z=Geopotential [m**2 s**-2]
>   timerange 3 P1 5 P2 11 TimeU 2  nx 62 ny 28 GDS grid 0 num_in_ave 14
> missing 0
>   center 98 subcenter 0 process 130 Table 128 scan: WE:NS winds(N/S)
>   latlon: lat  70.500000 to 30.000000 by 1.500000  nxny 1736
>           long -51.000000 to 40.500000 by 1.500000, (62 x 28) scan 0 mode
> 128 bdsgrid 1
>   min/max data 49641.9 56994.9  num bits 16  BDS_Ref 49641.9  DecScale 0
> BinScale -3
>
> And now wgrib output for new troublesome file:
>
> wgrib -V EMEMO1500_20080313_011_032.grb | more
> rec 1:0:date 2008031300 Z kpds5=129 kpds6=100 kpds7=500 levels=(1,244)
> grid=255
> 500 mb ECext=30 class=operations type=Ensemble means stream=ensemble
> forecasts Version=0001 *96-264hr ave:
> *  Z=Geopotential [m**2 s**-2]
>   timerange 3 P1 16 P2 44 TimeU 11  nx 62 ny 28 GDS grid 0 num_in_ave 14
> missing 0
>   center 98 subcenter 0 process 131 Table 128 scan: WE:NS winds(N/S)
>   latlon: lat  70.500000 to 30.000000 by 1.500000  nxny 1736
>           long -51.000000 to 40.500000 by 1.500000, (62 x 28) scan 0 mode
> 128 bdsgrid 1
>   min/max data 49871.8 56654.3  num bits 16  BDS_Ref 49871.8  DecScale 0
> BinScale -3
>
> As you can see, time definition is different, in hours instead of months.
>
> I create ctl file with grb2ctl.pl:
>
> perl $HOME/bin/grib2ctl.pl -verf EMEMO1500_20080313_011_032.grb
> EMEMO1500_20080313_011_032.gmp > EMEMO1500_20080313_011_032.ctl
>
> more EMEMO1500_20080313_011_032.ctl
>
> dset ^EMEMO1500_20080313_011_032.grb
> index ^EMEMO1500_20080313_011_032.gmp
> undef 9.999E+20
> title EMEMO1500_20080313_011_032.grb
> * analysis 2008031300 1stforecast 96-264 hr
> *  produced by grib2ctl v0.9.12.5p10
> dtype grib 255
> options yrev
> ydef 28 linear 30.000000 1.5
> xdef 62 linear -51.000000 1.500000
> tdef 4 linear 00Z24mar2008 7dy
> zdef 2 levels 850 500
> vars 6
> t2m  0 167,1,0  ** 2 metre temperature K
> mslmsl  0 151,1,0  ** Mean sea level pressure Pa
> probprec  0 228,1,0  ** Pot. evaporation [kg/m^2]
> probsnow  0 144,1,0  ** Volumetric soil moisture [fraction]
> tprs 2 130,100,0 ** Temperature K
> zprs 2 129,100,0 ** Geopotential (at the surface = orography) m**2 s**-2
> ENDVARS
>
> I run GrADS:
>
> grads -cl "open EMEMO1500_20080313_011_032.ctl"
>
> ga-> d zprs
> Cannot contour grid - all undefined values
>
> In attachment you have both files.
>
> Thank you in advance for every suggestions.
>
> Daniele Gandini
> ARPA Piemonte
> Servizio Meteorologico
> Turin - Italy
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://gradsusr.org/pipermail/gradsusr/attachments/20080319/db7ad8c6/attachment.html