Ensemble GRIBs decoding

[Davide Sacchetti]

if all the members have the same grib codes, as 7,100,500 ... no way.
7 -> grib code
100 -> the variable is referred to a pressure level
500 -> the level is 500 (hPA)
The only method I know is to split grib on the different members and
create different ctl. In other words: if grads cannot distinguish
between different members in a single file ... grads can distiunguish
different files ...
hoping it helps

bye bye
davide


On Sun, 2005-01-23 at 16:50 +0100, Mandelli Alessandro wrote:
> I would like to make use of the gribs found in
> ftp://ftpprd.ncep.noaa.gov/pub/data/nccf/com/mrf/prod/
>
> I downloaded enspost.t00z.z500 and the index file enspost.t00z.z500i
>
> Then I created the control file
> ga-> !perl grib2ctl.pl -verf enspost.t00z.z500 enspost.t00z.z500i
>  >enspost.ctl
>
> Result is
>
> dset ^enspost.t00z.z500
> index ^enspost.t00z.z500i
> undef 9.999E+20
> title enspost.t00z.z500
> *  produced by grib2ctl v0.9.12.5p33j beta11
> dtype grib 2
> options yrev
> ydef 73 linear -90.000000 2.5
> xdef 144 linear 0.000000 2.500000
> tdef 65 linear 00Z22jan2005 6hr
> zdef 1 linear 1 1
> vars 1
> HGTprs  0 7,100,500  ** Geopotential height [gpm]
> ENDVARS
>
> I just modified "zdef 1 linear 500 500" for consistency.
>
> Then I mapped the grib
>
> ga->!gribmap -i enspost.ctl
>
> So far so good. I'm proudly plotting all timesteps of the variable.
> Fact is that I'm pretty sure that in the grib all 12 members for the 65
> timsteps are included.
> In fact
> ga-> !wgrib -v enspost.t00z.z500
>
> gives fcst+0, -0, +1, -1, etc. for all the 12 members.
> Unfortunately I'm not able to extract the members from the grib.
> Anyone has a control file or a suggestion how to do it?
>
> Thanks
> Regards
--
Davide Sacchetti <davide.sacchetti at arpal.org>
ARPAL - CMIRL

______________________________________
For your security, this mail has been scanned and protected by Inflex