[gradsusr] Cannot Contour Grid - all undefined values error

William Rasch - NOAA Federal william.rasch at noaa.gov
Tue Jun 6 15:15:03 EDT 2017


Jennifer and Wesley,

Can't thank you guys enough for your help.  This is really great!!  But,
I'm still having issues, however, I am very excited the Wesley was able to
get
something working so I have gone from thinking I'll never get this done to
some hope now!!

Wesley,  I'm really sorry, but I need even more specifics of how you did
it.  Here is what I tried, please tell me where I have gone wrong...

1.)  I download ALL the pctile file from a certain date, in this case,
ftp://ftp.cdc.noaa.gov/Public/gbates/reforecast2/efi_grib2/pctile.2017060600.000_to_024hrs.grib2
all the way through the 156_180hrs files, so a total of 14 files (so not
the efi* files, only the .pctile files).

2.)  I then cat all of these into a "master" file (this technique works for
everything else I have done in the past).
cat * > efi.master

3.)  Run g2ctl on that master file, but I used the -b command this time?
Did I misunderstand that?
g2ctl -b efi.master > efi.master.ctl

4.)  Then run gribmap with the -b command?
g2ctl -b efi.master.ctl  (I have used -i in the past).

5.)  When I do the fifth step, I can a command line entry request of:
 Enter Name of Data Descriptor file:   What do I enter there?  For this
time I entered efi.master.ctl.

With those steps I still get the undefined error.

Can't thank you guys enough, really hope we can figure this out!!!!

Bill


On Tue, Jun 6, 2017 at 10:16 AM, Wesley Ebisuzaki - NOAA Federal <
wesley.ebisuzaki at noaa.gov> wrote:

> Bill,
>
> I had no problems plotting the efi* and pctile* files.  This is not to brag
> to to say that with a little more understanding, you can start making
> plots!
>
> g2ctl is a simple program.  When you make a ctl file from a single grib
> file,
> it looks for the records with the earliest, second earliest and last time
> stamp.  For example,
>
> earliest time stamp:   00Z jan 1 2017
> next time stamp:        06Z jan 1 2017
> last time stamp:         18Z jan 2 2017
>
> From these three time stamps, g2ctl knows
>
>    1) data starts at 00Z jan 1 2017
>    2) dt=6 hours
>    3) number of time steps = 8
>        -> grads ctl file:  tdef 8 linear 00Z01jan2017 6hr
>
> If you have data at 07Z jan 1, 2017, gribmap will ignore it because it is
> expecting data
> every 6 hours.
>
> g2ctl needs to make a list of variables for the control file.  To save
> time, g2ctl
> makes the variable list from the first and second time steps.  If the 3rd
> time step
> and a new variable, g2ctl will not find it.  Since you are  combining efi*
> and pctile*
> files, this may have happened.
>
> The GrADS data model only has one time dimension.  For example,
>
>    default mode:    analysis at 00Z Jan 3 2017  has the same time as
>                              24 hour forecast from 00Z Jan 2, 2017.
>
> To get around this problem, you have to be careful that g2ctl and gribmap
> using the same time mode.  For your files,
>
> default:  (no option) use verification time (end of time average)
> -0:          use start of forecast (initial time)
> -b:          use start of time average
>
> For my test, I used the -b mode.
>
> When you make your files, you should either make
>
>     cat pctile.2017060200 <(201)%20706-0200>.* >pctile.2017060200
> <(201)%20706-0200>.grib2
> or
>     cat pctile.*.000_to_024hrs.grib2 >pctile.000_to_024hrs.grib2
>
> Combining pctile and efi files is asking for trouble.
>
> Wesley Ebisuzaki
>
>
> On Tue, Jun 6, 2017 at 11:50 AM, Jennifer M Adams <jadams21 at gmu.edu>
> wrote:
>
>> There is no need to apologize for being a novice. You may find this doc
>> page on how GrADS deals with GRIB data useful:
>> http://cola.gmu.edu/grads/gadoc/grib.html
>> Please see more comments inline.
>>
>>
>> > On Jun 6, 2017, at 11:34 AM, William Rasch - NOAA Federal <
>> william.rasch at noaa.gov> wrote:
>> > I grabbed one of the files, only one, then cat'ed that to the efi.master
>>
>> I would haved run g2ctl on just the one file, before cat’ing it to
>> efi.master.
>>
>> >
>> > Ran:  g2ctl  efi.master.ctl
>> >
>> > Then, did a gribmap -v efi.master.ctl
>> >
>> > It then prompted me the "descriptor file", and I entered:
>> efi.master.ctl
>> >
>> > The feedback from that was this:
>>
>> In general, it would be better to send the actual text instead of an
>> image of it.
>>
>> >
>> > <image.png>
>> It doesn’t look you got a MATCH for any of the records in your file.
>> What is in your descriptor file?
>>
>> —Jennifer
>>
>> >
>> > I then tried to display tmp2m and got the same error.
>> >
>> > Thanks for any additional help and again, sorry for my incompetence
>> with this.
>> >
>> > Bill
>> >
>> > On Mon, Jun 5, 2017 at 6:20 PM, Jennifer M Adams <jadams21 at gmu.edu>
>> wrote:
>> > Hi, Bill -
>> > I find that the g2ctl output often needs a few tweaks to get evertyhing
>> right and some variables need a fourth grib code in the units field in
>> order to match properly. You also appear to be concatenating files with
>> different time stamps and this may be confusing the issue. Try setting up
>> your descriptor with only one file, and run gribmap in verbose mode (with
>> the -v) and look at what is matching/not matching. If records don’t match
>> what’s in your descriptor, gribmap will show you what the entry should
>> contain. Once you are matching all the records in one file, then add the
>> templating for multiple time steps. If it’s still not working please send
>> your descriptor and the gribmap verbose output.
>> > —Jennifer
>> >
>> >
>> >
>> > On Jun 5, 2017, at 11:33 AM, William Rasch - NOAA Federal <
>> william.rasch at noaa.gov> wrote:
>> >
>> > > Hi,
>> > >
>> > > I have what would seem to be a very simple data-set and "variable"
>> I'm trying to plot wiht GrADS.
>> > >
>> > > The grib files are located here...
>> > >
>> > > ftp://ftp.cdc.noaa.gov/Public/gbates/reforecast2/efi_grib2/p
>> ctile.2017060500.000_to_024hrs.grib2
>> > >
>> > > I am simply downloading several of these files, then do this:
>> > >
>> > > cat * > efi.master
>> > > g2ctl efi.master > efi.master.ctl
>> > > gribmap -i efi.master.ctl
>> > >
>> > > I'm simply trying to plot either:
>> > >
>> > > temp2m
>> > > wind10m
>> > > apcpsfc
>> > >
>> > > When I try to display the data, I get the error:
>> > >
>> > > Cannot Contour Grid - all undefined values
>> > >
>> > > Thanks for any help....
>> > >
>> > > Bill
>> > > _______________________________________________
>> > > gradsusr mailing list
>> > > gradsusr at gradsusr.org
>> > > http://gradsusr.org/mailman/listinfo/gradsusr
>> >
>> > --
>> > Jennifer Miletta Adams
>> > Center for Ocean-Land-Atmosphere Studies (COLA)
>> > George Mason University
>> >
>> >
>> >
>> >
>> > _______________________________________________
>> > gradsusr mailing list
>> > gradsusr at gradsusr.org
>> > http://gradsusr.org/mailman/listinfo/gradsusr
>> >
>> >
>> >
>> > --
>> >
>> > William Rasch
>> > Science and Operations Officer
>> > Weather Forecast Office
>> > Sacramento, CA
>> > _______________________________________________
>> > gradsusr mailing list
>> > gradsusr at gradsusr.org
>> > http://gradsusr.org/mailman/listinfo/gradsusr
>>
>> --
>> Jennifer Miletta Adams
>> Center for Ocean-Land-Atmosphere Studies (COLA)
>> George Mason University
>>
>>
>>
>>
>> _______________________________________________
>> gradsusr mailing list
>> gradsusr at gradsusr.org
>> http://gradsusr.org/mailman/listinfo/gradsusr
>>
>
>
> _______________________________________________
> gradsusr mailing list
> gradsusr at gradsusr.org
> http://gradsusr.org/mailman/listinfo/gradsusr
>
>


-- 
<https://twitter.com/NWSSacramento>
<https://www.facebook.com/NWS.Sacramento>
<http://www.youtube.com/nwssacramento>
William Rasch
Science and Operations Officer
Weather Forecast Office
Sacramento, CA
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://gradsusr.org/pipermail/gradsusr/attachments/20170606/42833c3f/attachment-0001.html 


More information about the gradsusr mailing list