[gradsusr] Global Control File Issue

[L.B.]

I found the issue. So I produced a sample output of that file using the
g2grb program to write out grib files from GrADS. From here, I use g2ctl to
create a control file. It turns out I needed one more grid point
longitudinally in my control file. I added that and it fixed the issue.

ydef 721 linear -90.00 0.25
xdef 1440 linear 0.00 0.25

should have been this:

ydef 721 linear -90.00 0.25
xdef 1441 linear 0.00 0.25

*-------*
William (L.B.) LaForce IV
Meteorologist


On Tue, Mar 12, 2019 at 10:38 AM L.B. <bcbass2989 at gmail.com> wrote:

> I've also attached a copy of the binary file and control file in case you
> would like to play around with it. The binary file contains just one
> timestep: 00Z01JAN
>
> *-------*
> William (L.B.) LaForce IV
> Meteorologist
>
>
> On Mon, Mar 11, 2019 at 8:18 PM L.B. <bcbass2989 at gmail.com> wrote:
>
>> I'm having an issue with a pesky control file that I created for a binary
>> file that I wrote out with GrADS using the fwrite function.
>>
>> I have a large dataset of multiple grib2 files containing 500mb
>> Geopotential by year. I used templating in the control file to run through
>> all the grib files. Here is the control file for that data:
>>
>> dset ^ERA5_500GEO_%y4.grb2
>> index ^ERA5_500GEO_Climo.grb2.idx
>> undef 9.999E+20
>> options template
>> title ERA5_500GEO_Climo
>> * produced by g2ctl v0.1.4
>> * griddef=1:0:(1440 x 721):grid_template=0:winds(N/S): lat-lon grid:(1440
>> x 721) units 1e-06 input WE:NS output WE:SN res 48 lat 90.000000 to
>> -90.000000 by 0.250000 lon 0.000000 to 359.750000 by 0.250000
>> #points=1038240:winds(N/S)
>>
>> dtype grib2
>> ydef 721 linear -90.00 0.25
>> xdef 1440 linear 0.00 0.25
>> tdef 58440 linear 00Z01JAN1979 6hr
>> zdef 1 linear 1 1
>> vars 1
>> GP500mb   0,100,50000   0,3,4 ** 500 mb Geopotential [m^2/s^2]
>> ENDVARS
>>
>> From here, I created a script to further reduce the size fo the
>> climatology file by writing out to a binary file the average 500mb
>> geopotential height every 6 hours for one year (data starts in 1979, but I
>> started the script in 1980 to include a leap year to get Feb 29th - hence
>> why I used 1461 for the first timestep).
>>
>> The script that outputs the 6hour average is here:
>>
>> 'reinit'
>> 'open ERA5_500GEO.ctl'
>> 'set gxout fwrite'
>> 'set fwrite 500HGT_Climo.dat'
>>
>> tt=1461
>>
>>  while (tt<=2924)
>>      'set t 'tt
>>      'q dims'
>>      rec=sublin(result,5)
>>      vdate=subwrd(rec,6)
>>      vday=substr(vdate,1,8)
>>      say vday
>>
>>      if (vday=00Z29FEB | vday=06Z29FEB | vday=12Z29FEB | vday=18Z29FEB)
>>
>>  'mean=(GP500mb(time='vday'1980)+GP500mb(time='vday'1984)+GP500mb(time='vday'1988)+GP500mb(time='vday'1992)+GP500mb(time='vday'1996)+GP500mb(time='vday'2000)+GP500mb(time='vday'2004)+GP500mb(time='vday'2008)+GP500mb(time='vday'2012)+GP500mb(time='vday'2016))/10'
>>
>>      else
>>
>>
>>  'mean=(GP500mb(time='vday'1979)+GP500mb(time='vday'1980)+GP500mb(time='vday'1981)+GP500mb(time='vday'1982)+GP500mb(time='vday'1983)+GP500mb(time='vday'1984)+GP500mb(time='vday'1985)+GP500mb(time='vday'1986)+GP500mb(time='vday'1987)+GP500mb(time='vday'1988)+GP500mb(time='vday'1989)+GP500mb(time='vday'1990)+GP500mb(time='vday'1991)+GP500mb(time='vday'1992)+GP500mb(time='vday'1993)+GP500mb(time='vday'1994)+GP500mb(time='vday'1995)+GP500mb(time='vday'1996)+GP500mb(time='vday'1997)+GP500mb(time='vday'1998)+GP500mb(time='vday'1999)+GP500mb(time='vday'2000)+GP500mb(time='vday'2001)+GP500mb(time='vday'2002)+GP500mb(time='vday'2003)+GP500mb(time='vday'2004)+GP500mb(time='vday'2005)+GP500mb(time='vday'2006)+GP500mb(time='vday'2007)+GP500mb(time='vday'2008)+GP500mb(time='vday'2009)+GP500mb(time='vday'2010)+GP500mb(time='vday'2011)+GP500mb(time='vday'2012)+GP500mb(time='vday'2013)+GP500mb(time='vday'2014)+GP500mb(time='vday'2015)+GP500mb(time='vday'2016)+GP500mb(time='vday'2017)+GP500mb(time='vday'2018))/40'
>>
>>      endif
>>
>>       'd mean/9.80665'
>>
>>   tt=tt+1
>>  endwhile
>>
>> 'disable fwrite'
>>
>> The control file for this binary file named: 500HGT_Climo.dat is where I
>> am having the issue. The control file is below:
>>
>> dset ^500HGT_Climo.dat
>>
>> title ERA5 500 hPa Height Climatology 1979 to 2018
>> undef 9.999E+20
>>
>> ydef 721 linear -90.00 0.25
>> xdef 1440 linear 0.00 0.25
>> tdef 1464 linear 00Z01JAN1980 6hr
>> zdef 1 linear 1 1
>> vars 2
>>  hgt500mean         1  99  500 hPa Mean Geopotential Height [gpm]
>> ENDVARS
>>
>> If I were to just plot the first timestep from this binary file, I would
>> get the first image attached named: 'fromdatfile.png'. This image is wrong,
>> however. The correct one (attached) is named: 'fromscript.png'. I can run
>> that same 6 hour average script but change it to output a contoured image,
>> and it is correct (the output is the 'fromscript.png' image). However, that
>> same exact code, when written to binary format and then plotted is wrong.
>> Somehow the control file is messing up, but I cannot figure out where.
>>
>> I have tested the 6hour average script using a smaller domain than global
>> (just the US) and adjusted the control file accordingly and it works fine,
>> but the full global domain doesn't want to plot correctly with that control
>> file. Any ideas on what might be going on? Thanks!
>> *-------*
>> William (L.B.) LaForce IV
>> Meteorologist
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://gradsusr.org/pipermail/gradsusr/attachments/20190313/ca74e908/attachment.html>