<html>
<head>
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 10pt;
font-family:Tahoma
}
--></style></head>
<body class='hmmessage'><div dir='ltr'>
Hi Diane,<br><br>First, to reply to my own previous mailing - My Bad! When I read the instructions properly, I found that I should be setting all those initial YYMMDDHH's with an "i" prefix. <br><br>Then gribmap worked fine.<br><br>Plus - it was obviously some problem with the 2.2 Gig concatenated big file - because now, with the Options Template set I am able to plot right through the full 120 hours!<br><br>Thank you kindly for the help and suggestion. That was brilliant,<br><br>Thanks again,<br>Theo<br><br><br><div><div id="SkyDrivePlaceholder"></div><hr id="stopSpelling">From: thibidottwo@hotmail.com<br>To: gradsusr@gradsusr.org<br>Date: Sun, 27 May 2012 04:33:48 +0000<br>Subject: Re: [gradsusr] Error plotting full 120 hour GFS forecast, stops working after t30<br><br>
<style><!--
.ExternalClass .ecxhmmessage P
{padding:0px;}
.ExternalClass body.ecxhmmessage
{font-size:10pt;font-family:Tahoma;}
--></style>
<div dir="ltr">
<br>Hi Diane - thanks for the suggestion - perhaps it is a big file problem.<br><br>Trying dset ^%y2%m2%d2%h2.gfs.t%h2z.pgrb2f%f2 and then running grib2map for my gfs files which look like this:<br>12052418.gfs.t18z.pgrb2f00<br>12052418.gfs.t18z.pgrb2f03<br>etc..<br><br>only the first hour gets gribmapped - on the second file it tries to scan <br>grib2map: scanning GRIB2 file: 12052421.gfs.t21z.pgrb2f03 <br><br>Where it is incrementing the 18 hour parts to 21 - but that should stay the same as with year, month and day?<br><br>I have never tried the templates option, is that %h2 the correct replacement to use<br><br>Regards,<br>Theo<br><br><div><div id="ecxSkyDrivePlaceholder"></div><hr id="ecxstopSpelling">Date: Sat, 26 May 2012 23:36:03 -0400<br>From: diane.stokes@noaa.gov<br>To: gradsusr@gradsusr.org<br>Subject: Re: [gradsusr] Error plotting full 120 hour GFS forecast, stops working after t30<br><br>
Theo,<br>
<br>
Have you tried using the template option in your descriptor file to
process the individual files rather than concatenating them into one
large file? See:<br>
<a class="ecxmoz-txt-link-freetext" href="http://grads.iges.org/grads/gadoc/templates.html" target="_blank">http://grads.iges.org/grads/gadoc/templates.html</a><br>
for more info.<br>
<br>
Diane<br>
<br>
On 5/26/2012 10:14 PM, Theo Carter wrote:
<blockquote cite="mid:BAY156-W638BE0563E5817E8705015A2070@phx.gbl">
<style><!--
.ExternalClass .ecxhmmessage P
{padding:0px;}
.ExternalClass body.ecxhmmessage
{font-size:10pt;font-family:Tahoma;}
--></style>
<div dir="ltr">
Hi Diane,<br>
<br>
This is a new attempt at downloading gfs and creating text file
output - so I cannot say it worked for me before: however, I
have used a similar way (cat the files, create .ctl and idx
files) to create graphical (not text) output using GrADS from
wrf modeling. The one difference is that the wrf I use only
forecasts out to 48 hours, not 120.<br>
<br>
I use gribmaster to download full gfs004grb2 3-hourly 0.5 deg
files from f0 to f120 from the NCEP server.<br>
<br>
I then concatenate the files together using "cat" into one large
file of 2.2 Gig:<br>
cat *pgrb2*>>gfsbig<br>
<br>
Then create .ctl and .idx using:<br>
./g2ctl -verf gfsbig >gfsgrads.ctl<br>
gribmap -v -i gfsgrads.ctl<br>
<br>
(Running gribmap manually shows a "Match" in every instance)<br>
<br>
My created .ctl file shows:<br>
dtype grib2<br>
ydef 361 linear -90.000000 0.5<br>
xdef 720 linear 0.000000 0.500000<br>
tdef 41 linear 18Z24may2012 3hr<br>
<br>
And from within grads:<br>
"q file 1" gives: ! Xsize = 720 Ysize = 361 Zsize = 26
Tsize = 41 Esize = 1<br>
<br>
So all the times appear to be available.<br>
<br>
My hardware environment is 5.8GiB Ram and 4 processors Xeon
<a class="ecxmoz-txt-link-abbreviated" href="mailto:E5520@2.27GHz">E5520@2.27GHz</a> allocated.<br>
My software environment is Scientific Linux 6.2_64 in a virtual
machine (Basically same as Enterprise Linux). <br>
Grads Version 2.0.a9 from the EPEL repository and wgrib2 from
the same. Built Fri Sep 3 17:16:06 UTC 2010 for
x86_64-redhat-linux-gnu/<br>
<br>
q config reveals:<br>
This version of GrADS has been configured with the following
options:<br>
o Built on a LITTLE ENDIAN machine<br>
o Athena Widget GUI DISABLED<br>
o Command line editing ENABLED <br>
o printim command for image output ENABLED <br>
o GRIB2 interface ENABLED <br>
o NetCDF interface ENABLED <br>
netcdf 4.1.1 of Dec 27 2010 21:12:42 $ <br>
o OPeNDAP gridded data interface ENABLED<br>
o OPeNDAP! station data interface DISABLED<br>
o HDF4 and HDF5 interfaces ENABLED <br>
HDF 4.2r5 <br>
HDF5 1.8.5 <br>
o GeoTIFF and KML/TIFF output ENABLED<br>
o KML contour output ENABLED<br>
o Shapefile interface ENABLED<br>
<br>
<br>
<div>
<hr id="ecxstopSpelling">Date: Fri, 25 May 2012 09:50:56 -0400<br>
From: <a class="ecxmoz-txt-link-abbreviated" href="mailto:diane.stokes@noaa.gov">diane.stokes@noaa.gov</a><br>
To: <a class="ecxmoz-txt-link-abbreviated" href="mailto:gradsusr@gradsusr.org">gradsusr@gradsusr.org</a><br>
Subject: Re: [gradsusr] Error plotting full 120 hour GFS
forecast, stops working after t30<br>
<br>
Theo,<br>
<br>
Answers to the following questions may help the forum point
you in the right direction.<br>
<br>
What is the source of your data?<br>
Are you using the template option to read a series of files,
or are you reading one large file?<br>
Did the same request work for you for GFS runs prior to
20120522?<br>
What version of GrADS are you running? (Please provide output
from 'q config')<br>
<br>
Regards,<br>
Diane Stokes<br>
<br>
<br>
On 5/25/2012 12:23 AM, Theo Carter wrote:
<blockquote cite="mid:BAY156-W1337ABA4DE7F4BE77AC2F5A2010@phx.gbl">
<style><!--
.ExternalClass .ecxhmmessage P
{padding:0px;}
.ExternalClass body.ecxhmmessage
{font-size:10pt;font-family:Tahoma;}
--></style>
<div dir="ltr"> Hi All,<br>
<br>
I am trying to plot a variable using 120 hours of a GFS
run, but for some reason I cannot plot past T30 (T41 being
120 hours). Everything plots fine up to that point, but
anything past T30 crashes out with the GRIB2 I/O error
specified below.<br>
<br>
Has anyone else come across this?<br>
Cheers,<br>
Theo<br>
<br>
<br>
<br>
ga-> set t 1 41<br>
Time values set: 2012:5:24:18 2012:5:29:18 <br>
ga-> d tmpsfc<br>
GRIB2 I/O error: fseeko failed <br>
Data Request Error: Error for variable 'tmpsfc'<br>
Error ocurred at column 1<br>
DISPLAY error: Invalid expression <br>
Expression = tmpsfc<br>
ga-> c<br>
ga-> set t 1 30<br>
Time values set: 2012:5:24:18 2012:5:28:9 <br>
ga-> d tmax2m<br>
ga-> d tmp2m<br>
ga-> <br>
<br>
<br>
<br>
<br>
</div>
</blockquote>
<br>
<br>
_______________________________________________
gradsusr mailing list
<a class="ecxmoz-txt-link-abbreviated" href="mailto:gradsusr@gradsusr.org">gradsusr@gradsusr.org</a>
<a class="ecxmoz-txt-link-freetext" href="http://gradsusr.org/mailman/listinfo/gradsusr" target="_blank">http://gradsusr.org/mailman/listinfo/gradsusr</a></div>
</div>
<br>
<fieldset class="ecxmimeAttachmentHeader"></fieldset>
<br>
<pre>_______________________________________________
gradsusr mailing list
<a class="ecxmoz-txt-link-abbreviated" href="mailto:gradsusr@gradsusr.org">gradsusr@gradsusr.org</a>
<a class="ecxmoz-txt-link-freetext" href="http://gradsusr.org/mailman/listinfo/gradsusr" target="_blank">http://gradsusr.org/mailman/listinfo/gradsusr</a>
</pre>
</blockquote>
<br>
<br>_______________________________________________
gradsusr mailing list
gradsusr@gradsusr.org
http://gradsusr.org/mailman/listinfo/gradsusr</div>                                            </div>
<br>_______________________________________________
gradsusr mailing list
gradsusr@gradsusr.org
http://gradsusr.org/mailman/listinfo/gradsusr</div>                                            </div></body>
</html>