[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Possible grib-8.0.jar bug

This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.

  • Subject: Re: Possible grib-8.0.jar bug
  • Date: Wed, 18 May 2011 22:55:48 -0400
Hi John

I don't know who originally created these two MFA or Meteo France grib tabs, 
but I do believe that there relatively new
and they appeared within the grib jar sometime after TDS 3.6.x.

Fabrice VOITUS and I have appreciated that someone created them because 
previously we were working a kludge type work
around by using an existing ECMWF grib tab table.

Please don't remove the meteofrance tabs, I believe that removal of the period 
after the Accum. snow variable within
both grib tab resource files will be sufficient. I've never seen use of a 
period on any of the other abbreviated
variable names within any of the other existing grib tab tables.

I've tested this within an internal NCDC development TDS server with good 
results. After deploying the enhanced
grib-8.0.jar and clearing the cache both OPeNDAP, NetcdfSubset and all of other 
access methods rendered the Accum_snow
variable properly.

Tomorrow I plan on deploying this change to our externally facing production 
TDS server which can be reviewed from the
link within my original email below.

Maybe after Robb returns from vacation he can take a quick look at the Meteo 
France grib tab tables and confirm whether
my solution is appropriate. I suspect that he may know about the origins and 
the author of these two new grib tab tables.

Thanks

-Bill



On 5/18/2011 5:46 PM, John Caron wrote:
> Hi Bill:
> 
> We are looking at a solution for this. Do you happen to know where 
> meteofrance_1.tab and meteofrance_2.tab files came
> from? (Robb is on vacation for two weeks.)
> The quick fix is to remove these, but we are also looking at a fix in 
> opendap, since this problem shows up elsewhere.
> 
> John
> 
> On 5/16/2011 11:35 AM, Bill Hankins wrote:
>> Hi John
>>
>> Regarding a possible issue with the new meteofrance_1.tab and 
>> meteofrance_2.tab grib tab resource files within the
>> grib-8.0.jar.
>>
>> Line item 65, i.e.
>>
>> 65:WEASD:Accum. snow [kg/m^2]
>>
>> I believe that the period after Accum is causing problems with OPeNDAP and 
>> NetcdfSubset access methods.
>>
>> While reading a grib type MFA SURFA dataset with OPeNDAP the Accum._Snow 
>> variable attributes were showing up within
>> global attributes.
>>
>> In NetcdfSubset the Accum._Snow variable won't render without an 403 error 
>> stating that Accum._Snow isn't a netcdf
>> variable or sometimes with just a content encoding error.
>>
>> e.g.
>>
>> http://www.ncdc.noaa.gov/thredds/catalog/surfa/mfa/y2008/m01/catalog.html?dataset=surfa/mfa/y2008/m01/MFA_2008011412_18.gz
>>
>>
>> Here's the wgrib dump of what should be the Accum_snow variable within the 
>> MFA SURFA dataset.
>>
>> 12:34290960:d=08010112:WEASD:kpds5=65:kpds6=1:kpds7=0:TR=0:P1=12:P2=0:TimeU=1:sfc:12hr
>>  fcst:NAve=0
>>
>> After removing the period from the meteofrance grib tab tables both errors 
>> were resolved while testing on our
>> development server.
>>
>> Please advise.
>>
>> -Bill
>>
>>
> 

-- 
Bill Hankins
Software Developer, STG, Inc., Government Contractor
Remote Sensing and Applications Division
National Climatic Data Center - NESDIS - NOAA
151 Patton Ave. Rm 445
Asheville, NC 28801-5001
address@hidden
Work-828-271-4910
Cell-828-458-2481