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

[THREDDS #VNI-792787]: WMS error



Hi Hualan,

Ah, I think we're getting somewhere now!  I'll be happy to add the 
customized grib tables to the TDS when I am confident that I'm not going 
to break something.  When I was looking at the gldas data files (CLM, 
Mosaic, Noah), I saw that all of them were using 
Center-Subcenter-Version 221-221-1.  However, I noticed a subtle 
difference on the parameter definition of at least one variable, and 
then there was the oddity of having the same parameter number map to 
multiple levels.  Strictly speaking, this sort of thing (deviating from 
one-to-one mapping) is not allowed, even if the "value" is the same at 
the different levels, and different datasets with different parameters 
should have individual grib tables for those datasets. Furthermore, 
before I hard code in a Version 1 table, I need to be sure that the 
table has never changed.  If some parameter was changed, or even added, 
the version number should increment.  That way, someone who created a 
dataset using the Version 1 table a couple of years ago won't experience 
a problem when  someone uses the "new" table to read their data.

All that said, I don't know much about Panoply, but the TDS does allow 
custom grib tables to be used at the local level:

http://www.unidata.ucar.edu/software/thredds/current/netcdf-java/formats/GribTables.html#GribTables

It's possible for someone running a TDS to include these tables locally 
for serving up data.  On the other hand, if I can get definitive, 
composite tables for the following datasets, I'll include them in our 
distribution:

7-4-130
7-12-130
7-138-130

173-4-130

221-221-1

Including them in our distribution would ultimately be the better 
option, as then anyone using our software would have the correct tables 
for these datasets.

Many Thanks,
   Lansing

On 1/10/2014 9:06 AM, Rui, Hualan (GSFC-610.2)[ADNET SYSTEMS INC] wrote:
> New Client Reply: WMS error
>
> Hi Lansing,
>
> I am Hualan Rui at GES DISC and I am responsible for NLDAS/GLDAS and other 
> hydrology data sets.
>
> Regarding to your inquiries about NLDAS/GLDAS GRIBTAB files, I would like to 
> explain that the parameters of NLDAS/GLDAS cannot be completely mapped to a 
> WMO standard GRIB table, that is why each NLDAS/GLDAS data set comes with its 
> own GRIBTAB file. That is also why the GRIBTAB files we use at DISC have head 
> line as "-1:-1:-1:-1".  Some other software, for example Panoply, allows to 
> input customized GRIBTAB file, but still requires to input information for 
> the center, subcenter and version of the GRIBTAB. Therefore, for the 
> convenience of Panoply users, we changed the headline to the center, 
> subcenter and version of the GRIBTAB. FYI, attached is the recipe, "How to 
> View GLDAS or NLDAS data in GRIB format with Panoply?"
>
> We get the information of center, subcenter and version of the GRIBTAB from 
> the output of wgrib (a software for reading GRIB files). For example, if I 
> run a command of "wgrib  -V GLDAS_CLM10SUBP_3H.A20131103.0000.001.grb", I can 
> find "center 221 subcenter 221 process 221 Table 1" from the output.
>
> Overall, yes, you are right, the NLDAS/GLDAS GRIBTAB files cannot be mapped 
> to any WMO standard GRIBTAB table exactly, therefore, customized GRIBTAB 
> files are needed. Since NLDAS/GLDAS data are in high spatial and temporal 
> resolution, have long temporal coverage, and contain many parameters,  the 
> subsetting service is heavily demanded. The current available subsetting 
> services are either a FORTAN program or GrADS Data Server (GDS),  but both of 
> them are suffering performance issues.  We have tested the subsetting via TDS 
> and found that TDS performed much better.  Therefore we are really hope that 
> the capacity for allowing customized GRIBTAB file can be added to the TDS, so 
> that the TDS will work with NLDAS/GLDAS data sets correctly (I mean correct 
> in terms of parameter names).
>
> If you need more information, please let us know.  Whenever you want to 
> discuss with the data providers, please let me know.
>
> Looking forward to hearing from you soon.
>
> Thanks!
>
> Hualan Rui
> GES DISC
>
>   
>
>> -----Original Message-----
>> From: Fang, Fan (GSFC-610.2)[ADNET SYSTEMS INC]
>> Sent: Friday, January 10, 2014 8:06 AM
>> To: address@hidden
>> Cc: Rui, Hualan (GSFC-610.2)[ADNET SYSTEMS INC]
>> Subject: RE: [THREDDS #VNI-792787]: WMS error
>>
>> Hi Lansing,
>>
>> You can write to Hualan (address@hidden) about your inquiries.
>>
>> Hualan, if needed, you may ask producers of these tables to
>> clarify/correct these issues.  Thanks.
>>
>> -Fan
>> ________________________________________
>> From: Unidata THREDDS Support address@hidden
>> Sent: Thursday, January 09, 2014 3:57 PM
>> To: Fang, Fan (GSFC-610.2)[ADNET SYSTEMS INC]
>> Cc: address@hidden; Fang, Fan (GSFC-610.2)[ADNET
>> SYSTEMS INC]
>> Subject: [THREDDS #VNI-792787]: WMS error
>>
>> Hi Fan,
>>
>> I'm going through the tables you provided, and I think it would be best
>> if you would put me in direct contact with the data owner you've
>> mentioned below.  The sample files that seem to correspond with the
>> tables look for 221-221-1, but the tables themselves are subtly
>> different, and have wildly different completeness.  As is, I cannot
>> include them.
>>
>> Thanks,
>>   Lansing
>>
>> Ticket Details
>> ===================
>> Ticket ID: VNI-792787
>> Department: Support THREDDS
>> Priority: Normal
>> Status: Open
>
>
>
> Ticket Details
> ===================
> Ticket ID: VNI-792787
> Department: Support THREDDS
> Priority: Normal
> Status: Open
> Link:  
> https://www.unidata.ucar.edu/esupport/staff/index.php?_m=tickets&_a=viewticket&ticketid=23032



Ticket Details
===================
Ticket ID: VNI-792787
Department: Support THREDDS
Priority: Normal
Status: Open


NOTE: All email exchanges with Unidata User Support are recorded in the Unidata inquiry tracking system and then made publicly available through the web. If you do not want to have your interactions made available in this way, you must let us know in each email you send to us.