NcML file generation

[marjy]

We have another few model runs that need to be uploaded to the COMT server soon, since CI wants to work with them so they can present them at the summer meeting in July.

We have a 20-30 year run of ChesROMS-1term, a new two year CBOFS-1term run and a five year run of what we call ChesROMS-ECB.

Rich - can you help us generate the NcML files for these runs? If so, that would be much appreciated! None of us really have experience with the ncml file generation.

The five year ECB run is already on the Rutgers opendap server (history files are at: http://tds.marine.rutgers.edu/thredds/roms/ecb/catalog.html?dataset=ches_fengjgr2015_history), and of course on our cluster here at VIMS. Can you grab the files off the Rutgers opendap, or is it best for me to sftp them to sura_ftp@comt.sura.org?

Thanks in advance for your help with this!

[rsignell-usgs]

@marjy , I still need to do the NcML, but I tried transferring the data by logging into the testbed server and using NCCOPY with the Rutgers OPeNDAP URL as input, which wrote a local netcdf file on the testbed server as output. It seems to have worked!

Here's what I typed:

cd /data/comt_2/cb_hypoxia/VIMS_ROMS/5year/Output
nccopy -k 4 -d 6 -m 50000000 "http://tds.marine.rutgers.edu/thredds/dodsC/roms/ecb/ches/fengjgr2015/his" fengjgr2015.nc

which shows up on the THREDDS COMT2 "full" catalog at http://comt.sura.org/thredds/catalog/comt_2_full/cb_hypoxia/VIMS_ROMS/5year/Output/catalog.html?dataset=comt_2_full/cb_hypoxia/VIMS_ROMS/5year/Output/fengjgr2015.nc

I took a look with the Godiva2 viewer and it looks okay, right?

[marjy]

Fantastic! Wow, that was fast! It generally looks about like what I would expect. One issue is that the date says Jan. 1, 2006, but I thought the run went from Jan. 1 2001 to Dec 31 2005.

Also, on here I see two names for the model - the THREDDS path says “VIMS_ROMS” and the heading of the figure reads “ChesNENA” which must have been inside the .nc files.

Can both of these be changed to the correct name “ChesROMS-ECB”? ChesNENA is simply wrong (sorry it is in the meta data that way - I hadn’t realized that), and VIMS_ROMS doesn’t make sense because there are lots of ROMS models being run at VIMS.

I would rather not have any of our COMT model runs linked to a specific institution, since people change institutions, and of course the code is open source and can be run at any institution. Would this be possible? If they have to be categorized some way, I would rather they be listed by the last name of the person who contributed the run.

Thanks again for all your help with this! Marjy

On Jun 20, 2015, at 8:24 AM, Rich Signell notifications@github.com<mailto:notifications@github.com> wrote:

@marjyhttps://github.com/marjy , I still need to do the NcML, but I tried transferring the data by logging into the testbed server and using NCCOPY with the Rutgers OPeNDAP URL as input, which wrote a local netcdf file on the testbed server as output. It seems to have worked!

Here's what I typed:

cd /data/comt_2/cb_hypoxia/VIMS_ROMS/5year/Output nccopy -k 4 -d 6 -m 50000000 "http://tds.marine.rutgers.edu/thredds/dodsC/roms/ecb/ches/fengjgr2015/his" fengjgr2015.nc

which shows up on the THREDDS COMT2 "full" cataloghttp://comt.sura.org/thredds/catalog/comt_2_full/catalog.html at http://comt.sura.org/thredds/catalog/comt_2_full/cb_hypoxia/VIMS_ROMS/5year/Output/catalog.html?dataset=comt_2_full/cb_hypoxia/VIMS_ROMS/5year/Output/fengjgr2015.nc

I took a look with the Godiva2 viewer and it looks okay, right? [6-20-2015 8-19-27 am]https://cloud.githubusercontent.com/assets/1872600/8267235/a426ffc6-1725-11e5-9650-5349234f641b.png

— Reply to this email directly or view it on GitHubhttps://github.com/ioos/comt/issues/36#issuecomment-113757027.