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

Re: [thredds] netCDF coordinate system (GeoGrid) criteria - requirements for THREDDS WMS



Stephane:

Your sever is not reachable by me, so i cant look at that dataset. You could try sending it to me, or putting it on an HTTP server.

If running a TDS, you can correct problems with NcML, see:

http://www.unidata.ucar.edu/projects/THREDDS/tech/tds4.2/tutorial/NcML.htm

Also, you should send questions to

address@hidden

since i may be on vacation.

On 6/7/2010 2:44 PM, Stephane Poirier wrote:
Hi John,
 
Thanks for your reply. I tried your recommandations and could not get a grid to display (selecting alsi row) with netcdf4.2tools (without further altering our netCDF file).
 
I understand we do not define the coordTransforms field. The parsing with CF-1.0 convention fails to find CoordTransBuilder name=polar-stereographic from Variable=float alsi(time=124, yc=174, xc=182); Info button details in http://www.oifii.org/do-org/dod/2010june01_netcdf42-tools_MRCC.4.2.0_amno_adj_alsi_ps_196012_r-nc.txt for opened netCDF file
 
My experimental installation of the THREDDS server is on http://132.206.251.239:8080/thredds (hopefully accessible from outside university's firewall, not sure), the file of interest is
http://132.206.251.239:8080/thredds/catalog/testAll/catalog.html?dataset=testDatasetScan/MRCC.4.2.0_amno_adj_alsi_ps_196012_r.nc (note: this is one of many automatically generated files we have, for which we seem to generate netCDF headers that do not comply to CF-1.0 convention, therefore, files for which we cannot easily use THREDDS software. I wish to find out how to make our data comply so we can reuse your great opensource/opendap software. All our data files are climate scenario time series periodically regenerated and currently always formatted like this one sample file).
 
Our projection is a polar stereographic projection. Our current climate scenarios generator uses the following parameters here below but it is not currently well defined in our netCDF output files.
 
PARAM for the Canadian Regional Climate Model (North American) grid:
"
     float yc(yc) ;
               yc:units = "m" ;
               yc:long_name = "y-coordinate of polar-stereographic projection" ;
               yc:standard_name = "projection_y_coordinate" ;
               yc:axis = "y" ;
               yc:coordinate_defines = "point" ;
               yc:actual_range = 22500.f, 7807500.f ;
       float xc(xc) ;
               xc:units = "m" ;
               xc:long_name = "x-coordinate of polar-stereographic projection" ;
               xc:standard_name = "projection_x_coordinate" ;
               xc:axis = "x" ;
               xc:coordinate_defines = "point" ;
               xc:actual_range = 22500.f, 8167500.f ;
           float lon(yc, xc) ;
               lon:units = "degrees_east" ;
               lon:long_name = "longitude" ;
               lon:standard_name = "longitude" ;
               lon:actual_range = 160.4761f, 332.2443f ;
       float lat(yc, xc) ;
               lat:units = "degrees_north" ;
               lat:long_name = "latitude" ;
               lat:standard_name = "latitude" ;
               lat:actual_range = 11.56135f, 87.47531f ;
       char polar_stereographic ;
               polar_stereographic:straight_vertical_longitude_from_pole = 245.f ;
               polar_stereographic:standard_parallel = 60.f ;
               polar_stereographic:false_easting = -2790000.f ;
               polar_stereographic:false_northing = -8113500.f ;
               polar_stereographic:hemisphere_of_standard_parallel = 1.f ;
               polar_stereographic:resolution_at_standard_parallel = 45000.f ;
"
 
In short, how can we properly define our X, Y and T coordinates? How can we properly define our projection parameters?
 
Regards,
 
 
Stephane
 
 
 
----- Original Message -----
From: John Caron
Sent: Friday, June 04, 2010 4:43 PM
Subject: Re: [thredds] netCDF coordinate system (GeoGrid) criteria - requirements for THREDDS WMS

Hi Stephane:

comments are embedded

On 6/4/2010 1:17 PM, Stephane Poirier wrote:
 
Hi All,
 
THREDDS WMS requires NetCDF files to identify the coordinates system used. Checking this (for one of my NetCDF file) by opening in the Grid Panel of the ToolsUI application, I get the following http://www.oifii.org/do-org/dod/2010june01_netcdf41-tools_MRCC.4.2.0_amno_adj_alsi_ps_196012_r-nc.jpg

This is the right screen to look at (although i'd recommend starting to use the 4.2 ToolsUI). The dataType column shows "grid" when its succcessfully recognized as a "GeoGrid", which in this case it is not.  Given that "coordTranforms" column in the middle table is blank, its likely not picking up the projection. Do you have the data variable pointing to it? The info button has some useful info. Also, try selecting "alsi" data variable in first table, then right click and do "try as Grid".

Send a dods URL if these pointers are insufficient.

 
http://www.unidata.ucar.edu/Projects/THREDDS/tech/tds4.0/reference/WMS.html page states, in order to comply to WMS requirements, a NetCDF files should have one or more variables shown as a GeoGrid. How can one tell a variable is shown as a GeoGrid?
 
Alternatively, when I look at one of UCAR's NetCDF file (that seems to comply with the GeoGrid requirement), I get http://www.oifii.org/do-org/dod/2010june01_netcdf41-tools_ucar-station-profiler-wind_example.jpg. How can one tell a variable is shown as a GeoGrid?

This is definitely not a Grid, its point obs data. To simplify, a grid will have lat(lat), lon(lon), but a point data has lat(pt), lon(pt).

 
Regards,
 
 
Stephane
 
 
--------------------------------------------------------------------
Stephane Poirier, M.Sc. Optical Physics
Remote Sensing Application Software Developer
Geography Department, Burnside Hall
McGill University, Montreal, QC, Canada
http://geog.mcgill.ca
address@hidden
Cell. (514) 994-3532
Fax: (514) 509-8833
--------------------------------------------------------------------
Stephane Poirier, M.Sc. Optical Physics
Remote Sensing Application Software Developer
3532 rue Sainte-Famille, Apt. 4
Montreal, QC, H2X 2L1, Canada
http://www.oifii.org
Tel.: (514) 509-8833
Fax: (514) 509-8833
--------------------------------------------------------------------
_______________________________________________ thredds mailing list address@hidden For list information or to unsubscribe, visit: http://www.unidata.ucar.edu/mailing_lists/


_______________________________________________
thredds mailing list
address@hidden
For list information or to unsubscribe,  visit: http://www.unidata.ucar.edu/mailing_lists/