NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.

To learn about what's going on, see About the Archive Site.

Re: [thredds] HDF5 Aggregation

  • Subject: Re: [thredds] HDF5 Aggregation
  • From: John Caron <caron@xxxxxxxxxxxxxxxx>
  • Date: Fri, 30 May 2008 08:18:35 -0600
thanks Rich, for answering that question.

BTW, im noticing that a number of email servers are refusing messages with zip 
attachments, since thats a possible security problem. Probably best to just 
include the URL when you want to provide a zip file. thanks.

Rich Signell wrote:
Dave,

I've been working on a similar problem:  using NcML to make the HDF
remote sensing imagery served by the NOAA Coastwatch East Coast
OpenDAP server CF-compliant.

I have been using aggregation type="union" to join the HDF data with a
netcdf file that contains the lon/lat values, since Coastwatch uses
the same non-uniform lon/lat grid (it's a uniform mercator grid).

The NcML looks like this:

<netcdf xmlns="http://www.unidata.ucar.edu/namespaces/netcdf/ncml-2.2";>
    <aggregation type="union">
        <netcdf
location="http://coast-enviro.er.usgs.gov/models/gom_interop/coastwatch_northeast_lonlat.nc"/>
        <netcdf
location="http://coastwatch.chesapeakebay.noaa.gov/dap/edac/avhrr/data/2008_135_0921_noaa15_ex1_sst.hdf"/>
     </aggregation>
</netcdf>

I can also aggregate the HDF5 images along a virtual time dimension
and then union with the netcdf file that has the lon/lat values, and
it works -- only remaining issue is that the scale_factor is lost for
some reason, which I'm looking into.

Try getting http://coast-enviro.er.usgs.gov/models/share/coastwatch.zip
and running "test_sst", then looking at the two NcML files "sst.ncml"

-Rich

On Fri, May 30, 2008 at 9:20 AM, David Robertson
<robertson@xxxxxxxxxxxxxxxxxx> wrote:
Hi,

If I want to serve aggregated HDF5 files, do I have to specify the
coordinate system (grid) with NcML, or is there a way to construct the
HDF5 file so THREDDS can get the coordSys directly from the HDF5 file? I
ask because if my only option is NcML I will ask my "data source" to
create NetCDF files instead since the coverage area is 1222 x 1183
points and they are not equal spaced.

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





------------------------------------------------------------------------

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


  • 2008 messages navigation, sorted by:
    1. Thread
    2. Subject
    3. Author
    4. Date
    5. ↑ Table Of Contents
  • Search the thredds archives: