Re: CEOS WGISS WTF-CEOP

Hello Bryan,

Good question.

The satellite data is coming from various space agencies (JAXA, NASA, ESA, E
umetsat - members of CEOS) in various formats (e.g. HDF-EOS, CEOS, etc.).  T
he data is being subsetted to 250 by 250 km squares centered over each of th
e 35 Reference Sites, its being regridded from satellite swath to a lat/lon 
grid, and its being output in a standard format which is BSQ, 2 bytes, binar
y.  So, we need to convert the data from BSQ (Band Sequential) to whatever. 
 We could convert it to Geotiff and put it on a WCS directly.  Our primary p
urpose in participating in GALEON is to learn the technology of a WCS client
 reading gridded data in NetCDF files.  In the CEOP project, in addition to 
(> 1,000,000 scenes of) gridded satellite data, we also have (about 30 to 50
 terabytes) of gridded model output data, so it will be useful to know how t
o view both the model output data and the satellite data in a WCS client.

Right now my job is to investigate the technology of and the effort required
 to convert a few scenes of gridded satellite data to NetCDF-CF format and g
enerate the necessary ancillary metadata in order to participate in GALEON. 
 I will also look at converting the same scenes to Geotiff format and view t
hem with the Space Time Toolkit.  After we have a clearer understanding of t
his we can decide how to go forward in our R&D efforts.  We're looking at ma
ny different methods and technologies for data integration in our system.  A
ll of these technologies will potentially be useful in our future systems an
d in GEOSS.

Please let us know if you have any questions or suggestions.

Thank you,

Ben
JAXA/RESTEC


At 14:50 2005/09/20 +0100, you wrote:
> 
> Hi Ben
> 
> I'm intrigued to know why you would want to put satellite data (which 
> presumably starts in HDF or something similar) in NetCDF for an 
> *interoperability* experiment based on OGC protocols? 
> 
> If the existing data format wont support an OGC server, then an approach b
ased 
> on converting to NetCDF and using Stefano's middleware and taking part in 
> Galeon makes perfect sense ...  However, if you have data which could be 
> served up directly by an alternative WCS implementation then I would have 
> thought you and the Galeon community would both benefit by showing how the
se 
> differing data host formats can be served up *together* using OGC protocols.
> 
> Of course it makes perfect sense to be involved using that data which star
ts 
> life in NetCDF as well ...
> 
> (Note the major assumption that someone has a good way of serving up the 
> satellite data via a WCS).
> 
> As a side issue, can anyone confirm that CF has a good way of encoding 
> satellite scenes?
> 
> Regards,
> Bryan
>