Kari Hoijarvi wrote:
I don't know the correct CRS for the output data, but as a programmer
I'd like suggest an approach from different direction:
Everybody should support at least EPSG:4326.
Because then you can overlay anything with anything else, with a simple
offset+factor calculation. This would solve our problems immediately.
Having additional projections is a nice plus, though.
I guess this would imply that everyone must support reprojection to a lat/lon grid. In
Unidata's case we've been trying to see whether WCS can be used for serving
"raw" data.