Roger is correct. GE will just drape your raster data over it's terrain mesh regardless of what vertical parameters your data has. It is essentially 2D with a z value and is interpreted as such. That is why EPSG:4326 works, because GE doesn't care about any other information. No conversions/shifts occur.
I export rasters from GRASS, generate hillshades, and overlay in GE using gdal2tiles daily. I seldom see any alignment issues, and when I do it is usually my data. Of course there are alignment issues in GE, but they will vary place to place regardless of topography. - Jamie On Fri, Oct 28, 2011 at 2:06 PM, Roger André <[email protected]> wrote: > I find this thread interesting. In my experience overlaying both raster > (orthophoto) and vector data in Google Earth with KML, EPSG:4326 has worked > with no problems whatsoever. So far as I know, GE does not apply a vertical > datum, it simply drapes the 2D data over its internal terrain model. > > Would it be possible for you to share some data where you have seen that > this is a problem? > > Thanks, > > Roger > > > On Fri, Oct 28, 2011 at 1:05 PM, Hamish <[email protected]> wrote: > >> Helmut wrote: >> > ... >> > > in my experience overlaying kml (vector and raster) >> > > with EPSG:4326 in google >> > > earth works relatively well in flat areas, in mountain >> > > areas there are >> > > sometime distortions therefore. >> Markus: >> > I can confirm this from my (limited) experience to export >> > data to KML. To correct for the geoid heights, you could use >> > the previously indicated map and calculate the local >> > ellipsoid-geoid difference and vertically shift your vector >> > data with v.transform. >> > Then export to KML. For raster, use r.mapcalc to shift. >> >> n.b., IIUC raster KML support in google earth is essentially a >> hack building on/abusing their raster-icon support. >> they may have improved things, but this is what I was lead to >> believe at the time of writing r.out.kml. >> >> probably it is useful to consider if the ellipsoid-geoid >> difference is important vs. the overall vertical differences >> in the data, if the result is just for viewing maybe it is not >> worth the trouble. >> >> >> Hamish >> _______________________________________________ >> grass-user mailing list >> [email protected] >> http://lists.osgeo.org/mailman/listinfo/grass-user >> > > > _______________________________________________ > grass-user mailing list > [email protected] > http://lists.osgeo.org/mailman/listinfo/grass-user > >
_______________________________________________ grass-user mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/grass-user
