Terrain-Correction
false
false
Latest Auxiliary File
Copernicus 30m Global DEM
BILINEAR_INTERPOLATION
true
0
BILINEAR_INTERPOLATION
Use projected local incidence angle from DEM
Use projected local incidence angle from DEM
AUTO:42001
true
false
0
20.0
false
false
false
false
false
false
false
true
false
0
0
I have tried with both Copernicus 30m Global DEM and Copernicus 90m Global DEM.
Hi,
It seems that there is another issue with SNAP for using Copernicus DEMs in TC step. I can read them in SNAP, but when I get output DEM I see it doesn’t convert the heights to the Ellipsoidal heights (h). Based on Help (which probably needs an update), even Aster is not converted. I haven’t tried all Global models, but using external DEM was OK and heights were converted using EGM96 geoid model. Hope this issue is resolved in new Software update. Thanks.
Sorry, this is not clear to me: You used Copernicus DEM and selected the option “Apply Earth Gravitational Model”, and then the heights were not correct?
Sorry, if it was not clear. That option (Apply Earth gravitational Mode) is only for external DEM, and not for Copernicus DEMs. Please see the attached screenshots from SANP. IT doesn’t apply Geoid model automatically for Copernicus DEMS. I think that option should be visible and applicable for all DEMs. Hope you can fix it please. Thanks.
thank you for clarification. Actually, I’m not quite sure about this. Maybe @jun_lu or @lveci can clarify if the EGM option only applies for external DEMs.
Can you please share a link where it is stated that the Copernicus DEMs require the application of EGM96?
Thanks. I don’t have any link which says which elevations these models present, but I know elevations in such Global models are relative to the Geoid. For calculations in SNAP (TC for example), they should be converted to the ellipsoidal heights (using Geoid model). I noticed, in my study area, the numbers at some sample points with known heights (both H and h), the Pixel info shows the H value (relative to the geoid, not h (relative to ref. ellipsoid). This means to me they need to be converted. Hope that helps.
According to Copernicus DEM Product Handbook, the vertical reference datum is the Earth Gravitational Model (EGM2008). Therefore, EGM96 should be applied to the DEM in Terrain Correction. A ticket has been created to track the problem (https://senbox.atlassian.net/browse/SITBX-849)
Ok, so it is a known problem. Please just let us know to which version of the software we would need to update SNAP to test the new release of TC with Copernicus DEM.
Thanks for your help,
Mauro
When will the fix to this problem be released? Note that this problem exists for terrain flattening as well as terrain correction as noted above: in both case, option to remove EGM only exists for External DEM.
Still wondering when there will be a fix for this issue. Using gpt, “Copernicus 30 m DEM” works correctly for “terrain correction” but does not work for “terrain flattening”.
Here are some of my observations about the Range Doppler-based Terrain Correction in the latest version of SNAP 9.0.0 using one Sentinel-1A SAR image and one ICEYE Spotlight SAR image:
If I go with the option of “Copernicus 30m Global DEM (Auto Download)”, it works fine. The tool will download a copy of the original Copernicus 30m (in tiles) to a folder like this “auxdata\dem\Copernicus 30m Global DEM”. NOTE these DEMs are still with original vertical datum (i.e., with EGM2008).
If I go with the option of external DEM and select a local Copernicus 30m DEM file already with WGS84 ellipsoidal height, without checking “Apply Earth Gravitational Model”, the outputs will have some geolocational offsets despite that the online help says the external DEM “must be with geographic coordinates (Plat, Plon, Ph) referred to global geodetic ellipsoid reference WGS84 (and height in meters).”. Also tried to checking “Apply Earth Gravitational Model”, which did not make any differences.
If I go with the option of external DEM and select a local Copernicus 30m DEM file that is still with original vertical datum (i.e., EGM2008), without checking “Apply Earth Gravitational Model”, the outputs are the same as the ones from “Auto Download”.