In the dataset, S1A_IW_GRDH_1SSV_20150511T012334_20150511T012359_005867_0078D9_A725.SAFE, if I look at the geoLocationGridPoint -> height values, the height values are not changing within a single row of GCPs. So in the array of 21x10 geolocation grid points, every row of 21 points contains the same height value for varying lat/lon. Why is that so?
Also, it seems like the height should range from 0 to 1200m for the area covered by this scene, but as per the geoLocationGrid, the height values range from 77 to 350m. So, not sure what are these height values representing?
These are the average terrain height values that were used to focus the data - average over the swath. You will find time-tagged entries of this in the terrainHeightList in the annotation. These height values do not point measurements like from a DEM.
Thanks @piyushrpt for the response. I really appreciate your help.
I checked the terrainHeightList values in the sample dataset that I mentioned in my original post, and the time stamps on the terrain values does not match the Z values. Also, there are only 5 terrain height values, as opposed to 210 GCPs.
So, not sure how these would correlate.
Thanks for the help so far.
What are you trying to achieve? Terrain correction will optionally include the DEM as a band so you will get height values at the resolution of the DEM.
I am trying to see if we can fit the GCPs with a RPC, as a fast way to georeference the data using the height values from the metadata file as specified in the geoLocationGrid. But, yes, using a DEM will solve the problem.
Using a DEM is the only way forward as the geoLocationGrid does not contain detailed heigh-information, as explained above.