Vertical Datum of a Sentinel-1 derived DEM

s1tbx Interferometry
#1

Hi, everyone. When a DEM is produced by the used of Sentinel-1 data, what its vertical reference system? I understand that to transform the relative height to absolute height, it is used a technique called DEM-Matching. If the auxiliary dem used is SRTM, the resulting vertical system of our Sentinel-1 dem is with respect to MSL.

Differences in SRTM data
#2

This depends on which method you select to convert the unwrapped phase to elevation. SNAP offers two:

  • Phase to height: No vertical reference, just relative heights (in meters).
  • Phase to elevation: You select a reference DEM and a coordinate reference system.
1 Like
#3

So, in Phase to elevation, the vertical datum will depend on the DEM used, or if the DEM had the geoid contribution in it. For example, if I remove the geoid contribution in the auxiliary dem the resulting Sentinel-1 dem will be with respect to the ellipsoid of reference?

#4

I would expect so, yes. But there can still be shifts introduced based on the unwrapping (e.g. falsely generated ramps).

#5

Yes, I would expect that too. I couldn’t found a reference that explicitly tells me what is the vertical reference system of the resulting DEM (Sentinel-1 in this case) using SNAP. “SNAP Help” make the reference to “A Beginner’s Guide to Interferometric SAR Concepts and Signal Processing” (Mark Richards, 2007) where it said that the resulted elevation is an ellipsoidal one. Clarify this is important at the time of making any analysis.