Is the SRTM 1Sec HGT used in the Terrain-Correction a DTM?


Is the SRTM 1Sec HGT a DTM?
The DTM (Digital Terrain Model) gives the elevation of the bare Earth, on the contrary of the DSM (Digital Surface Model).

I thought that the vegetation & the urban infrastructure wasn’t removed from the standard SRTM DEM, and then started wondering if that was the case, how come the SRTM DEM is used during the Terrain-Correction (TC) instead of processed DTM? Doesn’t this make the TC less accurate?


The SRTM elevation models are the height of the main reflection of C-band radar because it was done with C-band InSAR. It is not exactly a Digital Surface Model as the C-band radar penetrates some distance into forests. You are correct that there is no removal of vegetation and urban infrastructure in the SRTM DEM products.

If you want to do terrain correction of C-band InSAR data, then the SRTM elevation is likely the most appropriate to use as it is the same elevation that your interferogram is seeing. If you were to use a Digital Terrain Model that has the surface elevation, you would get residuals that are proportional to the height of the C-band phase centers.


@dmichelakis This is the closest answer i could find…
@EJFielding I would like to use a DSM created from LiDAR to improve upon the SRTM. Since an srtm created a dsm, we are supposing a dsm NOT dtm would be the most appropriate substitute?

The SRTM elevations will be somewhere between the ground elevation from a Digital Terrain Model (DTM) and the tops of the trees that a LiDAR survey would see. I would interpret the SRTM elevations as a type of Digital Surface Model (DSM), but the “surface” measured with different methods (LiDAR vs. C-band radar) will be different depending on the types and density of trees.

1 Like