Multiple Data Processing Back-Geocoding

Hello all,

In SNAP 5.0 it became possible to apply Back-Geocoding for multiple slaves. Hence, 1 master for multiple slaves. When this is done, multiple DerampDemodphases are produced. I use these phases to deramp the master bands after Back-Geocoding, using a Bandmaths operator. However, as these derampDemodPhases show slight differences I do not know which to use for deramping. I always thought these phases should not differ among themselves, when produced based on the same master band.

As an example, Back-Geocoding was performed on the following datasets:

S1A_IW2_SLC__1SDV_20161014T055805_blabla_VV_VH.dim (master)
S1A_IW2_SLC__1SDV_20161002T055805_blabla_VV_VH.dim (slave)
S1A_IW2_SLC__1SDV_20160920T055804_blabla_VV_VH.dim (slave)

When the “output derampdemodphase” function is set as true, two derampdemodphases are produced:


These two phases show a difference. As a consequence, a few questions arise:

  1. What causes the difference between these derampDemodPhases?
  2. Which to use for deramping my master band?
  3. Which derampDemodPhase is normally used by the BackGeocoding operator, when ran in the GUI?

Thank you in advance,


Hi Corné,

Why don’t you use the tool “Enhance Spectral diversity” instead of using a band math ? To coregister the S1 you can do orbital correction ==> Back geocoding => enhance spectral diversity"

The deramp phase are not the same because you have two images. If I’m not wrong, the deramp phase show the quadratic phase term introduced by this acquisition system (TOPS and so steering of the antenna) so the term is slightly different for each acquisition.

Then 2 deramp are produced to correct the azimuth error in your slave images you want to coregister with your master.

Hope it helped