C2 matrix generation from two Sentinel-1 SLC images acquired at different dates

Good evening,
I am trying to create a C2 matrix from Sentinel 1 SLC product. I am following the steps advised in this article (doi: 10.20944/preprints201911.0393.v1). The problem is that my area of interest is just between two different product that are not acquired in the same day.

S1A_IW_SLC__1SDV_20200103T170610_20200103T170637_030639_0382D5_E0E9.zip
S1A_IW_SLC__1SDV_20200110T165753_20200110T165820_030741_03864C_B943.zip

I would like to have one C2 matrix covering my entire area of interest. Is it correct even if they are from different dates and orbits? Whic processing steps should I follow?
For each product I tried:
TOPS split → orbit file → calibrate → back geocoding → TOPS deburst → merge
Than i tried to mosaic the two merged images (befor matrix calculation) but It gives me strange results.

Any help is really appreciated! Thank you

Can you please specify why you want to use both images?

According to these footprints, the entire land area is already covered by the image from 03.01.2020.

I expect these images to have larger differences in their incidence angle, so there could be always inconsistencies.

BackGeocoding is for stacking images for interferometry, so depending on your application we have to think about more suitable preprocessing.

Thank you for the reply. There is a small slice of land which is not coverd by the first image.
After the matrix generation I should do H-Alpha Dual Pol decomposition

Then you should take two products of the same day and the same track

You can simply apply this processing to each product
TOPS Split (select only one sub-swath and the corresponding bursts)

  • Apply Orbit
  • Calibrate (complex)
  • Polarimetric matrix generation
  • Deburst

Then you merge both with

  • S1 Slice Assembly
  • Terrain Correction

The processing described in the article you mention is unnecessarily complicated and is only required if you want to work with the data in PolSAR pro later.

1 Like