PolSARpro cannot recognize outputs from "Single Data Set" as inputs for "Multi Data Set"

Hi everyone,

I am testing Pol-TomoSAR in PolSARpro for Sentinel-1 and assume I need dual-pol complex vector (Sxx, Sxy) as the input.

These are the steps I performed to obtain (Sxx, Sxy):
(Please ignore the fact that I might not get good results as this is just a test :smiley:)

Processing in SNAP

  1. Radar>Sentinel-1 TOPS>S-1 TOPS Split
  2. Radar>Apply Orbit File
  3. Radar>Radiometric>Calibrate:
    Save as complex output.
  4. Radar>Coregistration>S1 TOPS Coregistration>S-1 Back Geocoding
  5. Radar>Coregistration>S1 TOPS Coregistration>S-1 Enhanced Spectral Diversity
  6. Radar>Sentinel-1 TOPS>S-1 TOPS Deburst
  7. Raster>Subset
  8. Manually edit the .dim file from 7:
    Remove colons and spaces in XML element name. Otherwise, Step 11 will throw an error.
  9. Raster>Bands extractor:
    Separate each acquisition in the stack product into individual products.
  10. Raster>Band Maths…:
    Calculate the kz parameter.
  11. File>Export>SAR Formats>PolSARpro
  12. Manually edit the config.txt file from 11:
    β€œdual” β†’ β€œpp2” for PolarType

Processing in PolSARpro

  1. Environment>Single Data Set (Pol-SAR)
  2. Import>Raw Binary Data:
    β€œMono Static” for Data Type
    β€œ(Sxx, Sxy)” for Data Format – (Sxx, Sxy)≑{s12, s22}={VH, VV}
    β€œ(S22, S12)” for Partial Polarimetry Data Format
    β€œReal / Imag” for Input Format
    Choose i_XXX.bin file for real and q_XXX.bin file for imaginary.
  3. Import>Extract PolSAR images:
    Select a different directory for outputs.
    Full Resolution
    β€œ(Sxx, Sxy)” for Sinclair Elements
  4. Repeat 13-15 for each acquisition.
  5. Environment>Multi Data Set (Time series / Pol-TomoSAR):
    Select every acquisition using β€œNew” button.
    PolSARpro throw a β€œNO DATA FILES FOUND…” warning after β€œSave & Exit”.

Basically, I followed these two posts

and this tutorial

to do the test.

Any ideas what could go wrong?

Any helps will be appreciated!
Thank you in advance~

2 Likes

Thank you for sharing your experience. I have currently prepared a manuscript to merge Sentinel-1 VV/VH with HH/HV (rare) acquisitions to synthesize a quad-pol product and process it in PolSAR pro, so I have also conducted most of the steps you report.

May I ask how you calculated the kz parameter?
As you saw in the other topic, I have also tested the TomoSAR approach (with ALOS data though) but failed at a later step (the missing kz parameter probably). Currently, it’s hard to tell why your workflow fails at step 5 but maybe we could put joint efforts in this (I also wanted to test the tomographic potential of Sentinel-1).

Hi @ABraun,

So for kz parameter image,
we need perpendicular baseline, slant range distance, and incidence angle other than wavelength.

For perpendicular baseline, you should have it after running S-1 Back Geocoding.

For slant range distance, I calculate it using slant_range_time (round-trip time in nanosecond) in the β€œTie-Point Grids” of the product.

image
? is the product number in your Product Explorer.
299792458 is speed of light in m/s.

For incidence angle (in radian), similarly, you can find it in the β€œTie-Point Grids”.

With this information, you are able to calculate the kz parameter :smiley: Just need to be aware of the unit!!!

Correct me if anything above doesn’t sound right to you :upside_down_face:

Cheers!

2 Likes

Thank you for sharing and explaining - sounds right to me!