Problems in terrain correction removal for S1 GRD product

Microwave Toolbox
#1

Hi there,

I’ve tried to process the Sentinel-1 GRD products.
And I applied the steps with the corresponded gpt operators as follows,

[Test 1]

  1. Apply orbit file (Apply-Orbit-File)
  2. Remove GRD Border Noise (Remove-GRD-Border-Noise)
  3. Thermal noise removal (ThermalNoiseRemoval)
  4. Calibration to Sigma 0 (Calibration)
  5. Subset (Subset)
  6. Range Doppler Terrain Correction (Terrain-Correction)
  7. Conversion to dB (LinearToFromdB)

It worked well, and for the next trail I would like to remove the terrain correction step as the study area is on the ocean. But I need to project the results into WGS84/Auto UTM. So, I used the reproject operator (link as reference).
As reproject operator needs the input files to be terrain corrected or ellipsoid corrected, I revised the structure as follows,

[Test 2]

  1. Apply orbit file (Apply-Orbit-File)
  2. Remove GRD Border Noise (Remove-GRD-Border-Noise)
  3. Thermal noise removal (ThermalNoiseRemoval)
  4. Calibration to Sigma 0 (Calibration)
  5. Subset (Subset)
  6. Ellipsoid Correction (Ellipsoid-Correction-GG and Ellipsoid-Correction-RD, I tested both respectively)
  7. Reproject (Reproject)
  8. Conversion to dB (LinearToFromdB)

And then, I mosaicked [Test 1] and [Test 2] results, and found that [Test 1] worked well, but most of the points were missing in [Test 2].

I am not sure whether I had a wrong procedure or not.
I would like to ask for the possible reasons.

Thanks in advance.

#2

why do you not select UTM (Auto) in the Ellipsoid Correction directly?

grafik
grafik.png800×617 17.4 KB

#3

Ahhh… well, thank you.
And I removed the reproject step, then it worked well.