Radiometric & Geometric Correction Workflow

are you sure you are using SLC data? In this case you would need to perform TOPSAR Deburst after Calibration in order to remove the black stripes.

Coregistration is only needed when you work with multiple datasets at once.

You could also make a subset earlier in order to save computing time .


I try to obtain the backscatter coefficient of Sentinel-1A using SNAP. Is the radiometric calibration the first step?

depends on the level of your product. Have you downloaded SLC or GRD data?

SLC, IW mode, VV and VH polarization

If I perform TOPSAR Deburst, Filtering and radiometric calibration, the results of VH polarization is Ok, but the values of VV polarization are zeros.

for SLC data I would suggest

  1. TOPS Split
  2. Apply Orbit file
  3. Thermal Noise Removal
  4. Calibration to Beta0*
  5. TOPSAR Deburst
  6. Radiometric terrain flattening
  7. (Speckle filtering)
  8. Range Doppler Terrain Correction

If you don’t need the phase information you can also download it as a GRD product and only apply the following:

  1. Apply Orbit file
  2. Thermal noise removal
  3. Calibration to Beta0
  4. Radiometric terrain flattening
  5. (Speckle filtering)
  6. Range Doppler Terrain Correction

or (if you don’t have much topography):

  1. Apply Orbit file
  2. Thermal noise removal
  3. Calibration to Sigma0
  4. (Speckle filtering)
  5. Range Doppler Terrain Correction
Sentinel1 callibration procedure
S1 Pre-processing chain and parameter setting
Sentinel-1A GRD product
Which one is 'Terrain Correction & Radiometric Correction' standard workflow?
What does the APPLY orbit file for?
Pre-Processing Sentinel-1-Data
Incidence angle correction of Sentinel 1
Normalize Incidence angle
Why SLC sentinel-1 product has 18 intensity and Amplitude
Snap Error: Source product should not be map projected S1
Problem with terrain flattening
Can someone tell me the preprocessing steps for Entropy-alpha decomposition of sentinel-1 dual pol SLC data?
Soil Moisture Mapping Using S1_SLC product and s1tx
Speed up georeferencing of GRD-Data
Use of VV & VH polarisations for Water Detection
Sentinel-1, Ascending and Descending orbit
Oil spill application on SNAP
Extract the backscatter coefficient value
Sentinel-1A IW SLC data
Estimating Forest Above ground biomass from sentinel imagery using gray level co_occurence matrix
SAR image view in SNAP
Pre-processing SAR
Basic SLC processing for marine object detection
Bands in Sentinel 1 Level 1 (GRD, RAW and SLC) products - basic information
PALSAR-2 ScanSAR processing in SNAP
Intensity and Coherence Correlation
Crop classification using SENTINEL-1A SLC data
Geocoded and orthorectified images to full resolution
Equalization SLC and GRD images with SNAP
How to calibrate, geocode ERS-2 and Envisat SAR Images using gpt
When do we need Apply-Orbit file
Sentinel 1 images SLC and GRD
Backscattering for Soil Moisture
Unit Conversion
Exporting sentinel 1 band to Geotiff
Image Fusion Using Sentinel 1 and Sentinel 2
How to load sentinel 1 data into the Snap?
Using Sentinel 1 in Arcmap
Merging of two different satellite data
Bands in Sentinel 1 Level 1 (GRD, RAW and SLC) products - basic information
Problems with the Range Doppler Terrain Correction tool
About Sentinel 1 vegetation index
Backscatter normalization in Sentinel-1
Land cover classification using sentinel 1 GRD data
RCS values for land surface types - Sentinel 1
Estimating Forest Above ground biomass from sentinel imagery using gray level co_occurence matrix
examine the max., min., mean, median and standard deviation of spectral signatures in S1 products
Sigma nought values change working with Sentinel-1 SLC
Problem with split the image
Alos palsar 2 orthorectification
Sentinel-1 Deburst
Calculate the ratio of the backscatter coefficients
Questions about the processing steps for multi-temporal sentinel-1 SLC data
Georefering S1 images
Soil moisture from sentinel-1 (GRD Data)
Elaborate S1-GDR images to obtain FSMH (forest stand mean height)
Workflow for S1 image

Thank your very much for your suggestions. Does the Thermal noise removal and Apply Orbit file are the necessary steps?

1 Like

I performed the SLC data using SNAP
1 TOPSAR Deburst
2 Calibration to Sigma0
3 Speckle filtering
4 Range Doppler Terrain Correction
For the radiometric calibration results, the VV is right, but the VH is incorrect.

Thermal noise is not present in every product. In case you don’t rely on highly accurate radiometric values you can also skip it.
The orbit file adds information about the geometric location of your product. If you want to combine it with other data it can improve the position accuracy.

Your suggested workflow is sufficient for most applications.

Why do you think calibrated VH is not correct?

1 Like

Thanks for your reply, I follow your suggestion to perform the Sentinel-1A data using SNAP. In the previous results, when the radiometric calibration was performed, the results of VV polarization are zeros, only the VH polarization is correct. Can you supply an email for me, thus I can transform the results to you.

sorry, I can’t help you outside this forum.

But indeed, zeroes as an output are obviously wrong. Was the data correct/complete before radiometric correction?

Sorry, the results of VH is correct, but the calibrated VV are zeros.

Yes, before the radiometric calibration both of the intensity of VV and VH are correct. When the radiometric calibration was implemented, the calibrated VV are zeros, and the calibrated VH is correct.

I can’t think of a reason why this could happen.
Is it zero or no data /nan?

What order should I take steps in ?I’ve seen one of the above threads says that it should be good after orbit to apply thermal noise removal in order to decrease radiometric variation between subswath.
Best regards!

could be a good idea :slight_smile:

Thanks for your reply .Do you mean orbit before thermal noise removal is better ?What are the differences between the various epolynomial degrees and which is recommended?
Thank you for help

I really don’t know, sorry. Try it out and compare the outcomes.

Thank you very much !

I understand the theoritical difference between beta0 and sigma0, if I want to do intensity correlation analyisis between 2 images, or in this case 3 images. I was usggested to use sigma0, however the area I am working is very hilly and has some mountains. The values of sigma0 and beta0 are ofcouse different from pxel to pixel, but only beta0 allows me to do terrain flatering.

What do you suggest I can do for my application in this case?
Thank you