Sentinel 1 Processing Chain

Microwave Toolbox
#1

Hello everybody,
i`m new with SNAP. So I have a question about the minimum steps for a radiometric and terrain calibrated Scene.
In my case, i have a SAR-Sent-1-scene from ESA with Level1, GRD and VV-Polarisation.

First of all, what is the difference between the magnitude-band und intensity band?

For an optimal resulting scene, I would do the following steps

  1. Calibrate -> output: sigma0
  2. Input:sigma0-> Speckle (lee5x5)->
  3. Rang-Doppler terrain Correction

Did I forget something? Should i do a terrain-flattening and if so, at wich position in my Chain?

I read that its recommend to execute a terrain-flattening, but after the calibration i get a sigma0 band and for terr-flatt i need the beta0 band.
Wich band should i use for speckle beta0 or sigma0 (In the case i even need a terr-flat)?

Thanks for help!
A5iveMM

#2

The intensity is just the square of the amplitude. It’s just available for convenience.

The terrain flattening is used to remove the bias from the terrain for classification purposes.

It depends what your intended use is. If you just want to mosaic or overlay on optical imagery and you don’t mind that the illuminated side of mountains appear brighter then just calibrate and terrain correct as you have.

#3

Thanks for the fast reply!
Finally I want to generate a cross-correlation scene between two time-different Input radar scenes.
For example:

  1. Scene: 1.1.2015, 2. Scene: 1.1.2016

The Earth-Surface represent with DN of radar-sensing moves with time. I need two radiometric and terrain calibrated scenes to correlate them later.
At the moment i am trying to use Cosi-Corr for the cross correlation, but i dont get good results. Do you know if SNAP include a tool for cross-correlation two time different radar-scenes?
Thank you

#4

First apply the precise orbits then use the DEM Assisted Coregistration. If you have movement in your scene which you want to analyze then don’t use the Dem Assisted Coregistration with Cross Correlation and Warp.
Instead, after the coregistration you could try the offset tracking. This applies cross correlation to many windows on the coregistered product to give you a velocity map.

1 Like
#5

Thanks a lot.
Do i use the coregistartion with the original files or with the radiometric/despeckled and terrain calibrated data ?
Thanks

#6

ok, i got it.
here is a processing chain form Alaska Satellite facility. Thanks for that!

1 Like