Interferogram Formation - Subtract topographic phase check box

Usually while creating an interferogram, I use the “Interferogram Formation” program within ESA SNAP. I usually do not check the box that says “Subtract topographic phase.” From there, I deburst, apply the topo phase removal, filter, multilook, then unwrap.

If I check “Subtract topographic phase” what exactly does that do to my image and how would I change the rest of the process after that? Would I skip topo phase removal at that point?

If anyone has an interferometry recipe that includes checking this box, can you please share?

1 Like

If you want to perform differential inferferometry (phase changes between two images) you usually remove the contribution of the topography to the phase. so you get the real phase difference between your images.

You can either check the box “remove topography” or use operator for topo phase removal.

@ABraun Yes, I want to know how much the ground has moved between 2 time periods. A true differential interferogram.

Can you quickly outline the start to finish SNAP steps to achieve this please?

I can tell you the steps but you will have to read a bit to understand what is going on there and how the parameters and scene characteristics influence your results. ESA has compiled very good documents for this: http://www.esa.int/About_Us/ESA_Publications/InSAR_Principles_Guidelines_for_SAR_Interferometry_Processing_and_Interpretation_br_ESA_TM-19
This is also a nice tutorial but it refers to an old version of SNAP and many things changed in the meanwhile: https://saredu.dlr.de/unit/insar_deformation

  1. coregister the images
  2. create inferferogram and coherence
  3. subtract topographic phase
  4. optionally multi-look or filter the interferogram
  5. unwrap the interferogram
  6. convert unwrapped interferogram to displacement
1 Like

@ABraun I understand interferometry well and have been in the industry for many years, however, I have been working with SNAP for maybe 2 years processing Sentinel data. It is different from other software I have used before, Sentinel data is also different, and I am trying to get the order of operations straight. I see that there are 2 ways to remove the topography or subtract the topographic phase.

If I don’t check the “Subtract topographic phase” box during the interferogram formation, this is the process:

  1. coregister the images
  2. create inferferogram and coherence
  3. deburst
  4. subtract topographic phase (topographic phase removal)
  5. filter
  6. multi-look
  7. unwrap the interferogram
  8. convert unwrapped interferogram to displacement
  9. Terrain Correct

If I do check the “Subtract topographic phase” box, how does the above process change? what is the new order of operations? Here is what I think it is, let me know if I am right:

  1. coregister the images
  2. create inferferogram and coherence (here is where I would remove the topo by checking the “Subtract topographic phase”
  3. deburst
  4. filter
  5. multi-look
  6. unwrap the interferogram
  7. convert unwrapped interferogram to displacement
  8. Terrain Correct

Someone else said that you can either check the box or run the topographic phase removal separately and get the same results. I would like to know if that is actually true. When I have some time, I will do my own comparison, but can you confirm if that is true or not based off your own experience?

I am sorry, I couldn’t know about your background. Some users expect recipes for workflows they don’t yet understand and in these cases it is hard to help. This is why I mentioned the ESA materials.

The “subtract topographic phase” checkbox was only recently added and I haven’t tried it myself but I expect the results to be the same. It just combines two steps at once.
If you are working with Sentinel IW data, coregistration is a bit different, you at best use the TOPS Coregistration in the very first step where you can directly define the subswaths and bursts you need (as described here) which reduces a lot of time if you only need a smaller part of the image. But you have probably already seen this. For me, debusting after the interferogram formation worked best so I would agree to your second workflow.

Thanks! Yea, I saw that checkbox show up when I updated my software recently and had to ask about it. I appreciate you.

Hello everyone
I would like to ask in the STAMPS output process about “subtract topographic phase”.
If I chose the subtract topographic phase in the interferogram formation.
Because the flow of STAMPS output is interferogram formation—>topographic phase removal —>StaMPS output
Will there be extra deducted terrain phase in the subsequent topographic phase removal?

Thank you

Removing the topographic phase during the interferogram formation is enough. These steps used to be separate ones, so they still exist as single tools. But since SNAP version 6 they can be done in one step.
Both is fine though, unless you don’t apply it twice.

2 Likes

Thank you for your reply.
I have done both before, so I applied topographic phase removal twice.:pensive:
It seems that this is the reason why my all PS LOS velocity is very small … it’s wrong… Orz
Thank you very much, I will try again.:smile::smile: