Difference between vertical and LOS displacement

You need an ascending stack and a descending stack.

And then?
How can process them together?

If you want to have a "sure" vertical displacement you would need to do PSI-processing on sufficiently large ascending and descending image stacks. This is not an easy thing to do. Processing a pair of images will not work since atmospheric disturbances usually completely mask the deformation signal.

thank you so much.
I was hoping it worked out considering that I found many user guides, manuals and discussions on this forum explaining how to find vertical displacement with snap and snaphu.
In any case, I do not know how to do the PS method and it is very important that I deliver this work.
Is there anyone who has already done it and can explain to me step by step what to do, what other software do I need and how to install them?
I looked for on the web and I have found a few infirmations and the little information found only confused me.
Thanks for your help.

a good start is the work of Andy Hooper:

But we also have an own category about this in our forum: About the STaMPS category

1 Like

It is very demanding processing, I would guess that it takes several months of solid work&studying to understand what is going on. You might be able to duplicate someone else’s processing but that is always risky as every case is different and without deeper understanding of InSAR and PSI troubleshooting will be hard. Whoever gave you this task should also understand this. I hope this helps!


Thanks for your support…

Know someone who could make me a course or distance learning on this?
Please let me know.
Thank you.

@ABraun alrady provided some useful links. You could also have a look at materials from the MOOC Echoes in Space - they have also some videos on youtube.
Otherwise you could apply to attend our Land Training Course 2018 http://eoscience.esa.int/landtraining2018/ :slight_smile:

Ok thank you very much

In the meantime, are there any online courses or old courses that can be downloaded?
I looked at the site you indicated to me but there is no material to download

Good morning,
I have processed two images and I have obteined the displacement trought “Phase to displacement”.
Anyone knows how can I found the maximum and minimum displacement?

Good morning,
I have followed the following steps in order to find displacement:

  • 2 SLC_IW images
  • Stack–> S1 Tops Coregistration
  • Interferogram Formation
  • Deburst
  • Topo Phase Removal
  • Goldstein Phase Filtering
  • Subset in my AOI
  • Export to Snaphu
  • Unwrapping in Snaphu
  • Import of “Subset” and “Unwrapped phase” in snap
  • Phase to displacement
  • Terrain Correction
  • Terrain correction - pixel value of a pixel with low displacement
  • Terrain correction of coherence taken from the import of snaphu
  • Band math–>if coh>0.6 then 1 else NaN, in order to have only pixels with high coherence
  • Stack between (high coherence) and (Terrain correction of displacement-pixel value)
  • High coherence * Terrain correction of displacement-pixel value

I think that my steps are correct considering what I have found in many user guide and in the suggests of this forum.
But I would like to know if the parameter setted and the conditions flagged or not for each steps are correct or not.
Following the screenshot for each steps:
I hope that someone can tell me if everything is correct or not.




technically, your workflow seems right to me. It would be interesting however how the interferogram and coherence layer look.

Thanks for your reply.
I have obtained the following values of displacement:

min: -0.096m
max: 0.053m

The strange thing is that I am sure that in some point of this area the displacement should be -10m.
Now I am tryng to do the same steps in an area where there was an heartquake in order to verify is my process is correct or not.
So, I have inserted and flagged the right data in all the steps that I have done?

it is important to understand that DInSAR gives information about the spatial trends of a displacement. It is not a good idea to trust the values of single pixels as many factors are impacting the result.

Additionally, DInSAR measures the displacement along the line of sight, not the actual vertical displacement. Maybe this explains some of the discrepancies.

As suggested before, why don’t you assume the displacement of 10m at the area where you are sure and use it as the reference height for your displacement raster?

I’m sorry but neither DInSAR not PSI can detect a large movement in a single pixel. Not even in the case where the scattterers were undisturbed, which is absolutely not the situation in your case. As I said earlier PSI might be able to detect a precursor subsidence at the site of the collapse, but since it is entirely possible that there was no precursor-movement, even that effort could well be futile.

Because theoretically I have to found that value…

Thank you.
At this point, however, I do not understand what this method is used for.
I have read many posts in this forum that explain how to get both LOS and vertical displacement by applying the following formula:
(Unwrapping phase * 0.055) / (- 4 * PI * cos (rad (incident_agle)).
I also found many ESA guides about the interferometry in order to obtain vertical displacement and the steps are the same of those I have applied.
Why I cannot obtain the desidered results?
Please explain it to me.

Any movement in a single pixel (with stable scatterers) that is larger than 2PI in phase is ambiguous. It’s impossible to know whether that movement was Y or Y + N x 2PI where N is any integer. In other a words movement of Y, Y+2PI or Y + 76 x 2PI all produce the same observed phase. There’s no way around this limitation and there’s no way to unwrap the phase of such single pixel to retrieve the actual displacement.