Displacement from Sentinel1

[MarcoSal]

Hi,
I’m interested in determining the displacement of Earth’s surface from Sentinel data. Do you know free software that allows it or a step-by-step guide?

thanks in advance
MarcoS

[ABraun]

A nice tutorial with TerraSAR-X data is given here:
https://saredu.dlr.de/unit/insar_deformation

InSAR processing with Sentinel-1 data is demonstrated here:
http://sentinel1.s3.amazonaws.com/docs/S1TBX%20TOPSAR%20Interferometry%20with%20Sentinel-1%20Tutorial.pdf

Theory behind InSAR
http://earth.esa.int/workshops/ers97/program-details/speeches/rocca-et-al/

This source describes how fringes can be interpreted:
http://vldb.gsi.go.jp/sokuchi/sar/mechanism/interpretation-e.html
http://vldb.gsi.go.jp/sokuchi/sar/mechanism/mechanism04-e.html

These topics may also clear some questions:

Some explanations about concepts of fringes of interferogram and coherence

Dear all, 1) The phase difference can have contributions from five different sources: Δφflat is called flat Earth phase which is the phase contribution due to the earth curvature.  Δφelevation is the topographic contribution to the interferometric phase.  Δφdisplacement is the surface deformation contribution to the interferometric phase.  Δφatmosphere is the atmospheric contribution to the interferometric phase. It is introduced due to the atmospheric humidity, temperature and pressure chan…

Subsidence map in 3d view

is there any formula exist for converting the unwarpped phase into metric unit..... and how we create a 3D map of subsidence.... someone please help me....

[MarcoSal]

thank you very much. eccellent materials!!

but, there are services that do this type of calculations at a low cost?

[ABraun]

none of the mentioned data or programs is linked to costs, they are all free.

Speaking of services, the Geohazards TEP platform provides online generation of interferograms:
http://terradue.github.io/doc-tep-geohazards/tutorials/diapason-iw.html
https://geohazards-tep.eo.esa.int/#!pages/informationProcessing (DIAPASON)

[5h0hr3h]

This is an image of my area. why the left portion of displacement map is like this?

[falahfakhri]

That’s mean the slave image doesn’t match exactly the master image, consequently there is no data to be calculated.

[5h0hr3h]

Thanks for your answer and your consideration. What work do i need to do that this issue doesn’t happen?

[falahfakhri]

Nothing you could do unless, your slave image matches your master image, so this missing data doesn’t related to any error.

NB/ In case your data are matching each other might be this missing data could be appeared in case you use internal or external DEM, doesn’t match the entire image.

[sentinelgyb]

Have you solved this problem? I met the same problem with you.

[zahra0729]

Hi dear, I have a quastion about interferogeram unwarping, when I unwarppe whole interferogeram image, it gives a big number as displacment value, but when i unwrappe just focal region the number get change and give small displacement value for me…what is the reason?? which value is the true one in your opinion… !! pls explain the reason to me.

[ABraun]

unwrapping is an iterative process. SNAPHU choses a random point where to start and adds up the fringes outgoing from that. So different sizes can potentially lead to different results, but the differences shouldn’t be too large. Always also look at the coherence of these areas. Low coherence often means unusable results.

[zahra0729]

Thanks for your answer, but when I use small size of my ifg image, displacment value gets about -30cm for focal region…but when I use whole image it gets about -98 cm…so i dont know wich value is true!!

[ABraun]

hard to tell without seeing the data. It often comes down to interpreting the fringes of the interferogram correctly. Do they look different in the two subsets? Furthermore, you should never stick to the value of one single pixel. Maybe you define a small polygon and take the average of it using the histogram/statistics tool.

[zahra0729]

This is my whole image…and when i do unwrapping on this image, displacement values are about 1meters for my interested pixcel.

[zahra0729]

But when i just work on small piece of image, result gets different

[ABraun]

As you see, there are only few aeras with usable information. I personally wouldn’t trust any values in the noisy areas at all.
You might also try to set the coherence threshold (e.g. 0.4) at the step of interferogram generation.

What is important are the patterns of the fringes. They have to be the same in both images.

[zahra0729]

Yes they are the same, but when I do snaphu export from them and do unwrapping seperately , the result for displacement is different!
I have another quastion too, this is my final image before unwrapping step…and i applyed filters and other steps correctly…but why the result image looks noisy yet!!

[ABraun]

but why the result image looks noisy yet!!

filtering can only smooth the fringes you see already. If there is no coherence you cannot expect the results to get better after filtering.

This is a very good document: https://www.esa.int/esapub/tm/tm19/TM-19_ptA.pdf

[zahra0729]

Thanks alot,
Then the difference between displacement results maybe due to noises, so we can tell when the size of image to get smaller and we select smooth and unnoisy pieces of images, result of unwarping is more correct!!??

[ABraun]

yes, try other image pairs with better coherence. Have a look at this document http://www.esa.int/esapub/tm/tm19/TM-19_ptB.pdf
Section 1.3 Selecting images for InSAR DEM generation