yes, elevation_VV is the InSAR DEM.

The problem in your case is probably the perpendicular baseline. If it is too small, the fringes in the interferogram represent too much elevation difference.

Please have a look at the InSAR Stack tool to check the perpendicular baseline and the altitude of ambiguity of the image paris. The latter determines how much elevation difference is represented by one fringe.

Can you please also show the interferograms of the three pairs?

1 Like

Thanks for reply. I will try to show main process steps image.

Perpendicular baslines;


Filtered Phases;

Snaphu export parameters for all pairs;

Unwrapped phases;

Coherence maps;

1 Like

according to these guidelines a perpendicular baseline of 150-300 m is recommended. While the first one is clearly beyond that, the second ones are rather short.

InSAR Principles: Guidelines for SAR Interferometry Processing and Interpretation (ESA TM-19)

I suspect that the first one suffers from baseline decorrelation (please check the notes on the critical baseline), the second ones are not detailled enough to provide information on fine topographic variations.
Lastly, temporal decorrelation clearly leads to low coherence and random phase signals in the second and third. Is there much vegetartion in your area?

1 Like

Yes, there are many agricultural lands in my study area.

Do you think, is there any problem unwrapping parameters?

What would you suggest I do to find a solution to my problem?

the unwrapping parameters can help to remove smaller errors or tweak the best possible results, but they cannot restore information lost due to phase decorrelation.

Ideally, you have a 1-day image pair with 150-300 m of perpendicular baseline from the dry season with the least possible vegetation cover. That’s all you can do from the data acquisition side. If this still results in bad quality, there is too much vegetation in the area for proper interferometry.

Thank you very much for your answers.

As far as I understand, it is difficult to produce a healthy DEM using these data, isn’t it?

I’m afraid, yes. Any place on earth with vegetation is difficult for interferometry unless you have bistatic data.
Please also have a look here: Generate DEM in forest area

1 Like

Thank you very much for your advice. Best wishes