The phase diagram is illogical after unwrapping

Hi, I recently analyzed the deformation field of Qinghai earthquake. But my unwrapped phase diagram is very illogical。
f5caf24b04bb922ba6619bece546d01
But the phase value is about this
d7458ffe2d6134f56e51241c2697206
It seems that some symbols are opposite, but whether using different images or different unwrapping methods (MST, MCF), the final result is the same


The curve of the second half is always about - 1, which is not the conventional 0。I don’t know why, but I’d appreciate it if you could give me some advice. Thank you

1 Like

The 2021 magnitude (Mw) 7.4 Maduo Earthquake in Qinghai is difficult for measurement with the Sentinel-1 InSAR pairs. The area near the main fault rupture has very large and complicated deformation that reduces the InSAR coherence to very low values. The combination of very low coherence and very large deformation makes it very difficult to unwrap correctly. In addition, the long fault rupture extends across almost the full width of the Sentinel-1 IW swath, so there is no easy way to determine the correct phase unwrapping between the two sides of the fault.

You correctly analyzed the unwrapping results that are illogical because there are many unwrapping errors. The first thing you should do is look at the coherence layer and ignore the unwrapped phase in the area where the coherence is less than about 0.4. There may still be large-scale unwrapping errors that affect the relative phase between the north and south side of the fault. You would need to estimate what is the remaining phase shift between the two sides by assuming the phase goes back close to zero far away from the fault rupture.

4 Likes

@Ori3570

I suggest one more approach - try to reduce the bursts to certain limit (2 -3) and try again from the beginning. You should be lucky if the epicenter is in one sub-swath.
It was recognized that to some extent this eliminates the unwrapping errors.

Do you use both types of orbits? Could you please provide the names of the products?

1 Like

Orbit : S1A_OPER_AUX_POEORB_OPOD_20210609T121911_V20210519T225942_20210521T005942.EOF
S1B_OPER_AUX_POEORB_OPOD_20210615T111837_V20210525T225942_20210527T005942.EOF.3
product:
S1B_IW_SLC__1SDV_20210526T232754_20210526T232829_027082_033C48_DA5D.zip
S1A_IW_SLC__1SDV_20210520T232820_20210520T232847_037978_047B84_031C.zip

I had ignore the unwrapped phase in the area where the coherence is less than about 0.4.And i also reduce the bursts.But I don’t know how to assume the phase goes back close to zero far away from the fault rupture.I’m new to snap,so can you elaborate?

As far as I know, you can’t adjust the phase of one part of an interferogram inside SNAP. I do that as post-processing after exporting the final unwrapped interferogram and working with it in other programs, such as GMT or Python.

To be honest, I haven’t been in touch with GMT and I’m a novice on py. Can you share your code? and explain to use it .

I have limited the range to four debursts in one IW , but the unwrapping effect is still not ideal.

The big problem with the 2021 M 7.4 Maduo earthquake is that the rupture is very long and extends almost all the way across the full Sentinel-1 scenes, so there is no part of the interferogram width that is not affected by the earthquake deformation. Reduced coherence in the mountains and zero coherence in lakes add to the difficult of correct phase unwrapping.

3 Likes