Temporal Coherence masking

402 views
Skip to first unread message

Saeed Mhanna

unread,
Aug 21, 2023, 4:24:35 PM8/21/23
to MintPy
Dear MintPy users, 

I am using MinPy from the OpenSARLab notebooks while building a stack from ASF vertex. 

The entire workflow works nicely, but the temporal coherence mask generated masks out a large part of my AOI. The spatial coherence however covers it all. I tried changing Temporal Coherence threshold (0.7 to 0.4) but it that didnt work. Is there a limit on the number of pixels that can be inverted to time series? What can be causing the large masked area?
I also attached the numTriNonzeroIntAmbiguity.png in case it helps with the analysis.

Your help is highly appreciated, 
Saeed Mhanna
PhD student at the University of Neuchatel
temporal_coh.png
numTriNonzeroIntAmbiguity.png
avg_spatial_coh.png

Eric Fielding

unread,
Aug 21, 2023, 11:23:22 PM8/21/23
to MintPy
The temporal coherence is only high in the center where all of the interferograms have coverage. You might notice that the center also has higher spatial coherence. One big disadvatage of the ASF Hyp3 Gamma on-demand processing from ASF Vertex is that it only processes the parts of pairs where the two input slices overlap. In some locations, the input slices shift back and forth and they overlap only in a small area. I don't know of a way to combine multiple interferograms from different input slices to get full coverage with teh ASF Hyp3 Gamma processing. You may have to use the ISCE2 package that knows how to combine multiple slice input files to make complete interferograms.

++Eric

Saeed Mhanna

unread,
Aug 22, 2023, 7:54:44 AM8/22/23
to min...@googlegroups.com
Many thanks Eric for your response, 

Do you think using ARIA products from ASF might help in that regard?

My AOI is covered entirely by this dataset. 

--
You received this message because you are subscribed to a topic in the Google Groups "MintPy" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/mintpy/3oK2fZmQBcg/unsubscribe.
To unsubscribe from this group and all its topics, send an email to mintpy+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/mintpy/4854a690-9e42-4d0e-a61a-db7c80943e0en%40googlegroups.com.

Eric Fielding

unread,
Aug 22, 2023, 2:25:47 PM8/22/23
to MintPy
Yes, if there are ARIA GUNW files, then they will work much better. The ARIA-tools package can download the multiple interferograms for the same date pair and stitch them together for MintPy analysis of the whole area. You should use a larger area (about 150 km north and south of the area that you want to study) for ariaDownload script to make sure you get all the interferogram slices. 

Saeed Mhanna

unread,
Sep 9, 2023, 3:35:09 PM9/9/23
to MintPy

Thanks Eric, 

Indeed, the ARIA products have much higher coherence and covered my study area efficiently. I am now trying to follow the UNAVCO 2022 course. I am still using OpenSARLAb, but I do not have access to the UNAVCO conda environment. 

Is it available online? (did not find it on the github page). if not, would you please share the .yml and other requirement so i can install it? 

Best regards, 
Saeed

Eric Fielding

unread,
Sep 10, 2023, 1:17:45 AM9/10/23
to MintPy
Hi Saeed,
You can use the “insar_analysis” Conda environment in the OpenSARlab system. It has the same ARIA-tools and MintPy software as the UNAVCO2022 course.You don’t need to install a new environment.

++Eric

Saeed Mhanna

unread,
Sep 17, 2023, 11:19:10 AM9/17/23
to min...@googlegroups.com
Hello Eric, 

I am facing some issues while trying to build a stack for ALOS-1 data using stripmapStack. I submitted an issue on the isce2 github. 

I would really appreciate any comments from you and the mintpy group. 

Best wishes, 
Saeed

Saeed Mhanna

unread,
Sep 18, 2023, 7:46:33 AM9/18/23
to min...@googlegroups.com
Many thanks as always Eric, 

It seems that there are much more images available on ASF for 1.1 level data. Is there a way to use those instead of starting with the 1.0 ALOS-1 images? 

Cheers, 
Saeed

Eric Fielding

unread,
Sep 18, 2023, 3:54:56 PM9/18/23
to MintPy
Hi Saeed,

I expect that all ALOS-1 scenes at ASF that have level 1.1 products also have level 1.0 products, so I would check your search method. It is possible that some scenes don’t show level 1.0, but the ASF has them and your search did not find them.

I believe that ASF is processing L1.1 from the L1.0 data, so they should have the L1.0 somewhere.

It would take a small amount of Python programming to make a version of the unpackALOS program to work on SLC or L1.1 data.

Saeed Mhanna

unread,
Oct 11, 2023, 9:01:36 AM10/11/23
to min...@googlegroups.com
Hello Eric,

Thanks for getting back to me.

Can you please have a look on this issue i created on the ISCE github repository?

Would really appreciate your help!

Saeed

Eric Fielding

unread,
Oct 15, 2023, 12:15:50 PM10/15/23
to MintPy
Hi Saeed,

I found out that there is a `ALOS_SLC` sensor now in ISCE2, so you can use that to process the ALOS L1.1 data. It seems that recent ALOS data at ASF may only be L1.1 and L1.2.

++Eric

Saeed Mhanna

unread,
Oct 16, 2023, 4:32:19 AM10/16/23
to min...@googlegroups.com
This is great! I will check it out. 

I asked the ASF Vertex team about this and they said JAXA is no longer providing L1.0 data so its great to have the SLC processor in ISCE for ALOS. 

Cheers, 
Saeed

Reply all
Reply to author
Forward
0 new messages