Generate Sentinel-1 coherence time-series

[yutingzhou]

Hi all,

I am trying to process the S1 SLC data into coherence time-series. However, I had no previous experience on that and I got stopped at the first step.

My question is can I use any two images to generate coherence or I have to use images with a 12-day revisit with the same orbit and slice number? Thank you.

Yuting

[falahfakhri]

No need to create a new topic, a lot of topics talk about coherence and more details you could find out step by step.

Bug in Create Stack - Stack of Coherence + Intensity

Dear Step Forum, I am about to create a complex RGB composite following the tutorial from step user mfoumelis here: http://seom.esa.int/landtraining2015/files/Day_5/D5P1a_LTC2015_Foumelis.pdf On page 50-52 a data stack is created for the partly preprocessed Intensities and the coherence. I try to follow the exact workflow of all slides, but the “Create Stack” function does not work, since there is a complain saying: S1 TOPS SLC products should use TOPS Coregistration. So I tried the TOPS Co…

[ABraun]

for Sentinel-1, any multiple of 6 is valid, as long as the same track/relative orbit is used. But the longer the period between the image, the lower is the coherence.

Subsidence map in 3d view

thank you, @heinzollerketchup. Generally, the longer the temporal baseline (=the time between two images) the higher is the decorrelation which leads to small coherence values. It also depends on the cover of the analyzed surface. Bare ground mostly shows slower decorrelation than one with high vegetation cover. As you can see here (tested with S1 data), coherence decreases from 0.7 to nearly 0.1 within 6 12-day-cycles (12, 24, 36, 48, 60 and 72 days, blue to red) [image] Source: http://ieee…

[yutingzhou]

Hi Falah,

Thanks for your remind. This is my first time to use the forum and I was not sure how to ask a question. I will be more careful in future.

Yuting

[yutingzhou]

Hi Andreas,

Thanks for the quick answer. I have already read the post you gave.

My study area is in US so the revisit time is 12-day. The study area is in the overlapped region of Orbit 34 Slice 17 and Orbit 107 Slice 17. Can I use these two images to generate coherence? I guess the answer is no judging from your response. It seems like only products with the same orbit and slice numbers should be used to generate coherence. Just want to make sure my judgement is right.

Yuting

[ABraun]

exactly. I made an illustration here: INSAR / DINSAR - Perpendicular baseline calculation

[falahfakhri]

No, problem, here you could find a lot of topics,

[yutingzhou]

Thanks, Andreas. This answered my question.

[yutingzhou]

Thanks, Falah. I will keep reading topics in the forum.

[yutingzhou]

Hi Andreas,

I have a question again. My study area (the yellow square) is in the overlapped area of two images. The top image is S1A_IW_SLC__1SDV_20160705T002810_20160705T002837_012006_01288F_4175 and the bottom image is S1A_IW_SLC__1SDV_20160717T002753_20160717T002821_012181_012E4D_DF35. Do you think it is ok to use them to generate coherence? Thank you.

Yuting

[ABraun]

yes, you can combine them because they are of the same track. The usable area is only the overlapping part, of course.

[yutingzhou]

That is a good news for me which means that there are more data available to generate coherence time series.

I was thinking they cannot be combined as they have different incident angles. Seems like I have to read more about InSAR. Thanks again.

Yuting

[ABraun]

there was a similar discussion here: S1A-S1B coherence of partially overlapping images
Please have a look at the explanations of mengdahl.

[yutingzhou]

Thanks, Andreas. That is a really useful discussion for me.