Remove incidence angle effect in Sentinel-1?

[marifo]

Hi
I am using a portion of my time series sentinel-1 images (both IW and EW). The problem is that I didn’t get the result that was expected and my features evolution in time is strange. Well, they are all over the red square in the image attached and I thought that maybe it is because of the effect of incidence angle.
For the pre-processing I uased :Thermal Noise Removal–> Apply oribit file–> Calibraiton to Beta–>Radiometric terrian flattening --> Range Doppler terrrian correction.
My Question is that is there any other pre-processing module I can use for incidence angle correction OR is there any easy way or equation that I can use for ?
Thanks a lot for your help in advance

[mengdahl]

Can you post a zoom and describe in more detail what you think is unexpected? Please note that the last two steps in your processing depend only on the observation geometry & the DEM, so if you are using the same orbit track & DEM these corrections do not evolve in time.

[marifo]

Hi =, Thanks, Nothing is clear in the zoom in of just one image, but I can explain that I am looking at small lakes here and their ice phenology through about 70 images. However, it does not follow the rules comparing to their size and depth. or for example some backscatter like water in the very freezing time of the year, and show ice in the warm time, which is also not in consistent in the whole image. I use a threshold for defining open water from ice in each image and seems something going wrong. I thought that maybe the problem is in incidence angle differences for different lakes in the image. I need to remove its effect so that I can analyse the ice phenology pattern better. Is there anything you can think of to correct its effect?
Thanks for your reply in advance

[mengdahl]

Wind on water and old ice have higher backscatter than very new ice or still water that can be close to the noise-floor. I don’t think there’s a way to solve it with simple thresholds.

[marjanmarbouti]

Hi, if our incidence angle difference between near range and far range is very small in order to 1 or 2 degree, can it bring any effect on our result?
Thanks

[ABraun]

I’d say this depends on how you generate the result. Do you work with backscatter intensity or coherence/phase?

[marjanmarbouti]

Coherence/phase…

[marjanmarbouti]

Actually I did not see any effect over backscatter intensity, coherence, phase, unwrapped result; the only problem was with phase to elevation that I saw the effect on that.
I guess maybe it is because of incident angle effect but from near to far range; the difference is 2 degree that I do not think so can make effect??

[marjanmarbouti]

connected to this topic

A wierd gradient over my result

I did but it has same form.[quote=“marjanmarbouti, post:1, topic:13511, full:true”] Dear all I used TanDEM-X data and made a DEM from them. Steps are listed in below: Interferogram formation unwrapping Phase to height Applying boxcar filter (5 by 5) But if you look at my final result, there is a blue gradient over my filtered phase to height. How can I remove it? [PNG] Cheers [/quote] @falahfakhri I did but same shape.

I remember, you introduced me some references for it (phase ramp) but I only think that maybe it also be because of incidence angle…

[ABraun]

coherence/phace impacts caused by the global incidence angle are compensated by the flat earth estimation of the interferogram operator.

[marjanmarbouti]

Thanks.

1. Yes, thanks and as I know, this is something that SNAP does in automatically when we make an interferogram. Is not is?

2. one more question; when we unwrapped the phase; then what is the unit for it? I got this plot by using profile plot in SNAP.
   
   

[hriston_bg]

@marjanmarbouti the unit of the unwrapped phase is number of abs.phases.
this what i always get after unwrapping - see below

[marjanmarbouti]

@ABraun

coherence/phace impacts caused by the global incidence angle are compensated by the flat earth estimation of the interferogram operator.

Thanks for help

[marjanmarbouti]

I tried to find out that large and small incidence angles have any different effects on DEM retrieval or not? Would you please suggest me any paper or idea about it?
For example, if satellite has 40 degree incidence angle in comparison with 20 degree; is there any difference in height or DEM retrieval?

[ABraun]

this study reports higher coherence for small incidence angles because of higher Signal To Noise ratio: https://www.sciencedirect.com/science/article/pii/S092427161730093X#f0020

But the actual phase is not mentioned. I think there is no genral truth, because the incidence angle interacts with the steepness and orientation of the topographic features.

This study provides a more systematic investigation: http://earth.esa.int/fringe03/proceedings/papers/33_eineder.pdf

The relationship of incidence angle and topographic distortions (layover/shadow) is analyzed here: http://www.esa.int/esapub/tm/tm19/TM-19_ptA.pdf (Figure 1-11)

[marjanmarbouti]

Yes, looks clear for me. Thanks.