Interferometry experiments using ICEYE data

[Nijat]

hi all
I’ve been studying on the forum for a while now and I really enjoy the various discussions about SAR. I tried interferometry with ICEYE data and it worked fine… Below are the results of the interferometry experiments, any discussion and suggestions would be greatly appreciated

Dataset provided by ICEYE here[Beyond Change Detection: Measuring the Changes that Matter]
Below are two images for interferometry (preprocessed for viewing only, SLC required for interferometry)

I used SNAP for data processing, the steps are the same as regular interferometry steps; InSAR stack information:

and the the coherence plot and the interferometric phase plot:

(It should be noted that the flat-earth phase has been removed from the interferometric phase)

In regions of high coherence, the interferometric phase has good continuity

[panjibrotoisworo]

Looks great! Can you share the workflow you used to create the data?

I tried to create a stack and coregister it using CreateStack but it have errors saying something about an empty or null string somewhere.

[mengdahl]

Try DEM-assisted co-registation instead.

[Nijat]

thanks sure
My processing steps are consistent with the ESA “Interferometry Tutorial with Radarsat-2” manual, and I used “Radar—— Coregistration——Coregistration”

The processing steps and the obtained products catalog are as follows：

Hope it helps you

[panjibrotoisworo]

The initial results are not so good when I try to use DEM-assisted co-registration. Perhaps because ICEYE is a high resolution SAR and the free DEMs are not so compatible due to wavelength/resolution differences. But I will keep trying with it.

Is it possible to declare this as a possible bug @mengdahl ? I’m able to run the coregistration process if it’s using the SNAP GUI as mentioned by Nijat but this means that I have no way of automating this process due to errors with the graph. I’m using SNAP Sentinel-1 Toolbox 9.0.2.

[panjibrotoisworo]

@Nijat Do you mind sharing the parameters you’ve used for the workflow? I just tried it with default parameters and it’s not so good.

When I try to process ICEYE, my coherence is only up to 0.4 and the coherence map is really unreadable. But I see your coherence map is really good.

EDIT: I got better results once I used the April 6 and April 7 pair. April 5 and April 6 don’t seem to work that well. Here are the results with completely default settings following the workflow you shared.

[Nijat]

Sure. the coherence between April 5 and April 6 pair is really not good.
The following are the registration parameters （April 6 and April 7 pair）

and the Multilook parameters:

the coherence map:

[ercad]

The coherence between 05.04 and 06.04. also is not so bad.
Try a larger fineRegistrationWindow. I think the initial
offset between master and slave, calculated based on the orbits,
is larger for this image pair.

my registration parameters:

the multilook parameters:

the coherence map:

statistics of the coherence map: