Sentinel-1 Burst Overlap Double Difference Interferogram

Dear all,

Digging into SNAP functionalities, I found an operator that retained my attention : S-1 Double Difference Interferogram with the description :

The operator computes the double difference interferogram for co-registered Sentinel-1 TOPS product for overlapped area between adjacent bursts.

To be sure : does-it mean that we substract from the interferogram built with burst i the interferogram with burst i+1?
(EDIT : I found my answer here : https://github.com/senbox-org/s1tbx/blob/master/s1tbx-op-sentinel1/src/main/java/org/esa/s1tbx/sentinel1/gpf/DoubleDifferenceInterferogramOp.java)

Related code
final double[][] mIBack = getSourceData(mstBandI, overlapInBurstTwoRectangle);
final double[][] mQBack = getSourceData(mstBandQ, overlapInBurstTwoRectangle);
final double[][] sIBack = getSourceData(slvBandI, overlapInBurstTwoRectangle);
final double[][] sQBack = getSourceData(slvBandQ, overlapInBurstTwoRectangle);

final double[][] mIFor = getSourceData(mstBandI, overlapInBurstOneRectangle);
final double[][] mQFor = getSourceData(mstBandQ, overlapInBurstOneRectangle);
final double[][] sIFor = getSourceData(slvBandI, overlapInBurstOneRectangle);
final double[][] sQFor = getSourceData(slvBandQ, overlapInBurstOneRectangle);

final double[][] backIntReal = new double[h][w];
final double[][] backIntImag = new double[h][w];
complexArrayMultiplication(mIBack, mQBack, sIBack, sQBack, backIntReal, backIntImag);

final double[][] forIntReal = new double[h][w];
final double[][] forIntImag = new double[h][w];
complexArrayMultiplication(mIFor, mQFor, sIFor, sQFor, forIntReal, forIntImag);

final double[][] diffIntReal = new double[h][w];
final double[][] diffIntImag = new double[h][w];
complexArrayMultiplication(forIntReal, forIntImag, backIntReal, backIntImag, diffIntReal, diffIntImag);

final double[][] ddiPhase = new double[h][w];
for (int i = 0; i < h; ++i) {
    for (int j = 0; j < w; ++j) {
        ddiPhase[i][j] = Math.atan2(diffIntImag[i][j], diffIntReal[i][j]);
    }
}

return ddiPhase;[/details]

If yes, what is the point of doing this ? I was thinking about phase closure but it needs 3 interferograms.

Thanks in advance,

QG

did you try it in the meanwhile and find out what it’s good for? :slight_smile:

Am also still searching for good examples using the the ThreePass Interferometry module.

Yeah I’ve more than tried.

I imagine that you went back to this subject because you read the articles I gave to you (Multi aperture interferometry) :wink:

Results are encouraging and the method I’m developing in my thesis partly use this. Double difference interferometry is a kind of MAI for lvl1 S1 SLC data only on burst superposition areas. The goal is thus to infer azimuthal displacements. I created the subject because I was very surprised SNAP implemented it, because almost nobody uses this technique. What do the developers had in mind when creating this?

I don’t want to share now my results but as we are almost neighbors, we will probably have opportunities to meet and talk about that in the future (EUSAR, IGARSS, EGU, AGU, … ) :slight_smile:

1 Like

thank you for your response. That sounds very interesting - would be nice to meet you at a conference and talk about it. Unfortunately, I won’t be attending any of the larger ones in the next months. But let’s keep this idea up.

1 Like