processing multilooked fringe

[bbuz...@odu.edu]

Hi Yunjun,

I've got a stack of full resolution interferograms produced with FRInGE, which I then multilooked and unwrapped.

prep_fringe.py is giving me problems because of the sizes of my custom multilooked vs what it's reading from the geometry files (which are full resolution).

I've tried a couple things, like hardcoding the pix_box to the correct size, and multilooking the geometry files. Hardcoding pix_box gets things to run but there are other metadata that I think are incorrect and mess up the results later on.

Any tips on how to proceed? There are a lot of places metadata is getting updated and narrowing my search would be much appreciated.

Thanks,

Brett

[Zhang Yunjun]

Hi Brett,

Below are a couple of things I could think of on top of my head:

1. the geometry files need to be multilooked into the same size as the interferograms as well. looks.py from ISCE-2 could do that, which is equivalent to multilook.py -m average, but you may want to use multilook.py -m nearest for a more meaningful geometry operation.

2. prep_fringe.py is hardwired to use the full resolution geometry files (https://github.com/insarlab/MintPy/blob/7b81971823e2c19e165395c9a4263995dc4fd8af/mintpy/prep_fringe.py#L140), you could change that for a quick solution, or add an `--geom-ext` option to expose this in the command line. The few lines below are the places that exact/update the size info, you could debug there if error occurs.

3. Besides the unwrapped phase, the inputs to the "--coh-file" and "--ps-mask" options also needs to be masked.

I hope this helps,

Yunjun

[bbuz...@odu.edu]

It does indeed, thanks a lot.