Coherence seams along whole-degree meridians in NISAR GSLC interferograms over Alaska

[chowoon97]

I found narrow coherence-loss seams along whole-degree meridians in interferograms generated from NISAR L-band GSLCs over Alaska using the polar stereographic projection (EPSG:3413).
For one GSLC pair (Track 121 / Frame 33, 2026-07-04–2026-07-16), the seams coincide with approximately −147°, −148°, −149°, and −150° longitude. Mean coherence measured along these lines is about 33–43% lower than in nearby areas, and localized wrapped-phase discontinuities are also visible.

As a control, I independently processed the corresponding RSLCs, and these meridian-aligned seams do not appear in the resulting interferogram. I also have not observed similar seams in GSLC interferograms from other UTM zones.

Because the artifacts are aligned with whole-degree meridians and appear in the polar-stereographic GSLC products but not in my independently processed RSLC interferogram, I wonder whether they could be related to DEM tiling/stitching, reprojection, or another DEM-related step during GSLC geocoding.

Higher resolution of PDF can be found in https://smu.box.com/s/y403hyiw4nk4kqukwtvfjuo7p39ukvtt

[NISAR - hfattahi]

Thank you for reporting this edge case. We were able to reproduce the issue. The source of the issue is the DEM at higher latitudes which seems to have a small discontinuity between tiles. We are investigating a possible solution and will update when we have more information.

[chowoon97]

Thanks for reproducing the issue and identifying the source. Hopefully, it will be relatively straightforward to address.