Using sentinel data to create phase interferograms for my study site near Los Angeles, I am regularly running into problems where portions of the phase are ommitted in a type of regular banded formation (attached). I am wondering if this has been observed and what to do about it - the graph used to create images is attached and default parameters were used. I am wondering if iut has something to do with the co-registration being affected near the ocean.
I haven’t seen this error, but I suspect that the graph is too big and some of the data gets lost in the cache.
Maybe you can also try a different DEM for the Terrain Correction step as it is used for masking sea areas.