S1 swath discontinuities

I’ve got some interesting coherence results which I haven’t seen before with S1 data. There appears to be a discontinuity between swaths but not bursts. The overall coherence is isn’t high for my dataset due to regular land changes (eg. rainfall, agricultural), but it’s appears a lot of my ifgs to varying degrees.

It doesn’t appear to propagate to the filtered or unwrapped ifgs. It also doesn’t appear to be related to temporal threshold as I’ve got 12, 24 and 36 day ifgs, and it can appear in all three thresholds.

Has anyone else seen this issue?

No errors (20150716-20150728):

Extreme case (20160205-20160217):

Subtle case (20160229-20160324):