Hi, I want to measure distribution of movement of the large area at once with DInSAR using S1 TOPSAR products, and I am wondering if you think Ive made a mistake somewhere, because the resulting displacements are only negative now. Ive done processing of a smaller area before and the results were both positive and negative at the beginnig. So this is my processing chain:
-
Clip DEM only to the area of interest to trick the program and not process the rest.
-
Slice Assembly of two products (one product does not cover the whole area).
-
Subswath 1: TOPSAR split (bursts 1 to 20)-> apply orbit file ->Back-geocoding(using the clipped DEM) -> Interferogram->TOPSAR deburst
-
Subswath 2: TOPSAR split (bursts 1 to 20)-> apply orbit file ->Back-geocoding(using the clipped DEM) -> Interferogram->TOPSAR deburst
-
Subswath 3: TOPSAR split (bursts 1 to 20)-> apply orbit file ->Back-geocoding(using the clipped DEM) -> Interferogram->TOPSAR deburst
-
TOPSAR Merge of subswaths 1-3 -> TopoPhaseRemoval -> Goldstein Phase Filtering -> Multilook (4 range looks)
-
Use band maths to remove areas with coherence below 0.2 from the phase information-> merge old coherence with new phase band
-
Export to Snaphu (DEFO, MCF, 2 tiles, 16 processors, 0 overlap, 500 tile Threshold) -> Snaphu import -> phase to displacement -> Terrain correction.
What I am uncertain about are whether the slice assembly can be used here and whether the subswaths can be merged. Does it influence the results? Can I remove the phase with the low coherence before unwrapping in snaphu or should I do this in a different way? I am also uncertain how the number of tiles influence the results.