Dear Maryse,

Sometimes even with precise orbital information, residual orbital fringes can remain. It sounds like that may be what is happening for you. Ideally, the Refinement and Reflattening step will remove orbital fringes. Perhaps the GCPs that you provided for that step were not located on orbital fringes, and so those fringes were not removed well?

There is a document on the Sarmap web site (developers of SARscape) about how to choose GCPs for different purposes. It might help to read the section there about GCPs for persistent scattering:

http://sarmap.ch/tutorials/GCP_Manual.pdf

Best,

Peg Shippert
Harris Geospatial

Maryse,

I was also curious that you mentioned QGIS. Did you also view the results in ENVI, and did you see the same fringes there? Or do you think this is perhaps something specific to the way QGIS displays the data?

Best,

Peg

Dear Peg,

thanks for providing me with the GCP manual. I didn't use any GCPs up to now, to be honest. I thought these would only be mandatory in the "Refinement and Reflattening" step in the SBAS processing, but would be optionally in the PS processing. Maybe, I'll try to include GCPs also in the PS processing.

Regarding the visualisation: Yes, I also checked the results in ENVI. Both files PS_geo_disp_75_meta and PS_geo_disp_VD_75_meta don't show any ramps. As I wanted to visualise the PS results with other geo data in a GIS, I merged all the _VD.shp files in QGIS and chose the attribute "velocity" for the visualisation. After seeing the ramp, I also implemented a function in Matlab to calculate the linear trends of the PS points out of the single displacements given for each date (also provided in the attribute table of the _VD.shp files). I got the same result. So, obviously, the results of PS_geo_disp_VD_75_meta and the ones of the merged _VD,shp files are not the same. Do you know, how the PS_geo_disp_VD_75_meta file is generated? Was it wrong to assume that the PS_geo_disp_VD_75_meta is in fact created by SARscape by merging the _VD,shp files?

Thank you for your help,
Maryse

Dear Peg,

sorry, I think I was looking at the wrong files in ENVI. I now opened the file PS_75_meta, saved the velocity band in an extra file and visualised it in ENVI. Here, I also see the ramp in the data. So, I think I will follow your recommendation and introduce GCPs in the geocoding step of the PS processing. Hopefully, this will reduce the residual orbit errors.

All the best,
Maryse