Forums » GAMMA Processing » SAR Data Preprocessing » Sentinel-1 »
Nan gaps between adjacent images when RTC-correcting from SLC using multi_look_ScanSAR
Added by Miguel González 2 months ago
Hello everybody,
I am RTC-correcting SLC images using GAMMA, but output tif files of adjacent images does not spatially match, existing nan values in between.
The workflow I am using is the following:
1. Open each SLC's sub-swath .tif file using par_S1_SLC .
2. Join all sub-swaths' burst using multi_look_ScanSAR
3. Apply bunch of tools for RTC-correcting.
For GRD images, I fixed this just by indicating e_flag to 1 (permit pixels without the full number of looks) when applying multilook, not to loose pixels close to nans. Nevertheless, in multi_look_ScanSAR using SLC this kwarg is not available, so I guess this is the reason that certain pixels are loss when multilooking and hence nan gaps are present in output files.
I wanted to ask then if this is the real reason of existing nan gaps between adjacent RTC-corrected images. Maybe I can pass this e_flag keywarg somehow to multi_look_ScanSAR ?
Thanks a lot in advance.
Regards,
Miguel
| Sin título.png (155 KB) Sin título.png | Nan gaps between images |
Replies (2)
RE: Nan gaps between adjacent images when RTC-correcting from SLC using multi_look_ScanSAR - Added by Redmine Admin 2 months ago
Dear Miguel,
It looks like you need to concatenate the two adjacent frames.
In general, the quality of S1 data is quite stable. Occasionally the raw data can be mismatched and a single burst may be missing, but that does not appear to be the issue here.
From your screenshot, I think the two frames were processed separately. Regarding the misalignment, please check the geocoding step: did you use the same kind DEM product and update the RDC height file with a fine lookup table? If the SAR images are well aligned with the DEM, this kind of dislocation should not occur.
If you don’t need to concatenate the entire orbit, I suggest using SLC_cat_all to mosaic the adjacent frames in radar coordinates. This script ensures precise co-registration in the overlap area and applies amplitude correction.
Best regards,
Sen
RE: Nan gaps between adjacent images when RTC-correcting from SLC using multi_look_ScanSAR - Added by Miguel González 2 months ago
Hi Sen,
Thank you very much for your prompt response.
You are right, I processed both adjacent images on a separate way, using for each a LUT refinement process and the same 30-meters DEM.
I think I am not getting what you're saying...
Firstly, I don't understand the theoretical background of co-registering adjacent images. Is it only necessary in azimuth direction? I do co-register images, but only for scenes sharing the same footprint. This way, they can share LUT and RTC process is more efficient. I have never heard of this for adjacent images, where images acquisition geometry will be always different.
I would really appreciate if you could shed some light on this.
A the beginning I would not like to take into account nearby images, since I am automatizing RTC workflow and unknown if user will provide adjacent images.
Thank you very much in advance.
Regards,
Miguel
PD. A more panoramic view of the two adjacent images is attached to ease understanding (red box represents former image footprint).
Also, since I am using GAMMA version of 2022, only SLC_cat and SLC_cat_ScanSAR are available for me.
