Range and azimuth shifts to remove burst offsets during coregistration

Microwave Toolbox
#1

Hi,

I am using the most recent version of the Sentinel-1 toolbox and am processing Sentinel-1A interferograms by following exactly the steps in the manual http://sentinel1.s3.amazonaws.com/docs/S1TBX%20TOPSAR%20Interferometry%20with%20Sentinel-1%20Tutorial.pdf

So far I have processed 3 or 4 interferograms, which have worked fine. However, for these scenes:

S1A_IW_SLC__1SSV_20151201T105917_20151201T105941_008847_00CA32_9DEF
S1A_IW_SLC__1SSV_20151225T105916_20151225T105940_009197_00D407_D4AC

there are offsets between the bursts that I am having trouble removing (see figure). Looking through the manual and the forum I can’t really find a clear explanation of how to remove these.

I have coregistered the two scenes using the “S1 TOPS coregistration” option. I have then selected to use the “S-1 Range Shift” and “S-1 Azimuth Shift” steps before resuming the steps in the manual (interferogram formation, deburst etc). Is this the correct place to do these steps? I have found that using them is making no difference to my results.

Any advice would be appreciated.
Thanks,
Amy

Screen Shot 2016-01-19 at 13.41.58.png713×701 598 KB

#2

The range and azimuth shifts are estimated and removed using co-registered pairs as inputs (back-geocoding output) and before the generation of the interferogram. The precision of the orbits contribute significantly in the successful co-registration and removal of jumps. Please consider updating orbital information by “applying orbits” (Restituted or Precise) (https://qc.sentinel1.eo.esa.int/), The software downloads them automatically for you. Please note that precise orbits are only available after ~20days from sensing.

#3

Thanks for the reply. I have used the most recent precise orbits downloaded from ESA, but this has not fixed the problem. The size of the swath for these most recent acquisitions is actually smaller than for previous acquisitions - perhaps this has something to do with it. Is there anything else I can try? Will the orbits be updated again in the future or is this the final estimated orbit?