Open mohseniaref opened 1 year ago
Hi @mohseniaref I agree that it is computationally expensive, but I do not suggest it. The improved multilooking in phase linking by selecting SHPs, is to avoid mixing different scatterers and this is one of the main differences from SBAS. If you use multilooked SLCs, then the covariance matrix is even more biased and you are actually performing double multilook (with miaplpy). You mentioned you are working with full SLCs, do you mean you are processing the full frame (large scale)? then If you don't care about a bias of ~2mm/y in case you have fading signal, and you are looking at large scale long wavelength displacement, you better use SBAS. Phase linking with current implementation is more suitable for small areas.
Hi @mirzaees
Many many thanks for your response. I am working on landslide detection on large scale and see how much fading signal could impact landslide detection. So do you think is there any solution ?
Hi @mohseniaref, if you have access to high performance computers, you can go ahead and process with multiprocessing. But I would suggest to use SBAS for large scale and landslide detection, then process your false alarms locally with phase linking to see the impact.
@mirzaees Thanks so much for your response. I will go with multiprocessing on server.
@mirzaees I emailed you. Have you recieved my message? Best regards,MMA
Hi @mohseniaref, Sorry I didn't receive an email
Hi Sara,
I am working with full SLCs for InSAR time series. I think it could be computationally expensive for non linear phase inversion. Can I use multilooked SLCs instead?
Best regards, Mohammad