Dear falahfakhri, Thank you so much. I look at LOS to horizontal displacement and to my understanding, it is differential interferometric SAR (DinSAR) but I think you know that I am working on fast ice, then I do not have any DEM on sea ice (as ice is on sea) so I can not use DinSAR. Am I right?
Sorry for delay answer, But first of all the title of this topic is the interpreting the results … So any results of InSAR despite of the technique is one of two things the first one is the displacement and the second one is the height or DEM.
The second important thing in your case that you are trying out to measure the horizontal movement of ice, I’m not sure about the workflow of your processing and also I think you are using GRD or are using SLC?
But in case you are using SLC and you are trying out to measure the horizontal displacement for ice so in this case there is solution. I’m working in now.
Hi all,
Here are two interferograms obtained from two Sentinel 1B SLC products. The two points within the rectangle represent locations of two earthquakes of 5.1 mag. The polyline feature added is a valley that may be along some fault. The first interferogram was obtained after removal of Flat Earth Phase and the Topographic Phase, while the second one was made using only Flat Earth Phase removal.
***This region lies in Central Himalayan terrain.
The phase difference can have contributions from five different sources:
Δϕ flat is called flat Earth phase which is the phase contribution due to the earth
curvature.
Δϕ elevation is the topographic contribution to the interferometric phase.
Δϕ displacement is the surface deformation contribution to the interferometric phase.
Δϕ atmosphere is the atmospheric contribution to the interferometric phase.
Δϕ noise is the phase noise introduced by temporal change of the scatterers, different
look angle, and volume scattering.
Baseline-length affects the density of Fringes, and topographic phase removal should remove almost all topography-related fringes. Something seems to be off in your processing.
Also your temporal baseline is long, you would get better results by taking the shortest temporal baseline possible around the earthquake of interest (should be 12 days over the Himalayas).
I am not sure if I understood you properly. Lets say there’s an earthquake occurred in 10/6/2016 and I shall create an interferogram using image pairs before and after the disaster.
which two image dates should I choose
7/6/2016 - 12/6/2016 or
31/5/2016 - 18/6/2016 or etc.
Should I choose the image pairs with minimal temporal acquisition separation?
How would I know if the topographic phase removal was successful? I have applied it during the interferogram creation step and I still feel like the topographic effect still persists
you can output the topographic phase by the TPR operator and see if there is a difference.
Also I recommend using the DEM of highest resolution (here SRTM 1Sec AutoDownload)
Hello everyone, I would like to share with you some testing results of the following workflow on sentinel 1 imagery (first image 19/01/2021, second image 31/01/2021):
Orb, Corregistration with ESD, Interferogram formation
with subtract topographic phase, Golstein phase filter, Deburst, Multilook, Snaphu export, Snaphu unwrapping, Snaphu import, Phase to displacement, finally the displacements shown are only for Coherence > 0.6.
From the results is possible to see that all displacements are between -1 and 1 centimeter, however I would expect not displacement in urban areas, at leats not for the same day in different places around the world. Should I assume that there is an associated error of +/- 1 cm for the DINSAR results? In that case, how could I perform a time series analysis and don’t confuse any real displacement with the associated error? In advance thanks for your reply.