Hello. I am looking at a surface deformation in Indonesia, where the the deformation is mapped over 10 cm/year using Sentinel-1 data. The dates of the two observations was Late 2014 and spring 2017. With that in mind, I have a couple of question.
I believed you would want each image to be taken within 6 cm apart for c-band, since above 6 cm you would not get good coherence including urban area since deformation unwrapping is done on a circle. Is this true?
If it’s not true that you can take imagery as far apart in urban areas to obtain a general idea of deformation, are there any threshold limits? If it is true, are there any boundary conditions that I should be aware of?
Is there a reasonable way to monitor the deformation without using PSISAR.
Does anyone have any links or guidelines of do’s and don’ts to follow. Thanks.
Do you refer to an existing study here? Maybe you can share the source with us.
For an investigated period of this length, classic interferometry (InSAR) is hardly applicable because of temporal decorrelation (except for urban areas).
It is true that the detection of displacement is limited by the size of the wavelength. If you are working on an urban area, I would at least consider using several image pairs to reduce the temporal baseline and the effect of atmospheric effects in single pairs).
In fact, PS InSAR is more reliable, but requires a lot more images (e.g. a monthly base for the entire period), and time to get into. The preprocessing is currently not implemented in free software with a simple user interface. We collected instructions in this here: About the StaMPS category
It took me half a year to get to my first result, to be honest, because I was not used to using Linux and Matlab (required for the processing) using the required scripts. Even after that, I am quite insecure about the many different parameters ad outcomes.
Thanks for the update and filling a lot of holes in my knowledge. Let me see if the report is posted online and I will happy to send the link to you. Thanks.
Sorry for the late reply, but the report has not been posted anywhere. Unfortunately, I will not be able to share it with you.
@ABraun What is the range of the limitations of displacement due to size of wavelength when you say this?
For sentinel-1, wavelength of 5.6cm, does this mean with DInSAR, it can only detect displacement of less than 2.8cm in a single pixel. Therefore subsidence/uplift greater than 2.8cm are ambigous and are due to atmospheric or noise errors as it is impossible to retrieve?