sarproz-group

SBAS Analysis

Tagged: 

Viewing 3 reply threads
  • Author
    Posts
    • #5239
      gokhan
      Participant

      Hello to everybody,

      I want to process SBAS analysis with 20 pieces of CSK images. I just thought that multiple master option instead of generating a single one might be a better idea under this situation.

      I know there is no SBAS directly in SARPROZ but there should be a way to do that.
      So, what are the procedures to make SBAS? Which parameters/graphs do I need to choose?
      Could you explain me the steps in detail please?

      Thanks a lot in advance…

    • #5240
      periz
      Keymaster

      we never made such a tutorial for 2 reasons. firstly, because there is no “SBAS” button in Sarproz. secondly, because approaches that differ from the standard time series approach offered by Sarproz are specific to each case. So, if we would make a tutorial, it would be specific to that single case, not something to be replicated to each other case.
      Here, see a couple of similar questions/answers:
      https://www.sarproz.com/groups/sarproz-group/forum/topic/how-to-process-sbas-insar/
      https://www.sarproz.com/groups/sarproz-group/forum/topic/sbas-processing/
      Here, see a tutorial on a non-standard approach (it’s specific to the case there presented, but it shows you how to use Sarproz in a non-standard way):

      Download the PDF file .


      We will prepare more material in the next future (e.g. some glacier studies with unwrapped phase).
      With 20 CSK images, I do suggest you to do first a standard time series analysis. If the APS estimation is ok, in the MISP module you can do 2 things to improve the analysis and the spatial coverage:
      1. choose to process all pixels (not only local maxima as default in the initial mask generation)
      2. apply a “smart” analysis by writing the number “5” in the “smart” box, time series processing options in the MISP module.
      Best

    • #5241
      gokhan
      Participant

      Hi Daniele, thanks for your quick response as usual.
      I did standard time series analysis and radius was 12 Km for the interested area. Actually, APS was not solved perfectly. When I checked with test function, the trend of coherence graph was almost 0.5-0.6. They were not close to 1.
      What should I do to increase the quality of this step? I mean changing star graph to other one? or changing the parameters.
      After passing this step successfully, sure I ll try your recommendation.
      On the other hand, Applying a smart analysis generates cumulative displacement instead of linear one, right?

      Thank ypu so much again.

    • #5242
      periz
      Keymaster

      Hi Gokhan,
      I suggest we have a remote session. It’s difficult to give suggestions without seeing your data.
      Yes, the smart analysis (non-parametric temporal filter) is a non-linear analysis and it returns the time series plus the cumulative displacement.
      best

Viewing 3 reply threads
  • You must be logged in to reply to this topic.