sarproz-group

sentinel 1, problems with geocoding timeseries

Viewing 1 reply thread
  • Author
    Posts
    • #5854
      yvhout
      Participant

        Hi,

        I am processing around 60 images of sentinel 1 data for a mountainous area (a radius of 7 km) containing glaciers and lakes. I checked the boxes for using ext. DEM at coregistration parameters. I came across some problems, I did not come across solutions on the forum.

        1). I thought I should not estimate linear trend (as I am not expecting linear deformation) and neglect it (and fill in 5 for smart). However when I do that at the sparse points processing part, everything first seems to work well, but I do not obtain a velocity matrix in the end, which prevents me from geocoding time series and producing scatter plots (if I then change linear trend to estimate at sparse points processing, it does work). So my question is, is it not possible to combine this and thus I will have to use linear trend estimation?
        2). When using the Time Series module – time series space visualization and export, I get very weird deformation results which make no sense. They are not like the output from sparse TS Geocoding. I also (both sparse TS geocoding and visualization and export) do not get the graphs when you click on the points. I can only see the colour and the corresponding legend.

        3). Furthermore, do you always have to click on ‘External DEM and synthetic amplitude in SAR coordinates’ at site processing, even if GCP selection is already OK after generated automatically with keep current orbits? (In my case it only works when I do click on it)
        4). Lastly, if I fill in the threshold at sparse points selection (APS Processing) for amp. Stab. Index to be 0.8, I only get PSC Nr. 25, so I lowered this value to get more points (not sure if that is necessary?). My coherence seems to be fine at the sparse points processing so not sure why I get so little points at the APS processing.

        Thanks in advance!

      • #5858
        periz
        Keymaster

          hi,
          here some answers to your points
          1) if you use the non-parametric approach (e.g. smart=5), the software does not estimate a linear trend. what you can export instead is the cumulative displacement
          2) sorry, don’t understand your question
          3) ‘External DEM and synthetic amplitude in SAR coordinates’ is a function to generate those products. you don’t need to do it unless you need those products
          4) this question is not clear either. anyway you need a decent amount of points to process the aps.

          in any case, I would strongly suggest you to take a training course. if you are interested you can write to
          best

      Viewing 1 reply thread
      • You must be logged in to reply to this topic.