SARProz results analysis

This topic contains 1 reply, has 2 voices, and was last updated by  periz 2 years, 4 months ago.

  • Author
  • #702


    Hi,Professor Daniele

    When I finished the APS and SPP module operations,the result obtained in “Sparse geocoding” module are lots of points containing deformation information,including velocity,cumulative displacement ,etc. I’m not sure whether the velocity values and cumulative displacement values are absolute values or relative values with respect to reference point? How to understand these values?

    You must be logged in to view attached files.
  • #704


    a. before exporting data, I suggest you to use the “visualization tools” to explore the results. In particular, histograms and scatter plots are very useful tools. In the scatter plots, you can analyze the correlation between different parameters, you can plot them in 2D/3D in SAR coordinates or geographic coordinates, you can place many thresholds to extract sub-selection of points for investigating any kind of signals.
    b. after analyzing well your data and understanding their meaning, I suggest you to create an optimal selection of points discarding outliers. When this is done, you can export the optimal selection in Google Earth, so that you can easily visualize them on an optical layer.
    c. about your question on reference point: any parameter estimated through interferometry is by definition relative to a reference point. However, Sarproz tries to export the height using the GCP information you input at the beginning. So, the height of the reference point should not be zero. The cumulative displacement is the result of a double differencing operation. So, it is calculated with respect to a reference point, but it’s also the difference between last and first acquisitions. Both Velocity and Cumulative displacement should be zero for the ref. point (consider that, if you use a linear model, the cumulative displacement is directly linked to the velocity -no real additional information-)
    d. please consider evaluating all these infos using the Small Area Processing. It’s a very good training module for understanding well the estimated parameters

You must be logged in to reply to this topic.