I’ll appreciate you if answer my question:
Which parameters should i choose in “small area processing > multi
image analysis” module where is located in “reference point” section?
And why the results varies with choosing different parameters of
You can choose any parameter for your reference selection: temporal or spatial coherence, amplitude stability index, reflectivity map or whatever.
Or you can even pick it up manually (ticking the corresponding box) depending on other infos/observations (e.g. you want it on the ground, in a stable area or whatever).
InSAR is a differential analysis, so, the results are depending on a reference point. Each point can be affected by different phenomena and/or errors. So, if you change the reference point the final result may change.
In stable conditions, changes should be negligible. In complex situations, you can have big changes. E.g. if in your area you have high buildings, choosing the reference point on the ground or on the top of a building will drastically change your results.
Hi Dr. Perissin
Thanks so much for your response.
I have read your papers entitled as “High-Accuracy Urban DEM Using Permanent Scatterers” and “Validation of the Submetric Accuracy of Vertical Positioning of PSs in C-Band”. A few questions have come to me, I’ll appreciate if you answer.
Does “residual height” calculated only by subtracting the reference point height from the target point height?
Is “Height St Dev” an indication of the accuracy of the estimated height? In this case, we know that the theoretical accuracy of 3D positioning by means of the PSI technique is under 1 meter [submetric accuracy]. So why the “Height St Dev” values in the output file are in the range of several meters!!?
InSAR is by definition relative to a reference point. So, everything is calculated by subtracting the value corresponding to the reference point.
Residual height is the height relative to a reference DEM. Height = external_DEM_height + residual_height.
The deviation of the height estimation depends on many things, as number of images and baselines dispersion (as you can read in the papers). That could be the reason why you get several meters of height st dev.