Subsidence map in 3d view

is there any formula exist for converting the unwarpped phase into metric unit… and how we create a 3D map of subsidence…

someone please help me…


the unwrapped phase is the change of surface elevation in the looking direction of the SAR wave.

Radar > Interferometric > Products > Phase To Displacement
This should actually convert it into measurable units.

However, I wasn’t lucky with it so far and used another approach in band maths:

  1. (Unw_Phase * wavelength in mm) / (-4 * PI * cos(rad(incident_angle)))
    This gives you the relative change between two satellite images in milimeters
    Source:, chapter 2.3.

  2. Crop out areas with a low coherence (Layer properties > valid pixel expression, for example coh_HH_11Feb2008_04Mar2008 > 0.6).

  3. Select a point where you assume no change and read the value. Let’s say its 45 cm

  4. Subtract 45 from your image in order to have zero values at those areas with no change. The remaining variation should be due to subsidence between two overpasses.


Hi ABraun

This steps are the same the methodology of Surface subsidence monitoring with NEST?

I have question: the criterion for to determine point value reference is an point that not occurred change? you are sure?

I am applying this methodology

Thanks in advance

1 Like

yes, these are the steps from the SAR EDU page.
Edit: Now filed under Echoes in Space: >

A point with a low probability of change should be taken as reference height.

Good luck! Let us know if it works.

1 Like

Ok, thanks ABraun!!!

hello ABraun
the first method was not worked for me either… when i did the band math with the above mentioned formula a new band appeared which look promising , but when i moving the mouse cursor over this newly created band image it did’t show the latitude and longitude value…

what will i do…

you can see the band values in the pixel info tab. If it is closed you can re-open it in the view menu. Note that the lat/lon values are not of importance but the pixel values of the band. Please tell us what you found.
You can also post screenshots in here by Ctrl+V or the upload dialogue.

yes ABraun i can see the band value… what does this value mean ? is it the subsidence or LOS displacement in millimeter or some metric unit…
one thing what does the minus and plus value means is it the the subsidence to the west or east depending on the satellite orbit i.e descending or ascending orbit…

if you entered the wavelength in mm those are values of LOS displacement in mm. You should now mask the areas with low coherence so only “trustful” values remain. Then you can select a reference point.

hello ,
i could’t find the layer properties option…

right-click on the unwrapped phase, in the menu you find properties. There is an entry “valid pixel expression”, where you enter something like coh_HH_11Feb2008_04Mar2008 > 0.6.
Of course, the coherence name should be yours and the threshold can also be lower. Some say acceptable quality begins with 0.3 already.

1 Like

what does the minus and plus band value means is it the the subsidence to the west or east depending on the satellite orbit i.e descending or ascending orbit…

in your original image, the band after the applied formula of above, they are the relative changes of each pixel in look direction of the sensor between both images. After applying the reference height (setting no-change areas to zero) they are supposed to be the absolute change. Here, negative values imply subsidence, positive stand for uplift.


Excellent!!! all comments, help me to respond my doubts!


hello ABraun
when I enter the coherence file name it shows the error like… “expression is invalid . undefined symbol”… what is it

in band properties Manu unit of Raster band properties show null … what should I enter here…

you also need to enter the threshold value for the coherence in the valid pixel expression:
coherence... > 0.4 for example

hello ABraun when I enter the coherence file name it shows the error like… “expression is invalid . undefined symbol”… what is it

I explained it in the post above. Just entering the coherence is not a valid logical expression. In this field (valid pixel expression) you define at which locations your result is usable. As coherence is a measure for quality of the interferometric product you only want those pixels at a location where the coherence is larger than a threshold, let’s say 0.4.
Thus, just putting the coherence there is pointless because the expression is lacking a condition. Enter this instead:
coh_23Nov2012_08Feb2013 > 0.4
Then you will have bad quality pixels masked out and only the “good” ones remain.
Depending on your colour scheme it should somehow look like this:

As your temporal baseline is quite long (about 75 days between first and second image) coherence values could be quite small.

Hi @ABraun,

this post and your comments are a very good walk through, thank you all! Is the following quote your experience or are can you recommend literature?