Surface Subsidence

Dear ABraun

@ shishir.manandhar
But I think I figured this out. So basically I terrain corrected 1)Displacement 2)Unwrapped phase.
And then I collocated them (stacked), and then I applied valid pixel expression of coh>0.4 on the displacement. And I believe the result is good. What are your thoughts on this @ABraun? :slight_smile:

I followed above steps for 5 interferograms.

Here i given as shown in above figure.
Valid pixel validation : Unw_Phase_ifg_21Nov2017_27Dec2017_VV_slv1_21Nov2017>0.4

Is it correct for assigning the coherence value >0.4 ?

Then i created for all 5 interferograms in same way, finally stacked.
i want to know the displacement for that i added this displacements using Band Math, using this below command in edit expression.
($26.displacement_VV + $26.displacement_VV_slv2_27Dec2017 + $26.displacement_VV_slv4_20Jan2018 + $26.displacement_VV_slv6_25Feb2018 + $26.displacement_VV_slv8_21Mar2018)/5
but I’m getting this below error, please check it once.

masking by coherence is a good idea. Maybe you should include all displacements and coherence bands in one stack and then create an average coherence first and use this as a final mask (maybe with a lower threshold, because the average will always be lower).

Dear ABraun,
Here i stacked displacement and coherence and unwrap_* and intensity as shown in figure but i’m not getting where i wanna give coh>0.4.

you enter it in the valid pixel expression (remove what is already in there).
But I actually meant to have all coherence bands in one stack so you can calculate an average coherence as well which you can use for the masking of the average displacement.

Dear ABraun,
Here i added coh_IW3_VV_21Nov2017_27Dec2017_slv2_21Nov2017>0.35 in coherence band. as shown in below figure.

Is it correct?

No, you add coherence threshold in the valid pixel expression of the displacement to mask low coherence values.

1 Like

Thank you so much ABraun.

Is it correct?

Yes, like this. But this is only one image pair, right?

Thank you so much sir…
Yes sir, i will forward for all the 20 images stack.

I suggested to do the same with all coherence bands

Ok sir. thank you

I’m interested in the outcome :+1:

I applied coherence for all the displacement bands as shown in above, then stacked all the bands (1+2+…19). But creating the final displacement band in Band Math i’m getting below error.



If i open the final stacked, one displacement band it is giving this below error.

Dear ABraun,

I created combined stack of only displacement band (1+2+…19), using this created one average displacement band using Band Math(dis1+dis2+dis3+…dis19)/19.
And then same way i created combined stack of only coherence band (1+2+…19). using this created one average coherence band using Band Math (coh1+coh2+coh3+…coh19)/19.

Then finally i stacked both of this bands.
properties of the displacement band (new_band_14), placed coherence band value (new_band_13_slv57_21Nov2017) >0.35 as shown in below figure.

final output

Is it correct ?

looks correct - good job!

Thank you ABraun.

Phase to Displacement
The unwrapped phase is now a continuous raster but not yet a metric measure. To convert the radian units into absolute displacements, the Phase to Displacement operator (under Radar > Interferometric > Products) is applied. It translates the phase into surface changes along the line-of-sight (LOS) in meters. The LOS is the line between the sensor and a pixel. Accordingly, positive values mean uplift and negative values mean subsidence of the surface (if the master image is the earlier acquisition date).
Copied from
InSAR Displacement Mapping
with ERS data
Issued April 2020
Andreas Braun.

Please clarify this one.
Because if LOS is increase (means sensor to ground pixel distance increases (+ indication)) I’m getting subsidence (- indication) and vice versa.

Is it correct?

this depends on the order of yourr images. Which one is the master and which one is the slave?

For suppose master is 2014 January and slave is 2020 August. Then how can we define LOS (or upliftment and subsidence)?