I did a test converting the phase to displacement (sentinel 1) both manually (band maths) and using the phase to displacement operator.
The automatic results are exactly one order smaller than the ones obtained using (unwr * 0.056/4*PI). Knowing the sites, it seems that the manual conversion seems more realistic (min-max +/- 30 cm, while automatic min-max +/- 3 cm…over I mine site that I know is quite active.
Version: SNAP 5 and beta version 6.
Did anyone experience that? Which one is correct?
By converting the unwrapped phase into displacement either using the equation or the implemented method in SNAP, it should give you exactly the same answer.
By looking at the equation you posted, it seems that some parenthesis are missing so that you do not get the correct results. Hence, the equation should look like (unwr * 0.056 / (4*PI) ) and not (unwr * 0.056/4*PI).
We can test both equations by using a random value for the unwrapped phase and see what we get.
If we plug a random value (lets say 100) for unwrapped phase to the equation that parenthesis are missing, we get the following. (100 * 0.056 / 4 * PI) = 4.3 but if we use the same value to the other equation with the parenthesis placed in place we have, (100 * 0.056 / (4 * PI) ) = 0.43.
As we can see, if we missed the parenthesis, then the values of our results are larger.
By the way, you can add a minus sign in the second part for the equation (-4*PI) so that values with minus sign indicate subsidence.
Since you also mentioned the sign correction:
I guess the resulting sign of an interferogram should be dependent on which master is selected (older versus latest date). Does SNAP apply a correction, in order to have always negative values for deformation away from the satellite?
Using SNAP for phase to displacement conversion , we get minus sign where displacement occurred and plus sign where there is an uplift.
Using only two SAR images to derive a map subsidence is not recommended as it suffers from errors due to the contribution of the atmosphere. If you want to minimise this error, you can create a stack of multiple SAR images and then you can derive the map subsidence.
I have found subsidence using two images in year 2016 and two in 2017. I have placed 24 pins on the displacement map of the year 2016 and exported the values in a table having following coulmns Sr No. Pin Name X Y Longitude Latitude Coherence (coh) Displacement (meter) with coh >= 0.4
Now i wish to show the displacement at the corresponding points with identical X, Y, lat, long’s for the year 2017. so that displacement value at every pin position can be compared year wise.
Please guide how to import or export values at same pin positions in the two maps?
Thanks i got it.
One more query;
I have taken S-1A data for one year and S-1B data for another year with 44.88 and 44.35 perpendicular base line and 96 days temporal difference.
Would it effect the results?
my results seems twisted clockwise. why?
Sentinel-1A and 1B are fully compatible for InSAR, you can combine them. What is important is which date is selected as master and which as slave as this decides the direction of the movement (towards or away from the sensor) .
Makes sense, but specifically in which direction does this shift occur based on mst/slv date?
Example An ifg was calculated with two images in reverse date order:
September2017 Master and August2017 Slave results produce positive phase to displacement values and positive vertical displacement (heaving) via band maths calculations. Correct?
Or should these values be multiplied by -1 to indicate negative vertical displacement (subsidence) due to the temporal reversal of the images?
In searching the forums I found several seemingly conflicting or vague responses on this matter.
I have created a stack of 12 interferograms (split_orb_coregi_ifg_deb_dinsar_ml_flt_avg). Now i wish to export it to Snaphu, it gives an error saying product should be co registered. when i export (split_orb_coregi_ifg_deb_dinsar_ml_flt) to snaphu all products are exported. Now how can i run phase unwrapping of all products?
I have to go for phase to displacement in the end.
sorry, I don’t understand: Did the export work or not?
As far as I know, snaphu only works on image pairs (not stacks).
So the best way to process Sentinel-1 images is to create pairs by TOPS coregistration, create the interferogram, deburst it and export it to snaphu.
Then process each interferogram individually as shown here: https://www.youtube.com/watch?v=sH4Uhsgmpak
Then import the unwrapped phase (select the hdr file) back into SNAP. You can perform “phase to displacement” on this product then and finally terrain correct it.