Export N-E-Up Components

I would like to ask if it is possile to export N-S-Up Components (not the value at Line of Sight (LOS))from a wrapped Interferogram using SNAP.

Thank you

A single SAR interferogram only measures the displacement in the line-of-sight (LOS) direction. It is not possible to calculate the east-north-up components from a single interferogram with SNAP or any other software.

Thank you for the information.
Also, I would like to ask you if it is possible to export east-north-up components if I have two wrapped interferograms for a seismic event (ascending and descending) using SNAP??

If you have ascending and descending interferograms, then you can calculate two components of the ground displacement. The LOS directions of ascending and descending Sentinel-1 orbits make it possible to estimate the East component and a mixture of the Up component and the North component.

You will need to do the phase unwrapping and conversion to displacement first. I don’t think there is any function in SNAP to do the complete calculation of the components, but you can get an approximate calculation with the raster calculator. The approximate value of the East component is proportional to result of taking the difference of the Descending-Ascending displacements. The sum of the Descending and Ascending displacements will be proportional to the vertical component plus about 10% of the north component. Doing the simple addition and subtraction operations can give you a qualitative look at two components.


Hi dear, I am radar(ps_InSaR) researcher. I am working on a project that I must retrieve 3D deformation map in the region. I get sentinel scending and descending images for interested region. But now I am confused that how I can extract 3D deformation from this ascending and descening data sets. Please guide me.
Thanks regard​:pray::pray::pray:

You might want have a look at these topics first, they discuss this issue:

1 Like

Hi dear, thanks for your answer. I saw that solutions, but I want to extract 3d def using ascen and descen images. And the exist solution for this method has presented very general and in a complex way.

I understand what you want to do but maybe there is no easy solution to this problem? :slight_smile:

The most straightforward and elegant approach I have seen so far is presented in this study:

Please have a look at formula 1 and 2.

Thanks regards. :pray::pray::pray::pray:. That’s nice of you

How we can unwrapp a stack of interferograms!! That’s impossible! Do you mean that at first we seperatly do it for a single pair of ifgs, and then average from unwrapped ifgs?

yes, unwrapping is performed outside SNAP (by calling snaphu.exe) and cannot be applied to a whole stack.
But I think simply averaging all the displacement rasters will not give you what you need.

I mean that by considering EJFielding method, It maybe possible to generate ascending and descending averaged unwrapped ifgs for calculating 3D deformation based on suggested formulas?!!

as far as I understood his post, he refers to the combination of only one ascending and one descending interferogram (of similar date). This is why I responded to your question about an entire stack of interferograms.
But maybe he can clarify.

that is right. can i ask you, what is your proposed way to me now? Is there any complete tutorial?

I don’t know any, sorry. To me, the combination of ascending and descending displacements as proposed in the formula of the study i mentioned above seems to be a good way.

Thanks alot. I have another quastion from you Mr Abraun, If I want to apply ps method on 3 swathes of all images, is it suitable using of merge option after splitting each swath? How much memory will be required?

I don’t know, sorry. So far I have only worked on single swaths (with a maximum number of 4 bursts). I think this will require a considerable amount of disk space and the use of tiling into multiple patches (can be defined at the end of mt_prep_snap).

Hello, I was talking about a single pair of interferograms, one ascending pair and one descending pair. With two look directions, you cannot get all three East, North, and Up components of displacement. There is simply not enough information.

The SNAP program has a function now to call snaphu.exe (in Windows) to unwrap one pair. As far as I know, there is no function in SNAP to unwrap a whole stack of interferograms.


Good day, Dr. Fielding!

I came across this post looking for some good discussions on how to get the components of the derived LOS displacements given that I have both ascending and descending data pairs. I just have few clarifications.

  1. If I want to combine my ascending and descending LOS displacement results to get the components, should the path/track/orbit be the same? This is one thing that is unclear to me. I’ve seen some posts here saying that the data pairs must have the same path/track/orbit (i.e. ascending track 100, descending track 100). But I’ve read some published papers and observed that when they did the decomposition for inversion analysis, the path/track/orbit for both ascending and descending passes are different.

  2. Do you have any reference regarding this qualitative addition and subtraction of ascending and descending results? I just wonder about the reliability of this simple approach for calculating the east and vertical components of the LOS measurement.

Thank you in advance for your time.


You have to use data from different tracks to combine ascending and descending interferograms for an area. Any area is imaged by different tracks on the ascending and descending orbits, and the frame numbers are also different. The track has to be the same for making interferograms, but the combination is done on the interferograms after they are geocoded.

This is a useful reference for understanding the estimation of components of the deformation from multiple interferograms:
Wright, T. J., Parsons, B. E., and Lu, Z., 2004, Toward mapping surface deformation in three dimensions using InSAR: Geophysical Research Letters, v. 31, no. 1, p. L01607.

1 Like