Difference between vertical and LOS displacement

Hi all,
Is there any reference cited the below equation?
(unwrapped phase * wavelength) / (-4 * PI * rad(cos(incidence angle)))

http://www.esa.int/esapub/tm/tm19/TM-19_ptA.pdf54, chapter 2.3.

Good morning,
I have read your post and i have a questio for you.
If the vertical displacement is:

unwrapped phase * wavelength) / (-4 * PI * rad(cos(incidence angle)))
the horizontal displacement is:

unwrapped phase * wavelength) / (-4 * PI * rad(sin(incidence angle)))??
Please let me know.
Thanks.
Have a nice day

Good evening to everyone,
I’m new on this subject.
I was entrusted with a project that concerns the detection of the displacement of the land both horizontally and vertically.
I started to study the basics on Sentinel (different types of products, modes, etc.) and I downloaded SNAP.
I have read several guides and tutorials and thanks to your help I have obtained results.
I need to know if what I have done is correct and if I have finished with the steps or I need other operations to do before getting the final results.
The steps I have performed are the following:

1 I took from the copernicus hub two images of the same place at a distance of 12 days (S1A_SLC_IW).
2 I created this chart on SNAP and I ran it:

3 I exported the result to SNAPHU
4 I have performed unwrapping on SNAPHU
5 I imported the result on SNAP
6 Phase to displacement -> (unwrapped phase * wavelength) / (-4 * PI)
where according to my calculations the wavelenght is 0.055m
7 terrain correction
8 I have marked the value of displacement_Terrain Correction in a high coherence point (taken from the coherence band obtained by unwrapping)
9 terrain correction - pixel value

I then generated the vertical and the horizontal displacement like this:

10 Vertical:
(unwrapped phase * wavelength) / (-4 * PI * rad (cos (incidence angle)))

11 Horizontal:
(unwrapped phase * wavelength) / (-4 * PI * rad (sin (incidence angle)))

12 I applied steps 7,8,9 for both vertical and horizontal displacement

After that I moved on coherence:

13 I performed the terrein correction of the product obtained from step 2
14 I created a math band with the following formula:

If coherence band> 0.4 then 1 else NaN

15 I created a stack respectively between the product obtained from step 14 and steps 9 and 12 (three different stacks one for the displacement one for the vertical and one for the horizontal)

16 I ran the maths bands

result step 14 * result step 9
result step 14 * result step 15 (two different operations)

What I did is correct?

I’m finished or do I have to take other steps? (I read on the forum that maybe I have to use STAMPS for PSI)?

How can I tell if the results are right or not?

please help me.
thank you.

Hi,

When doing measurements with SAR sensor to identify movements on the ground, we can resolve only one surface displacement vector. What the satellite measures is the distance between the satellite sensor and the ground but we cannot say that the ground was moved horizontally or vertically using only one satellite orbit (ascending or descending). If you want, we can make the assumption that the ground was moved vertically (we cannot be certain for that). SAR is much more sensitive to vertical changes rather than to horizontal ones. We need to combine ascending and descending orbits to resolve for the ‘true’ ground displacement.

1 Like

Thank you for your answer.
I need to get the horizontal and vertical displacement of the ground.
So you’re telling me that to do this I need to process two images of the same point SLC-IW one with ascending step and one with descending step?
And if the images are more than two how do I do it?
Please let me know also because reading in this forum I read the various steps to follow (please tell me if the steps I did that I listed in the previous post are correct) but I never read that the images should be one ascendant and one descendent.
This means that I was wrong :frowning:
Thanks for your help

Strictly speaking, to get the optimum results in intereforometry, the satellite orbits should be the same. The limitation of this is that only one surface displacement vector can be resolved. So, many studies carried out using both ascending and descending orbits to isolate vertical from horizontal motion. have a look on these articles.


You can collect a few images derived by both ascending and descending orbits and do the interferometric processing and see what you get.
Your graph looks correct

1 Like

Thank you very much for you help.
I will let you know.
Have a nice week end.

do you know if are the te online courses on these topics?
not having basic knowledge I follow what I find in some tutorials or in this forum.
I would need to have some material available and take courses

For getting familiarised with SAR data you can check this online course:
https://eo-college.org/courses/echoes-in-space/

2 Likes

Dear all,
in order to calculate the vertical displacement I used the following formula:

(unwrapped phase * wavelength) / (-4 * PI * rad (cos (incidence angle)))

I got the incidence angle from “show tie point grid” inside math band when I applied the formula.
Is it correct?
If yes, can you tell me how can I know the numerical value of this angle in order to apply the formula without that snap automatically taking it?
I’m using S1A_SLC_IW images
Thank you.

I also need some advice from you.
I would like to buy sensors to interface with Arduino in order to have a comparison with the displacement obtained with SNAP.
Can any of you advise me one?
Thank you.

Dear all,
nobody can help me?
Please give me a feedback.
best regards.

Good morning,
could you tell me how I can obtain sentinel images in ascending and discending orbit?
I have tried with copernicus hub but i cannot find where I have to select this option.
Thanks.
best regards.

HI,

The website for downloading SENTINEL data is the following:
https://scihub.copernicus.eu/dhus/#/home

Once you are logged in, on the top left corner, press the symbol that looks like a cone. and a pop up window will open as shown in the picture below:
You can choose order by descending or ascending. If you want ascending orbits to appear first in your results, you can order by ascending

Ok thanks,
I thought that this field give the chance to change the order of the images based on the period (from new date to old date and the other way around) and not the orbit.
Thanks for your support

I’m sorry to bother you again but I need more help from you.
I processed two images with a descending step and I got the vertical displacement.
Now I should get the horizontal displacement and they told me to use the ascending step.
My problem is that I should do it on the same dates and I noticed that there are no images on the same day with both steps (ascending and descendent).
How I have to proceed?

Ascending and descending orbits are used for increasing the revisit frequency so that we have more datasets for our AOIs. When the satellite observes your AOI in ascending orbit, as the Earth rotates Eastwards, the satellite will observe the same area in descending orbit as well. So, do not expect the satellite to pass over your AOI exactly at the same date in both descending and ascending orbit.
Lets say you want to derive ground subsidence after a volcano eruption, you can download two or more images in ascending orbit (before and after) and another few images in descending orbit (before and after). Then, you can use two pair of images in ascending and descending orbits for interferometric analysis.

Ok thanks…
so If I want to analyze the vertical and the horizontal displacement of a city I have to:
VERTICAL DISPLACEMENT

  • use two images with discending pass and at the end of the process apply the following formula
    (unwrapped phase * wavelength) / (-4 * PI * rad (cos (incidence angle)))
    HORIZONTAL DISPLACEMENT
  • use other two images with ascending pass in different days respect to the first two and apply the following formula:
    (unwrapped phase * wavelength) / (-4 * PI * rad (sin (incidence angle)))
    Is it correct?
    if it is correct how can I verify if my results are correct using the pitagora theorem if the images are not the same?

You should create an interferogram from one ascending and one ascending orbit. Due to atmospheric noise, you better create a stack (e.g you can collect 10 images of ascending orbits and 10 images of descending orbits) of ascending and descending orbits.
take a look at the paper: Inferring_three-dimensional_surface_displacement.pdf (2.6 MB)

In order to verify your results, you need GPS measurements for your AOI