Combining ascending and descending data?

[simonevaccari]

Hi all,
is it possible to combine ascending and descending Sentinel-1A GRD IW VV data in a simple way, in order to increase the amount/availability of images over my area of interest? What I am thinking of is to follow such an approach (for each of the Sentinel-1A GRD IW VV ascending and descending data acquired):

1. Radiometric - Calibration;
2. Speckle filtering
3. Radiometric - Terrain Flattening
4. Geometric - Range-Doppler Terrain Correction.

After this I should have gamma0 for both the descending and the ascending data. At this point, could I simply average the two gamma0 data if I want to use both the images (providing that they are acquired with little time difference).

many thanks
Simone

[mengdahl]

In theory that should work but since the current implementation of Terrain Flattening does not work properly you might run into problems in hilly/mountainous areas. On almost flat terrain the approach should work already now.

[simonevaccari]

Thank you Mengdahl for your answer. so provided that the Terrain Flattening works properly, then it should be possible to just average the two data together?

[SAR2016]

Even if the terrain flattening works, how we can combine Asc and Des image values where two images have different sizes. Even though these two images are taken from the same area of interest, they do not overlap completely. In my dataset, Asce image covers a larger area and include some major parts of Descending image.

[falahfakhri]

Also might be this answer is interesting

Values of displacement from S1-A SLC pair

Any space-born SAR system has both ascending and descending orbits, for which acquisitions cannot be at exactly the same time. Selecting insar pairs from orbits close in time is a common practice, assuming no much change happened in between. Please note that asc & desc insar measurements are done along different line-of-sight (LOS) and cannot be combined simply using averages. You need to geometrically combine considering acquisition geometries. There should be several related publications in l…

[tarik44]

Hi,

I need to create DEM with InSAR. I have ascending and descending data. When I create DEM for each ascending and descending data, I saw layover or shadow, it means data loss.

My question is, how can I use together ascending and descending data for DEM generation? If You know pls write step by step.

Thanks

[ABraun]

Maybe you can try something like this:

1. Create a DEM of both the ascending and descending image pair as described here: The Order of DEM Creating Steps
2. Use the SAR simulation tool for both the ascending and descending image to create a layover mask
3. Use the layover mask of each product to mask out the invalid regions in your dems
4. Stack both DEMs and create a merge product with the band maths (e.g. the average of both DEMs)

But note that creating dems requires very good input data regarding temporal and perpendicular baseline and minimized decorrelation. What data are you using and where is it located?

[tarik44]

@ABraun

But note that creating dems requires very good input data regarding temporal and perpendicular baseline and minimized decorrelation. What data are you using and where is it located?

I am using TerraSAR-X Stripmap mode for Erciyes Mountain, Kayseri,TURKEY. I created DEM from SAR images with use external 5m resolution DEM.
I’m trying to SAR Simulation

[ABraun]

Alright, if the time between both images is short and vegetation is sparse, it could work out.

[Nizar]

Dear all,
Now, I am familier with the procedure for Sentinel1 Snap-StaMPS PSI (SLC) but by using one orbit (descending or ascending). I need to retrieve east-west displacements so I need to use both orbits. My question is how to combine them to get the east-west displacement?. What are the required additional steps?. Any tutorial, please?. Thank you.

[dap]

Hi everyone, I have a related question. I have processed daily Sentinel images to calculate surface water. For any given polygon, I have two series: one from the descending orbits, and one from the ascending ones. Is there any guidance for how to combine them? For instance, it is often the case that the ascending values are much smaller than the descending ones (i.e. 1 percent on average), but multiplying all ascending values by 100 produces some extreme values on flood days. Thank you so much.