Problem with S1-TOPS Coregistration

Hello
I am getting a problem using SNAP when I try to aplly the interferometry. This problem occurs while trying to coregistrate the two images. an error appears:
" Error: [Noted: Back-Geocoding] org.jblas.NativeBlas.dgemm…"

Does anyone know what is this ?

Which products of S1 do use?

What are the previous processes , did you apply?

Please give more details to get the right assistance,

I am using S1B_IW_SLC products, I am applying the first step of interferometry which consists to coregistrate the SLC products…

Please take a look at the processing steps in the following post, I explained in details,

Source of the post

I am following your processing chain but i got the same problem in the step of Back-Geocoding.

Have you installed all module updates?

Please show up the whole names of your two products, in the above panel you shared,

Yes, as @mengdahl mentioned please check for updates from help tab check for updates,

“S1B_IW_SLC__1SDV_20180825T122515_20180825T122543_012419_016E63_239A.zip”
“S1B_IW_SLC__1SDV_20180813T122514_20180813T122542_012244_0168F8_5062.zip”

The names of your two products refer to no previous processes were being done,

Yes I thought about it. But I am working in a server, I hope I am allowed to update it myself. Else I will wait till next week…

Of course, the are the SLC zipped products… The input of the Back-Geocoding steps have the names with the extension _split_Orb

This one I was looking for, I mentioned in my previous post,

Now, did you split the same bursts of each product?

Yes, I think as our friend mendahl, I need to do some updates…

In this case, please also take a look at this post, explains, how to select the correspondent slice and frame,

Source of the post

Dear all,

I’ve coregistered two S1 images the first time with SRTM 1sec HGT and the second time with SRTM 3sec.
In the first try the “mask out areas with no elevation” didn’t have any effect and the area coincident with the mediterranean sea was not masked out (image A).
In the second case the “mask out areas with no elevation” did have the right effect and the area coincident with the mediterranean sea was masked out (image B).

I have two question.
1. Why with the use of the SRTM 1sec HGT the sea isn’t masked out? Which 1sec DTM allows the masking of the sea?
2. Coregistering with the SRTM 3sec affects the resolution of the two images?

Thanks in advance for the suggestions
Daniele

  1. maybe the area is considered shelf or flat shore in the 3Sec SRTM and water in the 1Sec SRTM? If you check the “DEM” option during terrain correction, the underlying SRTM data is also written into the target product. You can then check with the pixel info tool what values are used for this area in 1Sec and 3Sec.

  2. The spatial resolution of the primary image defines the spatial resolution of the output. The SRTM is upsampled for this task.

thankyou @ABraun.

The suggestions were very appreciated :slight_smile:

Daniele