How to prepare Sentinel-1 images stack for PSI/SBAS in SNAP 5


#305

Do you know that results by toporemove tool with a stack of images are wrong in Snap 5.0?


#306

No, I do not know,

What should I do then

Thank you for quick response


#307

If you want use snap 5.0, in this topic you can find all info for the “Stamps preprocessing”. Alternatively you could use snap 6.0 preview 4 (preview 5 has an error in Stamps export).


#308

Dear Jefriza,
It seems strange for me why you have so many products to export to StaMPS. After creating a Stack with Backgeocoding (point 2 in my workflow) I have only two products to export to StaMPS with bands corresponding to images (or pairs). I attach the products which I export to StaMPS to illustrate this. My advise is to start from the beginning with Snap 6.0 preview 4. There was a link to download it.


Good luck,
Katherine


#309

Thank you for your response Katherine,

“I have only two products to export to StaMPS with bands corresponding to images (or pairs)”

Because I do not know how to deal with batch graph processing and so I choose to create graph manually and pairing 2 SLC images into each folder.
My desktop run very slow (12 GB Ram, Intel ®Xeon to deburst each image pair, and after subset it become faster to create dinsar (+elev band).

Ok. There is SNAP 6.0 preview 5 in the website, and I am still downloading it.

Thank you


#310

Probably you have to download preveiw4. Check this.

Snap 6 Preview 4 can be downloaded from

http://step.esa.int/downloads/6.0-pre4/
Good luck,
Katherine


#311

Hi @Katherine

I have the same problem that @jefriza, I followed the steps from @bayzidul, but now have one question for you
When you speaks of only two products for export for StaMPS, you maked (point 2) the Stack with all images that you had???for example I have 22 images s1, one is master and the rest slaves, the Stack is with the 22 images??? with Split and Orbit for each one??? and after Deburst?
Why you stack have lon_lat_TC???

Sorry for severals question :sweat:


#312

Dear Gabriela,
This is the text of point 2 of my guide.

2/Create stack using Backgeocoding ( Radar-Coregistration-S-1Tops_Coregistration-S-1Backgeocoding. Master should be the first in the ProductSetReader). 12-13 images are OK for each Stack. If more images in the Stack - it is better to split it into several Stacks each having the same master. Master can be chosen by Radar-Interferometric-Insar Stack Overview.

  1. I had to devide my 25 images into two stacks having the same master. In this case I exported 4 products to Stamps.
  2. My Stack has lon/lat bands as I exported them to ENVI or Gamma format to get binary files .lat and .lon (to have input to StaMPS geocoded) . Point 9 of my guide.
  3. About TC, please, read posts of Annamaria. By TC she managed to solve the problem of shifted results.
    Good luck,
    Katherine

#313

thank you for the step Katherine,

Now I am using SNAP 6.0 Preview 5. And I try to export 2 file but it said Error: [NodeId:StampsExport] The product ‘subset_5_of S1A_IW_SLC_1126_0113_Orb_Stack_deb_Ifg_dinsar’ already contains a band with the name ‘i_20170113.rslc’

thank you


#314

same problem…:wink:


#315

Some people had problems with SNap 6 preveiw 5. Probably it is better to try Snap6 preview4.

Snap 6 Preview 4 can be downloaded from

http://step.esa.int/downloads/6.0-pre4/
Katherine


#316

Hi all,
Firstly, so many thanks for your help (@bayzidul, @ABraun @katherine etc.). I used 32 Slave and 1 master for my study area and I divided them 10 Slave 1 master subparts for PSI. I applied PSI process according to your guidance. I think I achieved PSI export and mt_prep_gamma sections succesfully. pscand.ij files are not empty now. mt_prep_gamma created one patch without giving any error.

But I have problem with ps_load_initial_gamma. I tried to use this changed .m file from Katherina and Feiliu but still I got same error. I could not find this error anywhere in this forum and MAINSAR. Additionally, the usage date of my MATLAB full version expired and I am using classroom student MATLAB provided by university without some toolboxes . Can this error be related to version of MATLAB? I am at the last step and need your suggestions and help. Thanks.
Note: I changed processor to gamma using setparm. Also MATLAB version is 2016b. I tried with full version of MATLAB and got same error.


#317

I try not to subset. and StaMPS export work now.

but I want to make it in small area (subset) to more detail in and reduce time. Thank you.


#318

Hi @Fikretjfm

Would you like to give another try with this script. Just copy this in the stamps MATLAB directory.
ps_load_initial_gamma(changed).m (5.8 KB)


#319

Hi @bayzidul,

Thanks but I got same error. Could you share your mt_prep_gamma_changed file. Maybe it is related to previous process, because nobody faced with this error. I want to try mt_prep_gamma step again.


#320

Hi @Fikretjfm
Here we go…
mt_prep_gamma_snap(changed) (6.4 KB)


#321

Hi @bayzidul ,

I tried and got same error. Please tell me which Matlab version you used? I dont understand what I am doing wrong. I followed all instructions and at the last step I faced with this error that nobody faced except me. How can it be possible? I am using same files and I got error. :confused:
Please @ABraun, @katherine @FeiLiu check this error what will be reason?

Thanks


#322

Because you are using Windows, Try Linux~


#323

I followed bayzidul step.

A litle bit confusing,
in step 3 should be: create stack back geocoded images must followed by deburst (in order to read in stamp export)

and of course in step 4: results of step 3 + ifg +dinsar (+elev band).
step 5 makes reader confusing.

deburst --> ifg to shortcut step, if we perform ifg --> deburts,
you wasting your time for making the 2 files you need for stamps export which is (step 4 and step 5 in opinion)

so where I should perform subset?
without subset, deburts 5 images take time 1027 minutes!
now I upgrade ram from 12 to 20 GB.

best regard,

jefriza


#324

Hi @jefriza

I had posted in August and again edited in October, but now it is November, so things can change with time. Those were the steps with that I got some good results. I was mainly following some previous instructions which you can also check.
Deburst can be the second step and then subset to make further steps quicker. I was getting an error message to do subset before doing deburst. I think it does not matter too much if you do deburst beofre or after ifg generation. In general, you can adapt your steps and you can test it.