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

Microwave Toolbox StaMPS
#1063

Hi @ABraun,

But I also processed the same area before with SNAP v6.0 and StaMPS v3.3b.
At that time, I used 66 images of sentinel-1a during 2014/10/22-2017/10/30, and the result with a basemap is as followed (to emphasize the distribution of PS, I use the same color of them):

image
image.jpg1580×751 623 KB

Now, I used 24 images of sentinel 1b and 23 images of sentinel 1a during 2019/01/01 - 2019/09/28, and the result with a basemap is as followed:

image
image.jpg1583×746 645 KB

In the result I got before, the distribution of PS points seems randon. Also, I never got linear alignment of the PS in other cases before.

Because this is the first time I processed sentinel 1a and 1b data together with SNAP v7.0.2 and StaMPS v4.1 in Windows Subsystem for Linux (ubuntu 18). So, I’m not sure the result I have now is correct or not…

Thank you very much!

#1064

to my opinion, this is because you are investigating a shorter time (less temporal decorrelation), which leads to a higher density of usable PS

#1065

I just got the result (show as red point) using same imsges during 2019/01/01 - 2019/09/28 and parameters but processed with SNAP v6.0 and StaMPS v3.3b. They looks very different with the same data processed with SNAP v7.0 and Stmaps v4.1, the PS points show as black circle as followed:

image
image.jpg1584×751 661 KB

I don’t know which is correct …

#1066

I found what happen!

I didn’t notice that one parameter default set is different between StaMPS v3.3b and v4.1.
In v3.3b, the “weed_neighbours” is “y”, but in v4.1 it default set is “n”. So, I got more PS points!

But, which one is better of the parameter “weed_neighbours”?

Thank you.

#1067

weeding should be strong enough to remove faulty or noisy PS, but still leave enough PS for the later steps. If the pattern you see makes sense, leave them in.

1 Like
#1068

OK! Thank you for your help and explanation.

#1069

where we can get this ps_load_initial_gamma.m file? kindly if you can explain it…

#1070

Replacing scripts is no longer needed since StaMPS 4.1, you directly use mt_prep_snap

#1071

Dear Katerine, Daer other users,
I have follow the steps that Katerine gave. Everything is going well except the StaMPS export.
In the StaMPS Export, I have put the 2 right files but I faced a wrong message (“Orthorectified lat/long bands required. Please add the bands to the interferogram product”). I do not know why I received this particular message. Someone have an idea how I can solve it. Thanks for your support and your help.
Francois

image
image.png943×506 15.2 KB

#1072

did you check this option during the interferogram generation?

grafik

It is deactivated by standard.

#1073

Thanks for your support. It works now !

#1074

Hello,
Why this error message?
I worked the data with PSI.
Any suggestions please.

Error%20plot%20in%20matlab
Error plot in matlab.png1744×1032 326 KB

Thank you,

#1075

good morning, could you solve that problem? please

#1076

Hi all, I have the following error during the run of PATCH 4… what can I make???

error
error.png1920×1080 185 KB

I am using StaMPS v3.3b1!!!

Thanks for advance

#1077

have you installed triangle on your machine?

#1078

For variable ‘a’ you must run atmospheric correction from TRAIN you can find in https://davidbekaert.com/
And to plot ‘a’ you must add additional string e.g. ps_plot(‘u-a’,‘a_linear’).

#1079

Ready! I just installed it!!!
Run again?

#1080

yes, run step 4 entirely again.

#1081

ok running…

thanks Andreas

#1082

you also need to have snaphu compiled for step 4 (sorry, I don’t know which step you were at with the triangle error.)