Your problem motivated me to add a new feature to the StaMPS-Visualizer You can now use the “Subtr. Offset” Button to subtract the offset like you described it to align the time sereis to be 0 at the first date. Example:
The old way, before you use the Offset button (still available and the default plot after you select a point):
The new plot, after you have used the Offset button:
Enjoy 
You mean by subtract is to remove from my dataset/analysis the outliers (i.e., two spikes), isn’t? Thank you.
First, thank you for providing us this great visualization tool for StaMPS-derived results. It’s really very nice. I think I have to upgrade from the beta version to the latest one (v2.1).
Also, I am just curious if StaMPS also provides the temporal coherence of the selected PS? Or is it assumed that we do get relatively high temporal coherence with just using the ADI?
I have one suggestion also (since I am not a programming expert). Would it be nice if StaMPS-Visualizer can generate a smooth pixel-based displacement map overlaid with a displacement contour?
Thank you.
That I do not know, I have not seen an objcet like that, but it is worth to have a look into the objects within the Matlab environment after you have run StaMPS.
Actually, I once started programming a 2D KDE to interpolate the points, hopefully I will find the time and add this feature, contour lines from those interpolated raster should be no problem, thechnically, but I do not know if this is helpful every time. However, there are many potential improvements for the visualizer.
Thank you for giving some light on this. Because in SARPROZ or even in SqueeSAR results they provide temporal coherence of the PS points.
Anyway, I have one concern regarding the CSV file. I know that the first two export_res_XX correspond to the lon lat coordinates of the points. I just wonder if export_res_3 corresponds to the displacement for the first date? In the StaMPS Visualizer, there are 77 date-to-date points but in the CSV file, after the first two export_res_(1 and 2), I have 78 data points.
Thank you.
Have a look here
I got it. So export_res_3 corresponds to the mean velocity. I realized also that PS point no. in StaMPS Visualizer is PS point no. + 2 in the csv file (i.e., PS point no. 100 in StaMPS Visualizer is cell number 102 in csv).
Thank you again!
Hi, @ABraun and @mdelgado! Just a quick question. I was trying to find the answer in some posts here but I don’t remember where I saw the same problem I have now.
I got one pair (from 60 pairs in total) with zero mean amplitude prior to launching matlab for PSI analysis with StaMPS. I proceeded and adjusted some parameters, and executed step 1. I got an error. I suspect it’s because of that pair with zero mean amplitude.
What I did is to remove that pair from the analysis. The PSI analysis was successfully completed. Was it okay to delete that pair with zero mean amplitude?
Thanks in advance!
I believe so, you got a very good representation of 59 images. 1/60 is fine.
You could also had tried to re-process it.
But delete it is fine.
I agree with @mdelgado, the topic you are looking for was discussed e.g. here
and here
If there are more cases of zero mean amplitude it is worth redoing the preprocessing in SNAP, this always solved the issue. I highly recommend checking the results after each preprocessing step to figure out such problems, since they often appear in the first steps.
Thank you for your immediate response. I’ve only checked a few interferograms in SNAP before the stamps_export through the snap2stamps tool.
I am tempted to redo it actually but I noticed that towards the end of the coregistration and interferogram generation process, it takes more than 4 hrs to finish 1 pair. In the early part of it, it’s relatively fast. I can redo it but I am sharing the PC with another person who uses the same PC for his own research also.
Anyway, thank you again!
Stay safe and healthy always.
PS: Fortunately, that pair that I removed is quite distant from the main event.
If it is only one ifg, it might not influence the StaMPS process so much.
However, concerning your processing problem:
Check if you can increase the RAM that can be used by SNAP Increase snappy memory (beginner) snap2stamps uses the gpt API of SNAP, check out the snap2stamps Manual 1.1 Installation in order to set up the memory for gpt
If possible, write the outputs to a SSD, for me that really made a significant difference in processing
I did modify the gpt.vmoptions also in snap. I just wonder where can I find the snappy folder? I am not familiar with this one to be honest.
please have a look at my last post, I edited it since you are right gpt should do, snappy is not used, as far as I see. If @mdelgado can confirm would be great.
If this is the case, I know where to look for it. As I have mentioned, I’ve modified it following the one recommended in the snap2stamps manual.
Even the splitting part takes too long also when it progresses.
Thank you again.
Confirm that snap2stamps does not uses snappy, but some python wrappers that execute gpt with the customised xml graphs
Hi!
Can anyone please give me a reference to cite for pre-processing of SAR data before and up to using snap2stamps?
Thank you!
how about this one? https://gitlab.com/Rexthor/gis-blog/-/blob/master/StaMPS/2_stamps_workflow.md
I don’t think there is any official, scientific or peer-reviewed reference you could use.
I think this should work. I just want to include in my thesis. I did not see any name of the author/developer. What name should I use for this citation?
Thank you!
I’ll send you a private message.
