Data that does not appear when visualization of LiCSBAS results about licsbas HOT 25 OPEN

I want to ask, sir, how do I create 3D values in this LICSBAS process, such as separate EW and UD, is there a way to get 3D?

from licsbas.

Too many ifgs are discarded in step11 and step12 because of severe decorrelation and unwrap errors. Try changing the parameters for step11 and/or step12.

from licsbas.

Thank you sir for the answer, which parameters in steps 11 and 12 do I need to change? please explain sir

from licsbas.

This is my problem sir, I set it from 2015 to 2023, but if I plot the results as below, the data for 2015, 2018, 2019, 2022 and 2023 is not there, so what is the solution to my problem below?

from licsbas.

Discussion about the suitable parameters for a specific case is out of the scope of the issues in this repository. Please carefully read the LiCSBAS paper, understand the logic, try changing the parameters, and check the result yourself.

from licsbas.

Thank you sir for the explanation, I want to ask regarding the processing of one frame, there is a process that is not completed in step 14, do you have a solution to my problem, sir?

from licsbas.

I have no idea. Maybe resource (memory or storage) limitation?

from licsbas.

Hi Mr. Yumorishita

I have finished the processing on frame 143A_08674_131308, the results I get have a complete time series of 2015-2024 but why do I get a small area? even though I clip on "98.32/99.01/2.99/3.97". do you have any suggestions sir?

from licsbas.

Please upload the full log.

from licsbas.

this is sir
Uploading 202403241041batch_LiCSBAS_01_16.log…

from licsbas.

I cannot download it because the link is https://github.com/yumorishita/LiCSBAS/issues/335.

from licsbas.

Also please see mask_ts.png

from licsbas.

Here's the mask_ts.png, sir

from licsbas.

The data actually exists in the southern half and the problem is only in the visualization by LiCSBAS_plot_ts.py. I have no idea about the reason.

from licsbas.

Ah, the cum_filt.h5 includes only the northern half because there is no unmasked pixel before step16 filter. You will find the southern pixels by plotting only cum.h5, i.e., LiCSBAS_plot_ts.py -i cum.h5.

from licsbas.

OK sir, I'll try later
Sorry sir, I want to ask again, I tried reprocessing, but why does it get stuck, even though I have a good signal. Do you have a solution sir?

This stuck has been happening for the past 2 days

from licsbas.

Maybe it is because of the COMET server issue. Please try again.
#303 (comment)

from licsbas.

Excuse me sir, I want to ask why when we clip an area, after it is made into a tif the result when plotted is that the area is small and does not match the area that corresponds to the clip

from licsbas.

Can you show a real example?

from licsbas.

This is the AOI that I want to clip the image to with the coordinates that I adjusted

This is the result of the LiCSBAS process, which becomes small

202403290025batch_LiCSBAS_02_16.txt

from licsbas.

It is because most of the pixels were masked in step15 and lost in step16. Similat to this: #335 (comment)

from licsbas.

We cannot get 3D from 1 or 2 LOSs.

from licsbas.

Excuse me sir, I want to ask
how many steps I can set the parameters to filter COH values, I have tried several methods but have not succeeded can you help me sir

from licsbas.

You can set the coherence threshold in step04, step11, and step15. However, coherence is generally not a critical parameter in LiCSBAS.

from licsbas.

Hello @esykusuma ,
I checked the log and found that you set up loop_thre: 3.5 rad at step 12.
Is there any specific reason?
Also, I'd like to hear the view from @yumorishita regarding this matter.

from licsbas.