MintPy processing close to the water

[Sven]

Hi Yunjun,

sorry for disturbing you.

I'm processing RS2 data over my AOI, which is close to the sea.

Processing seems to be fine, as it goes smoothly till the end, but when I look at the "geo_velocity.kmz" file on Google Earth this is on the water, not over the land (see the attached file).

Moreover, the dimension of the figure is not the one I would expect, but maybe this was my fault, as I probably used wrong multilook factors (6, 6).

Also the processing template is attached for inspection.

Consider that I tried many combinations of mintpy.subset/mintpy.reference .yx and .lalo, with many error messages and aborted processing: the .yx combination in the attached template was the best one.

It'a a pity because the deformation signal in the AOI is very clear, in a good agreement with GPS data we have in the area.

Cheers,

Sven.

 

[Yunjun Zhang]

Hi Sven,

It seems like a problem during the geocoding process, and more specifically, the geometry file. From my rough estimation based on your screenshot, assuming your AOI is Campi Flegrei, the lookup table in geometryRadar.h5 file has lower (~6 times) resolution. I guess your inputs/geometryRadar.h5 file is not consistent with your dataset: did it go through multilooking twice? If you could send me your inputs/geometryRadar.h5 file and velocity.h5 file, I can check more.

Yunjun

--
You received this message because you are subscribed to the Google Groups "MintPy" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mintpy+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/mintpy/3a5dba34-c4ae-413d-942b-4ed02562de1e%40googlegroups.com.
<MintPy_RS2_Naples.png><RS2_Napoli.txt>

[Sven]

Hi Yunjun,

yes, my AOI is Campi Flegrei.

You hit the problem: my concern during processing was multilook.

I initially used a multilook factor (6.6) to comply with the DEM pixel spacing, in the ISCE "stackStripMap.py" script (ref. https://github.com/isce-framework/isce2/tree/master/contrib/stack/stripmapStack) and I got the result you saw.

Afterwards I tried also with the "run_multilook_geometry.sh" script from MintPy and again I got the same result.

Could you please guide me with the correct sequence to get a correct result? I mean, about multilook.

Should I insert multilook factors in the stackStripMap command or not? And the run_multilook_geometry script? Should I use it or not, and when if so?

Attached you'll find "geometryRadar.h5" and "velocity.h5" files, together with the related figures for quick inspection.

Thanks,

Sven.

[Yunjun Zhang]

Hi Sven,

The full resolution geometry file is loaded, as shown below, which caused the issue you have.

[yunjunz:~/Downloads] yunjunz% info.py geometryRadar.h5 | egrep 'file_name'
extra_file_name                      /home/pc-gamma/Scrivania/CSK_RS2_TSX/RS2/Radarsat2/merged/geom_master/hgt.rdr.vrt
file_name                            /home/pc-gamma/Scrivania/CSK_RS2_TSX/RS2/Radarsat2/merged/geom_master/hgt.rdr

Let’s say the root directory of your dataset is: /home/pc-gamma/Scrivania/CSK_RS2_TSX/RS2/Radarsat2. The command sequence would be:
1. Finish the README.txt from ISCE/stripmapStack, with -a and -r option specified for stackStripMap.py (-a 6 -r 6 in your case). You should use “--nofocus --zero” option as RS2 SLCs are zero doppler. Afterwards, you will have ./merged/geom_master.
2. Go to the root directory; copy $MINTPY_HOME/sh/run_multilook_geometry.sh file; run run_multilook_geometry.sh 6 6 #the same value used in stackStripMap.py. This will generate ./geom_master folder.
3. Create mintpy folder and go inside; copy your RS2_Napoli.txt here; then run smallbaselineApp.py RS2_Napoli.txt. To check if the geometry files are consistent with your interferograms, smallbaselineApp.py should also work without subset.

I hope this is clear enough,

Yunjun

-- 
You received this message because you are subscribed to the Google Groups "MintPy" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mintpy+un...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/mintpy/19c9fbd8-9efc-4a2d-8dbb-16bb4457d51d%40googlegroups.com.
<geometryRadar.h5><velocity.h5><Figure_1_-_geometryRadar.png.png><Figure_1_-_velocity.png.png>

[Sven]

Hi Yunjun,

got it :)) (see the attached file); now I'll generate also the waterMask file to see if I can improve the result!

Just a few questions more.

1) stripmapStack processing in ISCE doesn't create the "masterShelve" folder in the root dir, needed for "prep_isce": I picked up one of these folders elsewhere and put it into the root dir (they seem to be all the same). Is this ok?

I think so, as I got the right output at the end!

2) In order to create the *.kmz timeseries file, I was not able to get from processing the "geo_velocity_mask.h5" but only the "geo_velocity.h5" file, that I used. Is this ok?

3) Instead of copying every time "run_multilook_geometry.sh" into the root dir, may I add $MINTPY_HOME/sh/ into the PATH? I could afterwards run the command with "bash run_multilook_geometry.sh alook rlooks", right?

4) I was not able to run "plot_smallbaselineApp.sh" as i got an error message at line 26: how can I fix this?

B.t.w. many thanks for your guidance; now I'll try with more data, as this was just to test some RS2 data i had.

Sven.

Il giorno martedì 10 settembre 2019 15:14:39 UTC+2, Sven ha scritto:

[Yunjun Zhang]

This is great!

1) stripmapStack processing in ISCE doesn't create the "masterShelve" folder in the root dir, needed for "prep_isce": I picked up one of these folders elsewhere and put it into the root dir (they seem to be all the same). Is this ok?

I think so, as I got the right output at the end!

Yes, as you said.

2) In order to create the *.kmz timeseries file, I was not able to get from processing the "geo_velocity_mask.h5" but only the "geo_velocity.h5" file, that I used. Is this ok?

Yes. Note that geo_maskTempCoh.h5 file is recommended while calling save_kmz_timeseries.py, and the masking will be applied on the fly.

3) Instead of copying every time "run_multilook_geometry.sh" into the root dir, may I add $MINTPY_HOME/sh/ into the PATH? I could afterwards run the command with "bash run_multilook_geometry.sh alook rlooks", right?

Yes, just keep in mind this script need to be run at the root processing directory.

4) I was not able to run "plot_smallbaselineApp.sh" as i got an error message at line 26: how can I fix this?

I can not tell without more information.

Cheers Sven!

Yunjun

-- 
You received this message because you are subscribed to the Google Groups "MintPy" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mintpy+un...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/mintpy/084ca396-a793-4e2a-9ba7-b1bc51b0caa4%40googlegroups.com.
<MintPy_CampiFlegrei.png>

[Sven]

Hi Yunjun,

about the error message I had with "plot_smallbaselineApp", in the attached file you'll find the output.

About the use of "save_kmz_timeseries" I'm a bit confused about the flags I should use: on the MintPy web page the full command is

"save_kmz_timeseries.py geo_timeseries_ECMWF_ramp_demErr.h5 (in my case I have ERA_ramp...) --vel geo_velocity_mask.h5 --tcoh geo_temporalCoherence.h5"

We already discussed about "geo_velocity_mask": I didn't get this file during processing (how can I get it?), just "geo_velocity" but you told me I can safely use it.
And what about "geo_temporalCoherence"? You told me about the importance of "geo_maskTempCoh": should I use this latter instead of "geo_temporalCoherence"?

At the end, which is the full command for "save_kmz_timeseries", considering "geo_velocity mask" doesn't come out from processing?

I hope I was clear enough, sorry for confusion ;)
Sven.

Il giorno martedì 10 settembre 2019 15:14:39 UTC+2, Sven ha scritto:

[Yunjun Zhang]

Hi Sven,

I have updated plot_smallbaselineApp.sh script by removing the variable definition that caused the error you have. Please update your code and let me know if there is still a problem.

Run save_kmz_timeseries.py -h to the latest example usage, that's the most up-to-date version. I have removed the command line example from readthedocs, to avoid the confusion.

Thank you for pointing it out.

Yunjun

[francesco menniti]

Hi Sven

I am also working in the Campi Flegrei area (I am a researcher). I can kindly ask you to share the final speed and in case also the parameters used for processing! So you can compare the data.

[Sven Borgstrom]

Hi Francesco,

I'm a researcher as well, at INGV. I recently wrote a (monitoring) report over this area with very recent results, from S-1A/B data. Just give me a couple of days to check the mean LOS velocity, currently at a rate of about 10 cm/year and the parameters I used in the template.

Cheers,

Sven.

[Sven Borgstrom]

Hi Francesco, hi Yunjun,

sorry for late answer.

Francesco, I don't think people are really interested in our results so I think it's better, if you agree, to get in touch via email, avoiding to disturb here the other colleagues.

What do you think?

In order to communicate, is my email address visible here?

Cheers,

Sven.