Snap2stamps package: a free tool to automate the SNAP-StaMPS Workflow

Hi mdelgado ,
you were right it was the problem of coordinates
now i have this warning when i process “coreg” :

IFG: isTOPSARBurstProduct = true
INFO: org.esa.snap.core.dataop.dem.ElevationFile: http retrieving http://step.esa.int/auxdata/dem/SRTMGL1/N37E001.SRTMGL1.hgt.zip
WARNING: org.esa.snap.core.dataop.dem.ElevationFile: http error:http://step.esa.int/auxdata/dem/SRTMGL1/N37E001.SRTMGL1.hgt.zip on http://step.esa.int/auxdata/dem/SRTMGL1/N37E001.SRTMGL1.hgt.zip
INFO: org.esa.snap.core.dataop.dem.ElevationFile: http retrieving http://step.esa.int/auxdata/dem/SRTMGL1/N37E000.SRTMGL1.hgt.zip
WARNING: org.esa.snap.core.dataop.dem.ElevationFile: http error:http://step.esa.int/auxdata/dem/SRTMGL1/N37E000.SRTMGL1.hgt.zip on http://step.esa.int/auxdata/dem/SRTMGL1/N37E000.SRTMGL1.hgt.zip

is that will cause a problem???

Thanks a lot

Mmmm, in this case please check.

Is your area in the middle of the Mediterranean Sea?
It is normal that there is no DEM over water… and looking at the coordinates that gives the warning… this seems to be the reason (from my understanding)

Without DEM they will not be any topographic phase removal.
Please check if your area contains also other piece of DEM and not only the one giving the error.

I hope this helps

Hi Mdelgado ,

my area of interest include (IW3 and IW2) . my first work was the splitting of all my data with IW2, and i have done the process snap2stamps .
my question is how to combine the IW3 and IW2 to have my area of interest ???

cheers

Well…Using my scripts you may run the steps 2 and 3 for each of the IWs and then you may need to include also the merging in a latter step. This is not in the current snap2stamps version but it is quite easy to implement it and include it in a second release

Hi @mdelgado,

I just tried to use snap2stamps code to process images with SNAP v7.0.2.

My study is area just in one burst, so I have already deleted the ESD operator in coreg_ifg_computation.xml and coreg_ifg_computation_subset.xml.

But there still have the same error message as followed:
Error: [NodeId: Enhanced-Spectral-Diversity] Operator ‘SpestralDiversityOp’: Unknown element ‘useSuppliedShifts’

Could you help me to deal with this problem?
Thank you very much!

Let me try.

Which file had you modified? It seems that SNAP is still getting the one containing the ESD operator.

Secondly, SNAP 7 modified the ESD operator changing the parameters contained in the xml, but if you do not need it, you should not care about updating them.

I hope this helps

Hi @mdelgado

I modified these two files: coreg_ifg_computation.xml and coreg_ifg_computation_subset.xml

I don’t really understand what you mean. It means in SNAP 7, I don’t need to delete ESD operator even if I use one burst data? Thank you!

You remove the dependency of the ESD operator as I can see, but had you removed it from the xml? I can see some of the ESD tags on top of the interferogram block.

Please remove it totally from the xml graph with the subsetting, as it is the one used by default in snap2stamps and try again.

Let me know

It work after remove the ESD operator totally!
Thank you very very much!

1 Like

Hi @mdelgado,

Sorry to disturb you again.
I tried another study area, more than one burst. So, I keep ESD operator as default.
But I received the same error message as before I only processed one burst with ESD:

Error: [NodeId: Enhanced-Spectral-Diversity] Operator ‘SpestralDiversityOp’: Unknown element ‘useSuppliedShifts’

Could you help me again? Thank you very very much!

As said already, SNAP 7 changed the xml tags, so you should get them and update the graph provided within snap2stamps, as they were ready and aligned with SNAP 6.0

running the command gpt -h SpectralDiversiyOp should give you the current tags that SNAP 7 expects to get. Use that to update the graph provided by snap2stamps and it should be ok.

Let me know

Thank you! Now I understand what you mean before.
But I have tried running the command, and got an error message as below:
Unknown operator ‘SpectralDiversityOp’

The version of SNAP is the newest (v 7.0.3). Thank you for your help!

then run gpt -h so you can see all the operator list and then running the command I said to you last time with he correct operator name (Enhanced-Spectral-Diversity) you should get something like:

Graph XML Format:
< graph id=“someGraphId”>
< version>1.0
< node id=“someNodeId”>
< operator>Enhanced-Spectral-Diversity< /operator>
< sources>
< source>${source}< /source>
< /sources>
< parameters>
< fineWinWidthStr>string< /fineWinWidthStr>
< fineWinHeightStr>string< /fineWinHeightStr>
< fineWinAccAzimuth>string< /fineWinAccAzimuth>
< fineWinAccRange>string< /fineWinAccRange>
< fineWinOversampling>string< /fineWinOversampling>
< xCorrThreshold>double< /xCorrThreshold>
< cohThreshold>double< /cohThreshold>
int< /numBlocksPerOverlap>
< useSuppliedRangeShift>boolean< /useSuppliedRangeShift>
< overallRangeShift>double< /overallRangeShift>
< useSuppliedAzimuthShift>boolean< /useSuppliedAzimuthShift>
< overallAzimuthShift>double< /overallAzimuthShift>
< /parameters>
< /node>
< /graph>

You should get these parameter list (remove any whitespace character from here) and substitute them in the graph, using some default parameters similar to the ones are now in the snap2stamps graph.
Let me know

1 Like

Sorry for reply so late.

Because my study area is over 1,500 square kilometre and have 50 images, it take a long time to processed them.

I solved the ESD error with your help!
But got three kinds of error messages during coregistration and interferogram generation for a part of images.

  1. Error: Cannot construct DataBuffer
  2. Error: Java heap space
  3. Error: GC overhead limit exceeded

I think these error were all about the memory? I have 24GB RAM, and I set the CACHE = 16G. Isn’t it enough? Thank you!

1,500 km² is roughly 40x40 km - this should be manageable somehow. The AOI defined in the config file determines the required bursts. Sometimes, a slightly smaller AOI results in one or even two bursts less, depending how much is covered.

In my example, the red AOI requires the processing of 5 bursts, while the green AOI only covers 3 bursts. It is worth comparing if you can slightly reduce the AOI size to reduce the amount of data to be processed by 40%,

Besides that, 16 GB is not very much, especially the coregistration takes lots of memory. The second limitation is the reading/writing speed. If you have a computer with an SSD, you can move your data and process the data there. To my experience, this speeds up the processing. I am not sure if this solves the data buffer error however.

Thank you!
I will check my AOI and try to reduce the size for less bursts.

I don’t have SSD, but I could increase the CACHE to 19G or 20G which may solve these problems?
Or I need not only increase the CACHE but also the RAM in my computer first?

Thank you very very much!

Hi @mdelgado @ABraun

I check my ifg results and found that the area is much bigger than the AOI area I have set in project.conf.
(It is strange, because the result area was agreed with the AOI set in project.conf when I processed only one burst before.)

So, I back to check the coreg_ifg_computation_subset.xml, and opened by graph builder. Got an error message as follow:


And there have no subset operator, is that right?

In addition, I need to change the path of graphxml parameter from GRAPH+’/coreg_ifg_computation.xml’ to GRAPH+’/coreg_ifg_computation_subset.xml’in coreg_ifg_topsar.py?

image

Thanks for your help!

Make sure your prepared master is also smaller (same number of bursts as expected by the AOI).

But yes, the script with the subset is the one which should be called. Do you use the latest version of snap2stamps from GitHub?

No, I didn’t prepared master with the same number of burst s as expected by the AOI. I will modify it and try again.

I think I use the latest version of snap2stamps the version is v1.0.1. And I download again from GitHub. The subset operator still not exist in coreg_ifg_computation_subset.xml.

this xml is actually stored invthe graphs folder of the downloaded package.