C2RCC problem for MSI granules processing

[Sent]

Hello,

I’ve been trying to process multiple S2 images by calling the gpt command “c2rcc.msi” through the command line via an .XML file in Unix machine. However, it process only the 40% of the images without returning errors.

I’ll let here the XML file I’m using and the output from the command line.

Can anybody knows what might be the cause?

Thanks,
Giulia

[Sent]

This the output from the terminal. Here it stopped the processing after 30% of the image, and in fact when I open the image via SNAP only the upper part of the image is processed.

[marpet]

Hi Giulia,

I guess you have the cloud shadow option enabled.
There is currently an issue:

Issue with IdePix cloud shadow detection

That IdePix hangs if cloud shadow detection is enabled is a known issue (IdePix MSI hangs if Cloud Shadow computation is enabled · Issue #25) and has been fixed already on the development branch. The fix will be included in the next update of IdePix. But I can’t give you a specific schedule right now.

[Sent]

Hi Marpet,

Thank you for your fast reply. Unfortunaly I think the cloud shadow option in not my case because I’m not using IdePix. Also, It’s strange that every time the processing stop at a different percentage of the image even if I don’t change anything.
I’ll leave here my XML file. c2rcc_msi_v2_algo.xml (4.3 KB)

[marpet]

Oh, yes you are right. You are using C2RCC. Sorry!

I think in your case it would be good to split the graph.
Do in one step only the resampling and the regional subsetting and then do in a second step the rest.
It could be that you are at the limit of your computer capabilities. How much memory do you have?
At list it will help to find out where, at which node, your graph hangs.
Another option is to use the general resampling and not the S2Resampling. It is faster and, it uses less memory, I think. I’m not sure how sensitive C2RCC is regarding the differences in the geometry bands and if you can use it.

[Sent]

Yes, splitting the processing chain in two actually worked! I am using 16 Gb of memory and apparently is not enough for the full processing.

Thank you for your help!

Giulia

[binbinjiang]

Dear all,

how could i use that .xml run for my sentinel2 data . Could you show me the command line? thanks

Best

Binbin

[ABraun]

If you want to automate your processing with the command line you have to invest some time in understanding the principles. Especially because these things are very sensitive to your individual files, their location, product level etc, there is little sense in copy-and-paste commands and expecting them to work.

Marpet and the others did a good job helping you to get on the right track in the other topic, so please specify your question and take some time to build your workflow step by step. Start simple and only edit/add something once it worked. Putting too much in at once makes it hard to find errors.

[abruescas]

There are a few threads in this forum you could check:

C2RCC OLCI in GraphBuilder

Another question… just want to make sure it’s working correctly. The ancillary data are stored in nested directories first by year, then julian day. In one Julian day folder, there are 3 files- 1 AURAOMI_24h.hdf and 2 MET_NCEPR2_6h.hdf.bz2. Is the code going through all of these and matching the OLCI dates with the correct ancillary data date?

How to Process multiple files using C2RCC

Please, read the link marpet sent about batch processing with GPT and make a more concrete question about the xml or the command line order.

I am happy to help with specific questions.