C2RCC Normalised Water Leaving Reflectance (rhown) returns 0 values for band 7 & 8A

After running in batch mode C2RCC & C2X over a 100 Sentinel2 scenes, choosing rhown (Normalised Water Leaving Reflectance) as one of the outputs I noticed that for Band 7 & 8A I was always getting a value of zero whereas the rtoa, rrs and rhow are giving normal values. I find it a bit weird, is it a bug or I am missing something?

Thanks a lot!

This is indeed a little bug. The bands rhown_7 and rhown_8A should not be part of the resulting product. They are not computed by the neural nets. In the upcoming release they will not be included anymore.

Thanks for the report.

Thanks for the reply,


Is the C2X already available for snap? If so - how to include it?

C2RCC will be available with SNAP 6. C2X is an option within C2RCC.
Right now you need to build C2RCC on your own for SNAP 5.
It is explained at the Github repository.

Hi Marco,
thanks for your quick reply.
The reason why I couldnt find it, is because i used C2RCC only for OLCI yet - and it seems its not available for OLCI.
Am I right? In my version, I dont have an option to choose another NN as I have for other Sensors like MERIS etc

No there is no extreme net for OLCI. As far as I remember the standard net for OLCI can handle higher values as the nets for the MERIS or S2 MSI

Im new to SNAP. I tried C2RCC for processing S2 images in SNAP 6. I’m wondering if I can use this model for correction over very turbid water and use the results in publication? Please can u share any detail doc about C2RCC processing.

Please see the other thread.