Hi,
When processing Landsat data, I realised that c2rcc_flags bits 2 and 3 (= values 2 (Rtosa_OOR) and 4 (Rhow_OOR)) are always set to 1 (even on pixels outside the satellite swath). Is it sth expected / on purpose (it would be weird, though + it’s not the case for Sentinel-2/3)? 
PS: I processed multiple satellite products with different SNAP versions.
Hi @Marco_EOM,
1/ Is it a known issue? and 2/ do you know if a fix is likely to be released?
Best,
Arthur
This is probably an issue with the neural net. It defines the valid ranges but it seems something is wrong with the limits. Better you ignore these two flags for Landsat.
It is unlikely that there will be an update on this issue soon.
Ok, this is what I thought… Thanks anyway!