I’ve noticed that the MAJA output for L2A is different compared to previous S2 output, L8 remains the same. What this means, is when calling individual processors with the python scripts, is that there are bugs in the naming convention (folder starts with SENTINEL2 instead of S2, etc).
Is this a known problem? I’ve edited the python scripts for L3A and L3B to allow them to run, but I’m sure they are not perfect because the outputs have Tnone for the tile id, which probably isn’t correct. But the scripts run, and outputs are generated.
Are you aware of this change in output and is there any plans to update the python processor scripts?
Small side note: are you aware of the limited bit depth of the L8 output from MAJA? Seems like it is crunching the final values into a int16 with a range from 0 to approximately 50 or less. This works out to be less than 5bits of information. In contrast S2 outputs range from 0 to 2500-5000, which makes sense because they are a floating point number between 0 and 1 scaled to int16 between 0 and 10000. Just wondering if anyone knows why the L8 is not in similar ranges to utilize more of the 16bit range.