I have now used C2RCC on Sentinel 2 A and B data a couple of times, and my process is:
-Resample everything to either 20 or 30m resolution
-Run C2RCC, getting back remote sensing reflectance
-Result: RTOA, RRS, RHOWN
The problem I am running into is that it is not returning a SWIR RRS band. Is this in the design, or am I doing something to prevent SWIR from being processed?
This is intended. Only the bands needed to derive the water constituents are computed by the AC.
This is also not easily changeable. The neural nets need to be trained again.