Here are the things that I recall that need to be dealt with:
- Remove radiometer and open-path data (li7500, krypton, ec150) during cleaning events
- Try to recover odd Tsoil.grass data for about a week in early Oct. when scrambled. (Already have code to do this, but might be improved.)
- Remove TRH data when iFan is zero
- Adjust some early TRH data from probes that were swapped based on postcals
- Determine which set of boom angles to use
- Swap Cactus/Grass data from first day
- Determine what to do with Qsoil.c, Qsoil.c2 not matching. Qsoil.c data (when available) are self-consistent from the initial data, even through the lightning. Qsoil.c2 is always lower, but data were reasonable compared to one gravimetric sample. We will measure probe depths to see if this helps explain the differences. (Fix Qsoil probe that dies when cold.)