I have glanced at Tsoil and Qsoil values.  It seems that after 8 Jan or so the ground started freezing and reported Qsoil values started becoming smaller (since the EC-5 probes don't respond to the dielectric value of ice, only liquid water).  At many of the sites, there actually is an interesting large diurnal cycle to Qsoil as water melts and later freezes.  In any case, I think gravimetric sampling to compare to the EC-5 Qsoil measurements is hopeless until Tsoil values return to be positive.

It does bring up the interesting question of whether we should continue gravimetric samples to actually measure Qsoil, now that the EC-5 data is hard to interpret.  I would ask the PIs.  In the past, I've argued that frozen soil moisture isn't going anywhere, i.e. evaporating, so it didn't really matter what Qsoil was.  However, in this melt/freeze situation, it is more complicated.

BTW, the values we have been using are volumetric fraction, mislabeled as volumetric %.  I've changed the config, but after I started a recalculation of noqc_instrument.  noqc_geo should be correctly labeled.