So far, I haven't noticed much correlation between Qsoil and the TP01-derived heat capacity, except for one event at hiland3. Presumably, the soil is still "healing" from the probe installation. I'm pretty sure that I didn't pack the soil around the probes to the density of the undisturbed soil. However, there has been a fair amount of precip, both liquid and frozen, since installation, so they should have healed a bit.
Unfortunately, temperatures are now cold, so the soils are close to freezing. Tsoils are now in the range 0--2, though none have gone to below 0 yet. When they do, we should start seeing Qsoil values drop as the water becomes invisible to the EC-5 sensors. I suggest using the Qsoil values just prior to this drop as the actual soil moisture amount for the rest of the experiment, assuming that water in the soil is immobile once frozen. This would ignore any melted snow that might penetrate the soil.