In the context of the land surface, the ground absorbs heat from radiation (also some times from precipitation and advection), and releases heat through evapotranspiration, advection and radiation, and run-off. If it were a closed system, you would expect the net heat loss/gain to be zero (at least on a reasonably long time scale, that irons out daily and seasonal cycles, and noise). However, there is also heat from the mantle to consider, and also heat conduction to the oceans.
Are there other avenues for heat loss or gain from the soil? Are any of these significant in a climatological context? Or should the total heat flux into (or out of) the ground from the atmosphere be equal to zero (or close enough for practical purposes, over time scales longer than a year)?