Abstract
The Rand two-level general circulation model has been integrated to compute ground surface (bare land) temperature by solving: 1) the interface heat balance equation without soil heat flux; 2) the interface heat balance equation by including parameterized soil heat flux; and 3) a prognostic equation which includes the heat capacity of the soil as well as an explicit formulation for soil heat flux.
The integrations were performed for 48 hours for the month of January. A comparison of results shows that the most realistic distribution of the ground surface temperature with respect to the amplitude, diurnal range, and the phase relationship between the ground temperature, solar radiation, and soil heat flux is given by the solution of the prognostic equation.
