Studies on coupled transfer of soil moisture and heat have been widely carried out for decades. However, little work has been done on red soils, widespread in southern China. The simultaneous transfer of soil moisture and heat depends on soil physical properties and the climate conditions. Red soil is heavy clay and high content of free iron and aluminum oxide. The climate conditions are characterized by the clear four seasons and the serious seasonal drought. The great annual and diurnal air temperature differences result in significant fluctuation in soil temperature in top layer. The closed and evaporating columns experiments with red soil were conducted to simulate the coupled transfer of soil water and heat under the overlaying and opening fields’ conditions, and to analyze the effects of soil temperature gradient on the water transfer and the effects of initial soil water contents on the transfer of soil water and heat. The closed and evaporating columns were designed similarly with about 18 °C temperatures differences between the top and bottom boundary, except of the upper end closed or exposed to the air, respectively. Results showed that in the closed column, water moved towards the cold end driven by temperature gradient, while the transported water decreased with the increasing initial soil water content until the initial soil water content reached to field capacity equivalent, when almost no changes for the soil moisture profile. In the evaporating column, the net transport of soil water was simultaneously driven by evaporation and temperature gradients, and the drier soil was more influenced by temperature gradient than by evapo- ration. In drier soil, it took a longer time for the temperature to reach equilibrium, because of more net amount of transported water.
Simulation models of heat and water transport have not been rigorously tested for the red soils of southern China. Based on the theory of nonisothermal water-heat coupled transfer, a simulation model, programmed in Visual Basic 6.0, was developed to predict the coupled transfer of water and heat in hilly red soil. A series of soil column experiments for soil water and heat transfer, including soil columns with closed and evaporating top ends, were used to test the simulation model. Results showed that in the closed columns, the temporal and spatial distribution of moisture and heat could be very well predicted by the model, while in the evaporating columns, the simulated soil water contents were somewhat different from the observed ones. In the heat flow equation by Taylor and Lary (1964), the effect of soil water evaporation on the heat flow is not involved, which may be the main reason for the differences between simulated and observed results. The predicted temperatures were not in agreement with the observed one with thermal conductivities calculated by de Vries and Wierenga equations, so that it is suggested that Kh, soil heat conductivity, be multiplied by 8.0 for the first 6.5 h and by 1.2 later on. Sensitivity analysis of soil water and heat coefficients showed that the saturated hydraulic conductivity, KS, and the water diffusivity, D(θ), had great effects on soil water transport; the variation of soil porosity led to the difference of soil thermal properties, and accordingly changed temperature redistribution, which would affect water redistribution.
