Submitted to: Agronomy Abstracts
Publication Type: Abstract Only
Publication Acceptance Date: July 5, 2000
Publication Date: N/A
Technical Abstract: Soil thermal conductivity as a function of water content is needed for mechanistic models of the soil heat and water fluxes. We measured soil water content using time domain reflectometry and soil temperature using thermocouples at discrete distances below the soil surface (2.5, 5, 10, 15, and 25 cm) in three major soils of the Southern Great Plains, the Pullman silty clay loam (fine, mixed, superactive, thermic Torrertic Paleustoll), the Ulysses silt loam (fine-silty, mixed, mesic Aridic Haplustoll), and the Amarillo fine sandy loam (fine-loamy, mixed, superactive, thermic Aridic Paleustalf). Three replicate sets of measurements were made in each soil. For each vertically adjacent pair of depths, the diurnal temperature data for the upper layer were fit to a six-term Fourier series using a non-linear Levenberg-Marquardt regression routine. The fitted Fourier series coefficients were then used in a six-term solution of the Fourier heat law for one-dimensional heat flux, which had the thermal diffusivity of the intervening layer as the sole unknown parameter. Nonlinear regression was then used to fit the solution to the temperature data from the lower layer, thus solving for the thermal diffusivity. The mean water content of the intervening layer was used to calculate the layer's heat capacity, which was used in calculating the thermal conductivity of the layer.