Evaluation of Methods for Determining the Apparent Thermal Diffusivity of Soil Near the Surface

Abstract

AbstractField‐measured values of soil temperature were used to calculate the apparent thermal diffusivity of the upper 10 cm of soil with six different methods. The limitations of the six methods were analyzed both in terms of the calculated results, and for the quantity and quality of data required to make the calculations. Four of the six methods, Amplitude, Phase, Arctangent, and Logarithm, provided explicit equations for the thermal diffusivity. These explicit methods required only a few measurements of temperature, and calculations were simple to perform; however, the results were found to be erratic and in general inconsistent with known or more reliable estimates of the apparent thermal diffusivity. Two methods, Numerical and Harmonic, which made use of larger numbers of temperature measurements to implicitly solve for the apparent thermal diffusivity, generally provided more reliable estimates. Calculated values of the apparent thermal diffusivity by both methods were used in predicting soil temperature for comparison with measured temperature. Even under partly cloudy conditions both methods predicted temperatures very well. In general the data requirement of the Numerical method was 12 to 24 measures of temperature per day at three depths, while the Harmonic method only required 8 to 12 measures of temperature per day at two depths.