Comparison of four classical algorithms to determine the apparent thermal diffusivity of heavy clay soil in field and laboratory column experiments

Abstract

Prediction of soil temperature at an acceptable level of accuracy is essential for managing soil temperature at large scales since its measurement is costly and not practical due to its high spatial and temporal variability. This study was carried out for:: 1) Calculate the coefficient of thermal conductivity using classical methods and basing on data from measured soil profile temperatures in column and field conditions and 2) compare the results of the calculated and measured soil temperature, calculated from these data for both conditions. Soil temperature was measured with thermal sensors on both soil profiles in the field and soil column. The measured temperature values were similar in soil column and profile across 0.05, 0.10, and 0.20 m depths, but had the difference for 0.30 cm. The introduction of the second harmonic allows determining with high accuracy the parameters (T0, Ti, εi) of the temperature distribution on the field and column soils surfaces. For example, respectively for the harmonics n = 1 calculated the following statistical characteristics: R2 = 86.45%, σ = 3.26, A = 13.07%, UII=0.12 and for n = 2: R2 = 98.64%, σ = 1.09, A = 4.20%, UII=0.04. It was found that, mean values for soil thermal diffusivity (k), thermal conductivity (λ), damping depth (d) and heat absorptivity (e), calculated by amplitude, arctangent and logarithm methods, for the column were greater than those for soil profile, in all the cases.