Thermal diffusivity of peat, quartz sand, and their mixtures was studied at different water contents using the unsteady-state method. Quartz content in studied samples was 0.00 m3m−3 (pure peat), 0.05, 0.10, 0.15, 0.20, 0.30, 0.40, 0.50, 0.55, and 0.62 m3m−3 (pure sand). Thermal diffusivity of air-dry samples varied from 0.6×10−7 m2s−1 for pure peat to 7.0×10−7 m2s−1 for pure sand. Thermal diffusivity of moist samples grew by two-four times as compared to dry samples. Small amounts of quartz material with separate sand particles distributed within the peat didn’t contribute much to the heat transfer through the studied media, and the thermal diffusivity of mixtures with quartz contents of 0.05 and 0.10 m3m−3 was practically the same as that for pure peat. Vice versa, there were pronounced differences in thermal diffusivities between the samples with quartz contents of 0.55 and 0.62 m3m−3. It means that addition of small amounts of peat to the sand material results in significant lessening of thermal diffusivity. We assume that there is a kind of threshold between the quartz contents of 0.10 and 0.15 m3m−3, after which the continuous sandy chains are formed within the peat, which can serve as preferential paths of heat transport.
This work provides parameters of the soil thermal diffusivity vs. water content function for eight textural classes. The experimental thermal diffusivity vs. water content curves previously obtained for mineral soils of the European Russia were used. In the course of the study, the dataset for 77 undisturbed soil samples was analyzed. This dataset included loose sand, dense sand, sandy loam, light loam, medium loam, heavy loam, light clay, and medium clay. Thermal diffusivity of soil samples measured using the unsteady-state method varied within the dataset from 0.77 × 10–7 to 10.09 × 10–7 m2/s. To parameterize the soil thermal diffusivity vs. water content dependences the grouping method was applied. Soils were grouped according to textural classes using the classification either after Dolgov or after Kachinskii. For each of the chosen textural classes, the parameters of an average thermal diffusivity vs. water content curve were estimated from all the experimental data points for this class by approximating these points with the previously suggested four-parameter function. The approximation accuracy was estimated using Willmott’s index of agreement between the model-predicted curve and the observed values. The greatest value for the index of agreement (0.845) was obtained for the medium clays defined according to Dolgov and the smallest one (0.532) for sandy loams. The index of agreement for the whole dataset was 0.699 when the textural classes were defined according to Dolgov, and 0.688 when they were defined according to Kachinskii. In addition, the parameters of an average general curve for the whole experimental dataset were obtained. The index of agreement between this general curve and observed values was only 0.554, which confirms the efficiency of soil grouping by textural classes.
The thermal diffusivity of sieved quarry sand with a predominance of a fraction of 0.05–0.25 mm, lowland packed peat, and their mixtures was studied. The sand was mixed with the peat in various proportions; the peat content in mixtures ranged from 1 to 80% by dry weight. Metal cylinders of 10 cm in height and 3.8 cm in diameter were filled with sand, peat, and peat-sand mixtures. Thermal diffusivity was measured in laboratory by the unsteady-state method with a working temperature range of 20–26°C. The heating rate of the packed samples was measured after the samples were placed in a liquid-bath thermostat with a constant water temperature. For each sample, a series of measurements was carried out with a step-by-step change in the water content from the maximum one after capillary saturation of the sample to the minimum one corresponding to the air-dry state. The thermal diffusivity vs. water content dependence was almost linear for peat; for sand, it was a curve with a maximum. The lowest thermal diffusivity value was obtained for peat and mixtures with low sand contents; the highest one, for pure sand. Within the studied range of water contents, the thermal diffusivity of different samples changed by a factor of 1.3–2.8. The nonlinear character of the thermal diffusivity vs. peat content dependence was revealed. Small additions of peat to sand resulted in a noticeable decrease in the thermal diffusivity of the mixture; small additions of sand to peat had practically no effect on thermal diffusivity. The thermal diffusivity of the studied substrates increased with increasing bulk density and sand content of the sample and decreased with increasing organic matter content.
A series of models has been proposed for estimating thermal diffusivity of soils at different water contents. Models have been trained on 49 soil samples with the texture range from sands to silty...