Thermal conductivity of unfrozen and partially frozen managed peat soils

Detailed, accurate information on soil temperature is crucial for understanding processes leading to solute leaching and greenhouse gas (GHG) emissions from managed peat soils, but few studies have attempted to study these processes in detail. Drained peat soils have different characteristics from pristine peat. Cultivated peat soils, in particular, have high mineral matter content in the plough layer, due to mineralisation of peat and, sometimes, addition of mineral material. This study examined the effect of mineral matter content on thermal conductivity (λ) in partially frozen and unfrozen peat samples. Effect of change in temperature from −3 °C to −10 °C on thermal conductivity was also estimated. Three existing models for estimating the thermal conductivity of organic soils were assessed for their suitability for cultivated drained peat soils. The thermal conductivity of peat samples with three different levels of mineral matter content was determined, using the single probe method, in the saturated state and when subjected to at least two different matric potentials at five different temperatures (+10 °C, + 1 °C, −3 °C, −5 °C and −10 °C). The results showed that λ values differed between peat soils depending on mineral matter content, ice content and moisture content. The samples with the highest mineral matter content and bulk density had higher thermal conductivity at positive temperatures and to a lesser extent, at freezing temperatures, when volumetric water content and volume of water-free pores was similar. Most soil samples, especially those with no added mineral soil, were not fully frozen at −3 °C and −5 °C, but this had minor effect on thermal conductivity compared with values measured at −10 °C. The Brovka-Rovdan model proved reasonably good at predicting frozen thermal conductivity in sand-enriched peat soils, while the de Vries model proved best at estimating thermal conductivity for unfrozen peat samples. We provide a first estimate of the thermal conductivity of (partially) frozen cultivated peat measured using undisturbed samples. These results can be used to parameterise numerical heat transport models for simulating soil processes and GHG emissions.