Hydraulic characterization of the variably saturated zone in peatland permafrost mires

Abstract

Peatlands acquire only 3% of the terrestrial earth's surface; however, they store up to 30% of the global soil carbon. These peatlands also dominate the northern Hemisphere, which is covered by more than 25% of permafrost. An accelerating trend in the rate of permafrost degradation due to climate warming has been observed in most regions of the Northern Hemisphere. An indicator of permafrost degradation is the active layer depth, which is situated in the variably saturated zone. The hydraulic properties of the peatland top soil layer are influenced by its unstructured porous media, high organic matter, and high porosity. The water content in the variably saturated zone in cold regions is influenced by both drying-wetting and freezing-thawing cycles. The soil water characteristic curves (SWCC) define the relationship between unfrozen water content and matric potential. The soil freezing characteristic curves (SFCC) define the relationship between unfrozen water content and temperature around 0°C. The SWCC, SFCC, and the similarities between them have been intensively investigated for mineral soils in comparison to organic soils.   The three main goals of the study for peatland permafrost mires are as follows: (i) Determine the SWCC using inverse modeling of transient evaporation experiments. (ii) Estimate the SFCC using field-based volumetric water content measurements using a simple empirical function. (iii) Compare and develop a relationship between the SWCC and SFCC.  The lowland permafrost mires in the Abisko region, located in the northern part of Sweden, were investigated. For the SWCC evaporation experiments, 12 soil samples at six locations and two depths (10 and 25 cm) were taken. Inverse numerical modeling was used to fit and compare the three pressure-saturation functions: (i) Van Genuchten model. (ii) Peter Durner Iden (PDI)-variant of the Van Genuchten model (iii) PDI-variant of the bimodal van Genuchten model. The goodness of fit was checked by Root Mean Square Error and Akaike Information Criterion. It was observed that the PDI-variant of the bimodal van Genuchten model was most suitable for all the soil samples. 12 soil moisture sensors were also installed at the six locations and five depths (10 to 50 cm). An exponential logarithmic function with two parameters (transition temperature and temperature dispersion) was fitted to individual freezing and thawing curves from the soil moisture sensor data. The function showed a very good fit, and it was observed that the two fitting parameters were higher for thawing curves compared to freezing curves. A new SFCC function was developed based on the PDI variant of the bimodal van Genuchten model. This function was compared with the fitted SFCC logarithmic function. Reasonable differences were identified, which could be attributed to the field-installed soil moisture sensors and laboratory-conducted evaporation experiments. It is one of the few hydrological studies that has investigated the effects of bimodal behavior in organic soils on soil freezing and thawing. The measured parameters and datasets provide the necessary functions for developing cryohydrogeological models. The cryohydrogeological models can be used to assess the impacts of climate change on permafrost.