Lab-Scale Investigation of Multi-dimensional Relationships between Soil Intrinsic Properties to Improve Estimation of Soil Organic and Ice Content using Novel Core Imaging and Geophysical Techniques in Arctic Tundra

Abstract:

Shallow permafrost distribution and characteristics are important for predicting ecosystem feedbacks to a changing climate over decadal to century timescales. These can drive active layer deepening and land surface deformation, which in turn can significantly affect hydrological and biogeochemical responses, including greenhouse gas dynamics. Investigating permafrost soil intrinsic properties generally involves time-consuming and expensive lab-based analysis of few soil cores over a large area and extrapolating between points to characterize spatial variations in soil properties. Geophysical techniques provide lower resolution data over a spatially large area and when coupled with high-resolution point data can potentially estimate with greater accuracy the spatial variation of investigated properties, thus limiting the difficulty of collecting many soil cores in remote areas.

As part of the Next-Generation Ecosystem Experiment (NGEE-Arctic), we investigate multi-dimensional relationships between various permafrost intrinsic soil properties, and further linkages with geophysical parameters such as density from X-ray computed tomography (CT) and electrical conductivity from electrical resistance tomography (ERT) to evaluate how best to constrain estimation of properties as soil organic carbon content, ice content and saturation across low- to high-centered polygon features in the arctic tundra.

Results of this study enable the quantification of the multi-dimensional relationships between intrinsic properties, which can be further used to constrain estimation of such properties from geophysical data and/or where limited core-based information is available. This study also enables the identification of the key controls on soil electrical resistivity and density at the investigated permafrost site, including salinity, porosity, water content, ice content, soil organic matter, and lithological properties. Overall, inferred multi-dimensional relationships and related uncertainty enable probabilistic mapping of key parameters (organic content, ice content, etc.) using density and soil structural information from CT and bulk electrical resistivity.