Exploring the challenges of interpreting near-surface towed TEM data on saline permafrost

• Published in: GEOPHYSICS
• Main Authors: Lorentzen, Thomas Højland, Kass, Mason Andrew, Scheer, Johanna, Tomaškovičová, Soňa, Christiansen, Anders Vest, Maury, Pradip Kumar, Ingeman-Nielsen, Thomas
• Format: Article in Journal/Newspaper
• Language: English
• Published: 2024
• Subjects: Induced polarization (IP); Electromagnetics; Interpretation; Inversion; Permafrost; Greenland; Ilulissat; Ice; permafrost
• Online Access: https://orbit.dtu.dk/en/publications/9c2f38b2-7423-4956-9394-0d228b01ea31; https://doi.org/10.1190/geo2023-0221.1; https://backend.orbit.dtu.dk/ws/files/354020259/lorentzen-et-al-2024-exploring-the-challenges-of-interpreting-near-surface-towed-tem-data-on-saline-permafrost.pdf

This paper analyzes a continuous near-surface transient electromagnetic (TEM) survey on permafrost, around Ilulissat in western Greenland, an area characterized by continuous saline permafrost. The TEM data is severely affected by induced polarization (IP) effects, causing a large range of decay shapes. We identify seven unique decay shapes: “oversteepened”, ”sign-change“, ”all-negative”, “double-sign-change”, “no apparent IP”, “flat-spot”, and “positive-nonmonotonic”, of which the last two have not previously been identified in the scientific literature. A clear spatial dependency of the decay shapes is observed. Inversion of the data is carried out using a Cole-Cole model which poses a highly nonunique inversion problem with an extreme starting model dependency. A series of inversion and forward modeling experiments demonstrate these challenges, and show that a low-resistivity and highly chargeable layer with time constant (τ ϕ ) values between 0.31⋅10 −5 and 1.30⋅10 −5 s and frequency exponent (C) values above 0.74 is needed to fit the data used in the inversion experiment. Forward modeling further shows that low τ ϕ and high C values are needed to reproduce the observed flat-spot and positive-nonmonotonic decay shapes. Based on these observations, we attribute the IP effects to the orientational polarization of ice in the soil column. This mechanism allows for low-resistivity, high-chargeability layers due to partially frozen saline sediments; a combination which is difficult to explain by the IP mechanisms traditionally considered. Based on forward modeling of a realistic structural model of a layered sediment pack over dipping bedrock, we are able to reproduce all the observed decay shapes. This model provides a consistent framework for qualitative interpretation of the entire data set, and evidence for the presence of saline deposits in the central parts of the sedimentary basins.