Geophysical Techniques applied to permafrost investigations in Greenland

The presence and distribution of permafrost is a major concern in engineering related problems in arctic environments. The interstitial ice content of permafrozen sediments, which may even be in excess of the unfrozen pore volume, constitutes a risk of loss of strength and settling upon permafrost degradation, that may for instance be induced by construction work.

Permafrost distribution also influences the flow in and recharge to aquifers and therefore determines whether aquifers may be exploited for abstraction purposes. This thesis is concerned with the application of geophysical techniques to permafrost investigations in West Greenland and consists of three main parts.

The first part reports an integrated geophysical survey conducted near the town of Sisimiut. The second and third parts focus on theoretical studies and field tests of a ground based, multifrequency frequency-domain electromagnetic (FDEM) method and the complex resistivity (CR) technique, and evaluations of their applicability to permafrost investigations.

Two permafrost areas close to Sisimiut, West Greenland, have been surveyed with a range of geophysical techniques comprising DC resistivity, induced polarization (IP), VLF-R, ground penetrating radar (GPR) and seismic refraction methods. Especially the combination of resistivity, IP and GPR measurements proves powerful in determining the lateral distribution of frozen sediments and the depth to the frost table.

Indications of frozen ground are high resistivity, low normalized chargeability and a strong reflector at the frost table. Area 1 is found to contain one large highly resistive frozen body, area 2 has several frozen areas separated by unfrozen areas (taliks) caused by the presence of water bodies. Although many aspects of the 2D and 3D permafrost distribution could be clarified using multi-electrode resistivity and IP profiles, borehole information shows that the estimation of the thickness of frozen sediments was grossly overestimated in area one.

Furthermore areas of high ground ice contents are not uniquely identified, and other methods are therefore needed to improve reliability of such interpretations. The ground based, small coil, multifrequency FDEM technique is a fast and lightweight surveying method, requiring no electrical contact with the ground, and is thus potentially applicable to mapping both vertical and lateral resistivity distribution year round at low cost.

A theoretical study of misorientation effects in the response of a HCP configuration over a homogeneous half-space shows that both the real and imaginary parts may be severely corrupted even at small coil axes misalignments. With few exceptions, extreme errors increase with increasing axis inclination and are highly dependent on the direction of misorientation.

Six different apparent resistivity definitions are studied, but none of them are robust in all the studied situations. In general the modeling shows that currently available equipment cannot be expected to yield sufficiently accurate results to allow quantitative 1D interpretation even when applied in less hostile environments.

Application of the method at the Sisimiut field site confirms that the method is not well suited for permafrost studies in Greenland due to misalignment effects and limited responses of the resistive environment. Theory and laboratory studies indicate that the complex resistivity dispersions of geological materials may be depending on temperature and ice content.

A wide frequency bandwidth CR survey has therefore been conducted at the Sisimiut field site to evaluate the applicability of the method to permafrost mapping. The results suggest that a characteristic spectral response is observed in permafrozen areas, but the collected data seem to be obscured by errors possibly relating to electromagnetic and capacitive coupling effects as well as equipment related issues.

New modeling tools were developed to address these effects, and the models show that keeping electrode contact resistance low is crucial to avoiding capacitive coupling problems. Equipment tests showed that the current reference link introduces errors in the measured reference signal. Improvements have been made to the system, and a subsequent field test resulted in data that could be reasonably well fitted by geologically sensible 1D models.

If the observed permafrost responses can be confirmed with the improved system, the CR technique may prove valuable in future permafrost investigations.