Introduction
For several decades, geophysical techniques have been successfully applied to the characterization of ice properties in situ, and this work has contributed significantly toward our fundamental understanding of glacier structure, composition and dynamics. For example, borehole-inclination logs (e.g. Reference Hooke, Pohjola, Jansson and KohlerHooke and others, 1992) and echo-sounding techniques (e.g. Reference Jacobel and RaymondJacobel and Raymond, 1984) have provided information regarding mass balance and englacial deformation. Ground-penetrating radar (Reference Murray, Gooch and StuartMurray and others, 1997) has identified thrust faults, and cross-hole electrical resistivity experiments (Reference Hubbard, Binley, Slater, Middleton and KulessaHubbard and others, 1998) have imaged drainage features.
In this paper, we present data generated from an acoustic televiewer and recorded in two boreholes located in a temperate glacier. This instrument produces a magnetically oriented image of the borehole wall and is routinely used in petroleum, mining and ground-water investigations to locate fractures, identify lithologic sequences and assess well-bore stability conditions. In this particular application, the televiewer was deployed in water-filled boreholes drilled into ice, and the resulting data recorded in situ enable us to evaluate the capabilities and limitations of this tool when applied in a glacial environment.
Setting
Continuous ice cores were recovered from two boreholes penetrating Upper Fremont Glacier in the Wind River Range of west-central Wyoming, U.S.A. The study area is situated within the accumulation zone of this mid-latitude temperate glacier at about 4000 m a.s.l. The holes were located near the presumed flow center line (Fig. 1), and cores were collected, processed and analyzed for paleoenviron-mental studies (e.g. Reference Cecil, Green, Vogt, Michel and CottrellCecil and others, 1998; Reference Schuster, White, Naftz and CecilSchuster and others, 2000). Ice thickness ranges from about 60 to 172 m in the upper half of the glacier, and the ice flow is toward the northeast (Reference Naftz and SmithNaftz and Smith, 1993).
Locations of study site and boreholes (adapted from Reference Naftz and SmithNaftz and Smith, 1993).
The two boreholes were produced by a thermal drill (Reference Holdsworth, Kuivinen and RandHoldsworth and others, 1984), with a penetration rate of approximately 6 m h−1. This device consists of a non-rotating core barrel that is heated at the bottom. An annulus of ice is melted away by contact with this warm cylinder, and an ice core is collected within the hollow barrel. Water level in each hole was about 3 m below the snow surface, and ice was encountered at roughly 2 m depth. Hole DH98-3 was the first borehole to be cored. After reaching 49 m depth and returning to the surface to retrieve the ice core, the drillbit was deflected to one side on its subsequent lowering and would not follow the existing hole. Repeated efforts to return to the bottom were unsuccessful, and this hole finally had to be abandoned. A second hole, DH98-4, was initiated and advanced to the base of the glacier, extending to a total depth of 162 m. The two holes were approximately 10 cm in diameter and about 35 m apart. Ice cores were described, cut, and stored in 1 m sections for future laboratory-based analyses.
Acoustic Televiewer Logging
Equipment and principle of operation
Although the main scientific objective of this project was to extract a detailed isotopic record from the ice, the boreholes also afforded an opportunity to deploy an acoustic televiewer (Reference Zemanek, Glenn, Norton and CaldwellZemanek and others, 1970) in a glacial environment and evaluate its performance. This logging tool is 5 cm in diameter by 193 cm long and is connected by means of a standard four-conductor cable to an electric winch (Fig. 2). During logging operations, it is lowered or raised in the borehole at a trolling speed of about 1.5 m min−1, and data are transmitted through the logging cable to a computer at the surface. This instrument is equipped with a 500 kHz transducer that rotates at 12 revolutions per second and emits 256 pulses per revolution. The signals are transmitted through the borehole fluid, reflect off of the fluid–ice interface, and return to the tool where acoustic amplitude and transit time are recorded digitally. It should be noted that these high-frequency signals do not penetrate into the ice, and consequently do not provide any view of conditions beyond the borehole; they reflect off of the fluid-ice interface with the intent of developing an acoustic image of the borehole wall. Top and bottom centralizer springs attached to the instrument help to guide it in the borehole and minimize any systematic offsets in transit times produced by a decentralized tool.
Photograph of acoustic televiewer system deployed at study site, showing tool, cable, electric winch and thermal drill.
The televiewer’s principle of operation relies on a sharp reflection of acoustic energy at the fluid–wall interface. It is commonly used in boreholes drilled through rocks, where the acoustic impedance contrast between the borehole fluid and surrounding material produces a reflection coefficient on the order of 0.60. In a glacier, the acoustic impedance contrast between ice and water is less dramatic, and the resulting reflection coefficient is reduced by about half.
Field data
Acoustic amplitudes are converted into brightness (gray scale) or color, and the resulting image appears as a planar, “unwrapped” representation of a cylindrical surface (see left panels in Fig. 3). Acoustic transit times are also converted into a brightness scale for visual display (see right panels in Fig. 3), though these latter data contain spatial information that can be further processed to obtain radial distances and delineate cross-sectional borehole shapes. Non-horizontal planes that intersect these boreholes appear as sinusoids in the unwrapped images (Fig. 3). The particular orientations of these planes can be computed by a simple geometric exercise, where dip angle is proportional to the sinusoid amplitude and strike corresponds to the azimuth of its lowest point (Fig. 4).
Depiction of magnetically oriented, acoustic televiewer data obtained from (a) well in sedimentary rocks of the New York Basin, Pennsylvania, U.S.A. (from Reference Morin, Senior and DeckerMorin and others, 2000), and (b) hole DH98-3 in this glacier study. Unwrapped, planar images of borehole wall generated from acoustic amplitude data (left panels) and acoustic transit-time data (right panels) are shown. Outlined sinusoid locates a steeply dipping (63°) fracture in ice striking at N25° W
Diagram illustrating method for determining strike and dip of intersecting feature observed in televiewer image from depth scale and magnetic orientation.
Because of the deterioration in acoustic reflection coefficient associated with glacier applications, the processed images are not as crisp and clear as those typically generated in rock environments. This is illustrated in Figure 3, where televiewer images recorded in sedimentary rocks (Fig. 3a) can be compared to those obtained in hole DH98-3 (Fig. 3b). Steeply dipping features are observed in both media, though the view obtained in rocks is of better quality and exhibits more detail. Regardless of this inherent deficiency, the televiewer logs obtained at this site do detect englacial features. However, their exact physical interpretation may be problematic, likely requiring supplemental information from inspection of cores.
The planar feature intersecting hole DH98-3 at a depth of about 16 m dips to the east-northeast at 63° and has a strike of N25° W (Fig. 3b). Lower-hemisphere stereographic plots of all features identified in the two boreholes are presented in Figure 5. These data contain information that may be related to englacial structure or stress conditions at this study site (e.g. Reference Sharp, Lawson and AndersonSharp and others, 1988). The diagrams identify the presence of conjugate, high-angle features (> 45°) with orientations that are consistent for both boreholes and shallow dipping features (< 30°) that diverge significantly.
Lower-hemisphere stereographic plots of features intersecting (a) hole DH98-3 and (b) hole DH98-4. C.I. is the contour interval (%), and N is the total number of features.
Because the televiewer emits 256 acoustic pulses per revolution of its internal transducer, the travel-time data, such as those depicted in the right panels of Figure 3, can be converted to 256 measurements of radial distances as a function of azimuth; cross-sectional views of the borehole can subsequently be constructed. In Figure 6a, a circular, intact borehole is delineated at 39.8 m depth in hole DH98-4, whereas a damaged and enlarged borehole geometry is identified 6 m deeper (Fig. 6b). Gaps in the cross-section indicate no return of acoustic signal. This type of information may be useful in inferring stress conditions by identifying localized zones of brittle failure such as borehole breakouts (e.g. Reference Zoback, Moos, Mastin and AndersonZoback and others, 1985) or by monitoring creep processes through borehole deformation (Reference PatersonPaterson, 1977). These polar views of the borehole can also be stacked to produce three-dimensional cylindrical projections. For example, the layers depicted in the televiewer image from Figure 7a are alternating bands of bubbly ice and clear ice, as confirmed by a photograph of the corresponding core section (Fig. 7b).
Magnetically oriented cross-sections of borehole shapes recorded in hole DH98-4. Images depict (a) a circular, intact borehole at 39.8 m depth, and (b) an enlarged borehole at 45.8 m. Gaps indicate no return of acoustic signal.
(a) Cylindrical projection of borehole (hole DH98-3) developed from combination of amplitude and transit-time data showing alternating bands of clear and bubbly ice. (b) Photograph of corresponding core section.
In order for the acoustic images to be magnetically oriented, the televiewer is equipped with a three-component magnetometer. It also contains a two-component inclinometer, and the combination of these two instruments permits a continuous borehole-inclination log to be recorded during logging operations. This capability is incorporated into the tool in order to enable the computed dip angles of intersecting features to be corrected for the vertical deviation of the borehole. Inclination and azimuthal bearing are continuously measured at 1 cm depth intervals, and a cumulative plot of the hole trajectory is constructed from these data. Deviation angles ranged from 2° to 5° in hole DH98-4, and its spatial coordinates are represented by the polar plot of Figure 8 with depth markers at 8 m increments. This diagram indicates that this hole drifts to the south-southwest and is displaced horizontally by about 4.2 m at 78 m depth.
Polar diagram of inclination log for hole DH98-4.
Concluding Remarks
To our knowledge, this work marks the first attempt to log boreholes in ice with an acoustic televiewer. Numerous planar features intersecting the boreholes were identified by inspection of the digital images. Their depths and orientations were determined, and this information can yield insight into englacial structure with respect to depth. However, images generated in ice were not as visually crisp as those typically recorded in rocks, because of the lower reflection coefficient between water and ice. Therefore, although the features identified in the logs were found to be fractures, contacts between clear and bubbly ice, and/or localized zones of borehole enlargement, they could not be distinguished solely from the acoustic images. It was necessary to supplement these data with core descriptions in order to confidently interpret their physical significance.
This tool can only be deployed in fluid-filled boreholes; it will not function in air-filled ones. Obviously, video logs (e.g. Reference PohjolaPohjola, 1994) or optical televiewer logs are better suited for this latter condition. The acoustic tool offers the advantage over optical methods of recording radial dimensions as a function of azimuth from which cross-sectional views of borehole geometry can be constructed. This capability, as well as the concurrent inclination log, may be useful in studies designed to investigate ice-deformation processes.
Regrettably, the particular tool employed in this study was not adequately insulated to withstand the cold temperatures encountered in the boreholes and it failed to function properly after about 1 h of sustained immersion. This timeframe was sufficient to log the shallow hole, but was not long enough to completely log the deeper hole. Consequently, only about half of the deeper hole (from the surface to 78 m depth) was investigated. Thus, unless boreholes are shallow enough to be logged in about 1 h, improvements in insulation will be required to use this tool in glaciers.
Acknowledgements
Funding for this project was provided through an interagency agreement between the U.S. Department of Energy and the U.S. Geological Survey, No. DE-A107-98ID13598. Drilling services were provided by the Polar Ice Coring Office, Lincoln, Nebraska.
