Introduction

During July and August 1970, a Cambridge expedition carried out a programme of glaciological measurements and surveying on Roslin Gletscher (Fig. 1) in Slauning Alper, East Greenland.Footnote * The primary investigation was to determine the glacier ice thickness by means of radio echo sounding. Most radio echo sounding work to date has been carried out on polar ice sheets (Reference EvansEvans, 1967; Reference Evans and SmithEvans and others, 1969), and it was of interest to evaluate a similar technique for use on valley glaciers where there are the additional problems of unwanted reflections from the valley sides, and perhaps excessive absorption of the transmitted pulse in the ice, whose condition may vary over a wide range of temperatures, structure and impurity content.

Fig. 1 Roslin Gletscher. Radio echo sounding routes.

Instrumentation and Measurement Technique

It was originally intended to compare the performance of two sounding instruments operating at 150 and 440 MHz. The 150 MHz instrument was similar to the 35 MHz model used by the Scott Polar Research Institute in Antarctica (Reference Evans and SmithEvans and Smith, 1969), However, an unfortunate electrical fault prevented it from being used, and so all work was done at 440 MHz using a modified SCR 718 radio altimeter. The range and accuracy of the SCR 718 have been discussed by Reference Weber and AndrieuxWeber and Andrieux (1970). The internal power supply was altered so that it could be operated from 230 V a.c. A camera with a continuously moving film was used to record the intensity-modulated signals on an oscilloscope. It is only by continuously moving over the ice and at the same time continuously recording the echoes that bottom can be seen with certainty amongst scatter echoes. All the equipment was mounted on a sledge and man-hauled over the glacier surface. Two antennae of the rectangular “trough” type, 1.5 wavelengths wide by 1 wavelength long and with corner angles of 45° were used. The measured forward gain of each antenna was 8 dB.

To correlate echo strength with predicted absorption, ice temperatures were measured at a depth of 9 m, along the length of the glacier.

Several longitudinal and transverse ice-thickness profiles were obtained by radio echo sounding (Figs. 2 and 3). Vertical scale calibration marks were put on the recording film at regular intervals using a 500 kHz crystal-controlled oscillator (Fig. 4), and the corresponding locations on the glacier were determined by interpolation using an altimeter and odometer between points fixed by triangulation.

Fig. 2. Longitudinal profiles of Roslin Gletscher. Vertical: horizontal scale ratio ≈ 5:1.

Fig. 3. Transverse profiles of Roslin Gletscher, looking down-stream. Vertical: horizontal scale ratio 1:1.

Fig. 4. A film record of ice depths.

Results

Roslin Gletscher is 37 km long and has an average width of 2 km. The sledge routes are shown in Figure 1. One longitudinal profile of all but the upper 5 km was completed, as well as six transverse sections and several other profiles. A maximum ice thickness of 370 m was measured between b and l, and in many places the ice is about 350 m thick (Figs. 2 and 3).

Reflections from the valley walls were not believed to be present, except when the instrument was being operated within 50 m of the glacier sides. Thus, using directional antennae on the glacier surface is a satisfactory method of measuring the depth of valley glaciers. Reflections from valley walls are likely to be more serious problems if the equipment is operated from an aircraft due to the larger field of view.

Absorption was higher than that normally found on polar glaciers at similar temperatures. Attenuation per 100 m path length was determined by comparing the received echo traces with traces at known attenuations in the laboratory.

Several mechanisms have been examined to try to account for the high attenuation values from Roslin Gletscher. One possible cause is the presence of water, causing power to be scattered from the forward wave by changes in permitivity. Water may be present either as a surface or englacial layer, or at grain boundaries. Alternatively, the impurity content of the ice may increase the dielectric absorption.

Echo strengths were 10 dB stronger at night than during the daytime when melt water was visible on the surface of the glacier. It has been shown that at 440 MHz a layer of water 5 mm thick accounts for a 10 dB attenuation to the two-way path. The maximum attenuation for each water layer, 19 mm or more thick at 440 MHz, is 11 dB one way but there may be several layers in the line of sight, each contributing about 5 dB, on average, to the total one-way loss (Reference Smith and EvansSmith and Evans, 1972). These attenuations are almost independent of the purity and conductivity of the water.

At zones in which the absorption was determined, the ice temperature at 9 m depth varied between −6° and −11° C. If areas where there was appreciable surface water or reflections from within the ice are neglected, and if only echoes from the ice–bedrock interface are considered, then absorption values varied between 5 and 10 dB per 100 m path length, with a mean value of 6.8 dB per 100 m path length. These values are higher than those found in polar ice sheets where the absorption is found to be 5.7 dB per 100 m at −1° C and 2.7 dB per 100 m at −10° C (Reference Robin, Robin, Evans and BaileyRobin and others, 1969).

The range of the SCR 718 was found to be satisfactory for measuring the depth of most of Roslin Gletscher. Several interesting physical features emerged from the results. The transverse profile fxpw (Figs. 1 and 3), below the confluence of the main glacier and a large tributary glacier, indicates that there is a well-defined subglacial ridge which separates the two ice streams, and suggests they maintain their own identity farther down the valley. There is also some evidence to suggest that between b and l, and near h (Figs. 2 and 3), discontinuities (possibly due to the presence of rocks) may be present within the ice at a depth of 200 m. These may have resulted from past avalanche activity on the mountain west of h (Fig. 1). It is a pity that the extent of this feature cannot be determined because of insufficient data.

Conclusions

The SCR 718 radio altimeter operating at a frequency of 440 MHz has been shown to give satisfactory ice-thickness measurements on a valley glacier. An advantage of using the fairly high frequency of 440 MHz is that antennae are compact and have good forward gain characteristics. Equipment can also be light and compact, and this is often important on a valley glacier where surface travel can be very difficult. Absorption and scattering considerations indicate that a lower frequency might be necessary for good results on some glaciers.