The radiant surface temperature imaged by thermal infrared line-scanning equipment is a product of both surface temperature and emissivity. In detail it is the integrated product of the evolutionary development of the near-surface energy transfer regime. The variation of surface radiant temperature spatially and temporally contains information about the structure, composition, and thermal state of near surface materials. It is suggested that the study of ground ice, sea/lake ice, and the thermal regimes of mountain snow and ice bodies, are reasonable targets for the quantitative application of thermal mapping technology, and that considerable technical development effort be expended on spectral, spatial, and temporal operators for use in computer processing of pictures for glaciological data.

Thermal infrared sensing can provide much information about sea ice, and some of the physical conditions associated with sea ice suggest that surface temperature may be a good indicator of ice thickness. However, steady-state heat-flow calculations suggest that the variable thickness of the snow-cover and its low. variable thermal conductivity would preclude the use of surface temperature alone as a suitable indicator of ice thickness. Measurements of surface temperature, snow depth, and ice thickness suggest that, in an area of relatively uniform ice thickness, surface temperature might be useful as an indicator of snow depth if some surface data can be obtained.

The ground-wave is the most important mode of propagation of radio waves in Connection with glaciology. In cold regions, special conditions are often prevalent, involving propagation over non-homogeneous earth, presence of stratified media, and low values of conductivity and dielectric constant in the upper strata.

A radio wave which propagates along the Earth's surface is, however, also influenced by atmospheric refraction. As the frequency is increased, the roughness of the Earth's surface must also be taken into account. Thus seasonal variations are to be expected due to changes in the electrical and topographical properties of the ground as well as the propagation conditions in the atmosphere. It is, however, difficult to separate these various effects, a fact which reduces the possibility of using ground-wave propagation as a loot for obtaining information on the properties of the ground.

Though the propagation of the ground-wave has been dealt with both theoretically and experimentally for almost a century, some of the most valuable information of major importance in cold regions has been obtained during the last ten years. New theoretical papers on propagation over stratified media offer an explanation of the amplitude and phase variations of the ground-wave field, which have been measured, as well as suggesting new methods to be tested as possible aids in solving glaciological problems.

In many practical eases of ground-wave propagation in arctic regions, the formula for the ground-wave field strength can be written in a very simple way. Such a propagation model for frequencies above 30 MHz is presented in which account is taken of the Earth's curvature, the terrain irregularities, and the effect of the tropospheric refraction. This model also includes the recovery effect which occurs in propagation over mixed paths. At the higher frequencies the effect of depolarization becomes very important and sometimes overshadows field-strength variations due to the influence of the electrical properties. Finally some problems will be discussed which remain to be solved or have been given very little attention up to now.

This paper presents an overview of recent remote-sensing techniques as applied to geophysical studies of floating ice. The current increase in scientific interest in floating ice has occurred during a time of rapid evolution of both remote-sensing platforms and sensors. Mesoscale and macroscale studies of floating ice are discussed under three sensor categories: visual, passive microwave, and active microwave. The specific studies that are reviewed primarily investigate ice drift and deformation, and ice type and ice roughness identification and distribution.

This paper provides a review of the characteristics of airborne laser profilometry and its application to quantitative characterization of cold-regions terrain. The limitations of profilometry due to the profiler instrumental characteristics and instability of the aircraft platform (resulting from variations in aircraft altitude and attitude) are discussed. For typical aircraft speeds of the order of 100 m/s these limitations restrict the accurately measured roughness content to the approximate wavelength range 2 m to 300 m. Digital filtering and hardware techniques for removing the aircraft motion, and hence extending the long wavelength validity of the profile, are discussed.

Regarding terrain characterization, particular attention is given to Arctic, sea ice. Ridge height and spacing distribution models for sea ice in conjunction with digitally processed laser profiles allow efficient characterization of sea-ice ridging using only a few parameters. In particular, a single ridging intensity parameter has been found to allow reasonable estimation of the number of ridges encountered at any height level along a straight-line path. Examination of spectral characteristics of first-year and multi-year ice suggest that laser profiles may be used to identify the ice type of floes and ridges. Comparisons of laser data and submarine sonar data are made which suggest that ratios of c. 6.5: 1 can be used to estimate ridge keel depths from laser data. Use of laser profilometry to characterize tundra and indirectly 10 measure variation in snow depth is briefly discussed.

A continuous profile of the Arctic Ocean ice canopy from Spitsbergen to the North Pole was made with 48 kHz echo sounders mourned on a nuclear submarine. A semi-automatic digitizer was used to measure coordinates from the records at a frequency of about 1 000 points per linear kilometre of track. Methods derived for the reconstruction of subglacial relief from 35 MHz radio-echo sounding records were applied to eliminate part of the distortion due to the 20° beam width of the sounders. A corrected profile was used to obtain ice drafts at 2 m intervals. Data were analysed in 10 km sections and figures were summarized for each degree of latitude. The results include: mean ice draft, percentage of ice less than 0.3 m draft, and percentage of level ice. Histograms show the level ice drafts which occur most frequently, and these may indicate the ratio of first-year to older ice. The number of ice keels is listed together with their mean draft and draft distribution. Definitions are found to be of overriding importance in the comparison of data from different areas.

The dependence of the extinction coefficient on salinity was investigated for both NaCl-ice and salt-ice made from natural sea-water. Specimens were prepared under a variety of conditions and examined over the wavelength range 4 000 to 8 000 Â. The effects of scattering from air bubbles trapped in the ice were examined for ice made from distilled water. It was found that the method of preparing samples markedly affected their structure, but that, when prepared in the same manner, salt-ice made from natural sea-water and NaCl-ice did not show significantly different transmission properties. It was found that, for a wavelength of 6 328 Â, the data could be fit to the relation within an uncertainty of 26%, where ke is the extinction coefficient, x is the salinity of the ice in g/kg. Within an uncertainty of 10%, there was no variation in transmission for ice at the same temperature and salinity over the wavelength range 4 000 to 8 000 Â. All measurements were made at a temperature of -20° C.

Starting with the TIROS-2 weather satellite in 1961 which permitted synoptic viewing of large-scale areas with an 011-board television camera system, the capabilities of satellite observations for assessing snow and ice resources on Earth have been greatly improved through the utilization of higher resolution imaging systems and multispectral images in the wavelength range from 0.4 μm to 1.55 cm. The possibility that the variation in areal extent of the snow cover may be related by empirical means to the average monthly run-off in a given watershed was demonstrated by comparing run-off records from the Indus River Basin in south-east Asia with a series of snow-cover maps obtained from Nimbus-3 and 4 imagery. Similar studies using the higher spatial resolution available with ERTS-I imagery were carried out for the Wind River Mountains watersheds in Wyoming, where it was found that the empirical relationship varied with mean elevation of the watershed. In addition, digital image-enhancement techniques are shown to be useful for identifying glacier features thought to be related to extent of snow cover, moraine characteristics, debris coverage, and the like. Finally, longer wavelength observations using sensors on board the Nimbus-5 satellite are shown to be useful for indicating crystal size distributions and onset of melting on glacier snow cover.

The ERTS satellite’s sun-synchronous orbit covers the entire Earth horn lat. 81° N. to 81° S. every 18 d at approximately 10.00 a.m. ground time. The scanners on board have continued to relay images to Earth from the launch on 23 July 1972 to the present (summer 1974). Images from four spectral bands (0.5-1.I μm), with ground resolution of about 100 m. cover an area 185 km on a side. The resulting images have proved to be very useful for collecting certain basic data from glaciers.

Long-term surface velocities are readily determined by comparison of recent ERTS images with maps that have been produced from earlier data. Images have been used 10 measure velocities on the Malaspina Glacier over a 10 year period, surge displacements on the Lowell, Tenas Tikke. and Tweedsmuir Glaciers and Lednik Medvezhiy, and velocity at the margin of the Hubbard Glacier. Many surging glaciers are readily identifiable on (he images. Coverage from the satellite will allow surging glaciers to be identified world-wide which may help glaciologists to understand their peculiar geographical distribution. Images of large glaciers taken under conditions of low sun angle and complete snow cover show previously undetected subtle slope changes which can be interpreted as dynamic How features or reflections of subglacial topography.

ERTS imagery is providing new glacier data for several reasons: very large areas are covered repetitively with sufficient consistency to show subtle tonal changes, conditions of maximum and minimum snow cover, subtle morphologic features at low sun angle, and changes in position of glaciers.

Due to increased demand for electric energy, which in Norway is produced almost entirely by hydro-electric power plants, extensive glacierized basins are now considered for future power production, However, glacier behaviour has a great impact on the annual run-off from these basins, and information about expected annual deviations from a “normal” discharge must be obtained from mass-balance studies on representative glaciers within the basin. For this reason such studies are presently performed at eight selected Norwegian glaciers; the costs involved are in the order of $3 000-$10 000 annually for each observed glacier.

The transient snow line can be identified easily on conventional air photographs and its height determined by direct comparisons with reliable topographie maps (with contour lines on glaciers). The resolution in ERTS images is sufficiently good to use the satellite data for a similar height determination of the transient snow line on glaciers—at least on the larger valley glaciers and, of course, on ice caps.

Due to an existing correlation between the height of the transient snow line (at the end of the melt season) and the annual specific net mass balance, it is possible to construct a diagram expressing this relationship, provided detailed mass-balance studies have been performed during a period of several years of different mass-balance conditions (ideally comprising both years of negative and of positive net balances). Determinations of the transient snow-line height based upon ERTS data, obtained at the end of the summer season, can then be used to determine the mass balance for those glaciers where the above-mentioned relationship is established. This would be a great advantage in cases where there is a request for information on the net balance only; no separate data on winter balance or summer balance will be obtained by this method.

Experiments have been performed to delineate the transient snow line on glaciers directly from ERTS digital data (CCT magnetic tape). The results indicate that last winter’s snow is characterized by grey-levels 41-48 (of 63 possible grey-scale levels) on MSS-7, whereas exposed glacier ice is represented by levels 13—20 in the same multi-spectral band. Snow and ice can also be separated on MSS-4, but the difference in grey-scale levels between the two materials is much smaller.

An analysis of records of annual mean temperatures around Antarctica shows large-scale anomalies of thousands of kilometers extent with typical variations of a deg from one year to another From 1967 on, composite satellite photographs are available which show considerable variation in the sea-ice extent in different years up to about 5º of latitude. These largest differences seem to persist over entire seasons. In general there seems to be considerable association between the region around the Antarctic with the coldest temperatures and the regions of greatest sea-ice extent. An analysis of long-term records at a single location near the edge of the Antarctic sea ice indicates a strong correlation between variations in the annual mean temperature and the duration of the sea ice, such that a change of 1 deg in the annual mean temperature corresponds to about 70 d variation in the duration of the sea ice. A relation is obtained between variations of annual mean temperature and the mean extent of the sea ice, viz. a 1 deg change corresponds to approximately 2.5º latitude variation in the maximum sea-ice extent. The magnitude of the variations in the sea-ice extent observed from the satellite data in comparison with the large-scale temperature anomalies is compatible with the above relations, although some rotational shifts appear to take place.

Is it justified to adopt a two-dimensional continuum model for the movement and large-scale deformation of pack ice? A preliminary study oi’ ERTS-1 photography shows that the details of the ice movement are readily measurable; the problem is not in the accuracy of the remote sensing but in the inherent graininess of the sea ice. There is a spatial variation of ice velocity on a scale of several hundreds of kilometres; smaller-scale variations are superimposed on this, but their amplitude is not enough to obscure the large-scale trend. A continuum model is applicable, but, because of the small-scale variations in the velocity of the sea ice itself, it is not meaningful to specify continuum strain-rates on a scale of, say, 100 km to more than a certain accuracy, If ERTS pictures are available during the AIDJEX main experiment they could provide the necessary strain and displacement measurements for comparison with the predictions of the AIDJEX model.

The final session of the symposium was devoted to a “brain-storming” general discussion.

Bottom crevasses were detected at many locations in the southern portion of the Ross Ice Shelf during the 1973-74 austral summer. The crevasses which extended up about 100 m from the bottom of the shelf were detected by radio-echo sounding. These linear features were mapped in some detail at the RISP Camp. Wide-angle reflection velocity measurements, airborne radio-echo sounding, and other results of the R1GGS program will be included in the discussion.

In May 1974 two radio depth sounders were used on the Barnes Ice Cap, Baffin Island. One was a S.P.R.I. 35 MHz sounder and the other was a unit operating at 620 MHz. Bottom reflections were observed with both systems at similar depths, indicating no significant velocity change between the two frequencies. Used with a Motorola range positioning system, the 620 MHz unit proved an excellent depth-survey vehicle. Results along the so-called surge profile (Holdsworth, 1973) are presented. Depth and attenuation values are compared with previous data.

A short-pulse, long-wavelength radio-echo sounder has successfully measured the ice depth on the South Cascade Glacier. Depths up to 250 m were determined with resolution of about 5%. Bottom returns were clear and almost never ambiguous. Their accuracy was confirmed by comparison with hot-point drilling results. The secret for successful sounding in temperate glaciers is the use of a sufficiently low center frequency. Five megahertz was most successful. Tests at 15 MHz indicated an increase in coherent clutter which rendered the bottom return observable only with prior knowledge of its location. The cause of the clutter is probably water-filled voids in the ice which behave as Rayleigh scatterers.

The sounding system consists of an avalanche-transistor transmitter, which delivers a pulse to an acute-angle crossed-wire antenna. The pulse is shaped and given its center-frequency characteristics by the resonant properties of the antenna. The transmitting and receiving antennas are identical, consisting of wires and lumped resistors. The resistors reduce antennas ringing, thereby maintaining as short a pulse as possible. The receiver consists of an oscilloscope and a Polaroid camera. No preamplification is required for depths up to 250 m, but may be necessary for deeper glaciers.

The study of ice in the upper Great Lakes, both from the operational and the scientific points of view, is receiving continued attention. Both quantitative and qualitative field work is being conducted to provide the needed background for accurate interpretation of remotely sensed data. The sensor data under discussion in this paper were obtained by a side-looking multiplexed airborne radar (SLAR). These were supplemented with ground-truth data.

Radar, due to its ability to penetrate adverse weather, is an especially important instrument for monitoring ice in the upper Great Lakes. It has been previously shown that imaging radars can provide maps of ice cover in these areas. However, questions concerning both the nature of the surfaces reflecting radar energy and the interpretation of the radar imagery continually arise.

Our analysis office in Whitefish Bay (Lake Superior) indicated that the combination of the ice/water interface with the ice/air interface is the major contributor to the radar backscatter as seen on the imagery. The ice has a very low dielectric constant (<3.0) and a low loss tangent. Thus, this ice is somewhat transparent to the energy used by the imaging SLAR system. The ice types studied include newly formed black ice, pancake ice, and frozen and consolidated pack and brash ice.

Although ice thickness cannot be measured directly from the received signals, it is suspected that by combining the radar backscatter information with both meteorological and sea-state history of the area and with some basic ground truth, better estimates of the ice thickness may be provided. In addition, certain ice features (e.g. ridges, ice foots, areas of brash ice) may be identified with reasonable confidence. There is a continued need for additional ground work to verify the validity of imaging radars for these types of interpretations.

This paper is being published in full in another issue of the Journal of Glaciology.

The remote sensing of snowpack depth, density, and wetness with an airborne system would have important applications in water resource management and flood prediction. In this paper, the electromagnetic response of multi-layered models is analyzed. Normally-incident plane waves are assumed at frequencies ranging from 106 to 1010 Hz, and reflection amplitudes are calculated for models having various layer combinations. Each layer can have arbitrary thickness, and its own dielectric constant and conductivity, each of which can vary with frequency. Thus “lossy” media as well as “perfect” dielectrics can be employed in the models. An outline of the theory for the calculations is presented for an n-layered model. Because of the complexity of the equations, interpretation is accomplished by illustrative models, selected from seven snow types and seven earth types. The objective of this type of calculation is to establish the dependence of the reflection coefficient on the impedance transitions between two half-spaces. This paper is a theoretical study only, and does not include consideration of the size, weight, estimated cost, and other physical attributes of a flight system. These, and other matters of a practical nature, are being treated in other papers.

A revised version of this paper is being published in full in another issue of the Journal of Glaciology.

This paper discusses the work being done at the Defence Research Establishment Ottawa (DREO) on profiling of Arctic sea ice using airborne laser profilometers. Data from a flight in March 1973 have been partially analyzed and some results are presented. These include representative ridge counts and power spectra for chosen areas in Nares Strait and the Arctic Ocean. Some work has been done to try to improve the operation of the laser profilometer using real-time data processing. The basic ideas behind this scheme are outlined.