252 INSTITUTE NEWS
THE DEVONISLAND EXPEDITION
In 1959 the Arctic Institute of North of knowledge in those subjectsfor America undertook an integrated pro- Devon Island. gram of long term research on Devon In the latesummer of 1960 a main base Island in theQueen ElizabethIslands of was established on the north shore of arctic Canada. The co-ordinated studies Devon Island near Cape Skogn by an were designed to help understand the advance party of eight mentaken in interrelationships between the glacier- with theirmaterials by the Canada ice of Devon Island, the ocean in Jones Department of Transporticebreaker Sound,and the encompassing atmos- “d’Iberville”. Duringa 3-week period phere. They are being carried out over buildings were erected and routes inland a 3-year period under the leadership of and to the ice-cap explored and marked, Spencer Apollonio. The main effort is while an archaeological reconnaissance concentrated on attempts to evaluate of the Cape Sparbo area was made by a such factors as physical, chemical, and small party under Mr. Gordon Lowther biological variations inthe arctic waters of McGill University. Everything was of Jones Sound caused by discharging installed for a beginning of the 3-year glaciers; evaporationand transfer of moisture between the ocean waters and program in April 1961. the ice-cap and glaciers; and the overall During the months of April to Sep- influences of solar radiation energy on tember 196121 men worked on extensive the mass balance of the ice-cap, the programs in geophysics, glaciology, ma- biological production in thesea, and the rine biology and oceanography, meteor- growth and decay of sea-ice. Some sup- ology, and surveying. Intensive work plementary studies in archaeology and was also completed in archaeology and geology are included in the expedition’s geology. The following personnel took work because of the marked deficiency part: archaeology: R.G. Lowther McGill University M. Tamplin University of Toronto geology: J. W. Cowie Bristol University A. Ormiston Harvard University geophysics: K. Vogtli Generaldirektion, Post-Telegraph-Telephon (Switzerland) J. P. Greenhouse University of British Columbia W. P. Molson, Jr. McGill University glaciology: S. R. Ekman University of Stockholm R. M. Koerner University of Birmingham G.E. Stewart McGill University marine biology, S. Apollonio Arctic Institute of North America oceanography: B. Beck Arctic Institute of North America meteorology: L. Dahlgren University of Uppsala B. Holmgren University of Uppsala J. W. Fellows Arctic Institute of North America A. Gill Arctic Institute of North America C.W. Nicol Arctic Institute of North America surveying: R. Wyness Royal Engineers (United Kingdom) P. Cress University of Toronto mechanic: T. R. Welch Arctic Institute of North America pilot: R. Carswell ArcticInstitute of North America INSTITU TE NEWS INSTITUTE 253
In addition to support from Arctic In- Research (U.S.A.), theQuartermaster stitute general funds, the Devon Island Corps of the United States Army, the Expedition has received major support Royal Societ.y, the Signal Corps of the in funds or equipment from the Defence United States Army, the United States Research Board of Canada, the Geo- Steel Foundation, and the United States physical Research Directorate of the Weather Bureau. In addition alarge U.S.A.F., the Hudson’s Bay Company, number of private individuals and com- Massey-Ferguson Ltd., the Meteorolog- panies gave significant help by contri- ical Branch of the Department of Trans- buting funds, services, or equipment. port (Canada), the Mount Everest Thegeneral summary that follows Foundation, the National Research and the preliminary reportsare all sub- Council of Canada, the National Science ject to revision after study of the data Foundation (U.S.A.), the Office of Naval obtained.
Report of the field leader, cap meteorological station was found to April-September 1961 be on ice 750 metres thick. Resistance properties of snow, firn, and ice were The main geophysical, glaciological, determined and two distinct resistance andsurveying work was done on a characteristics were observed in the ice. glacier that forms anoutlet for the Some attempts havebeen made to relate Devon Island ice-cap and enters Jones these differences to the origin and his- Sound at about 83”lOW. For easy refer- tory of the ice. Samples were collected ence it has been tentatively named the in order to make aspectographic deter- “Harald Sverdrup Glacier”. The glacier mination of the ionic composition of the itself is about 10 miles long and 2 miles ice in an attempt to relate that to its wide, flowing from south to north be- electrical properties. In August two vis- tween valley walls thatrise almost iting engineers from the U. S. Army 1,300 feet above its surface. Ithas a attempted depth measurements on the prominent medial moraine, andabout glacier by a new method using radio 5 miles from the coast a small outlet altimetry,and the results obtained glacier flows back to the south where it agreed generally with results obtained almost joins a second and much smaller by the electrical resistance method. The outlet glacier. An ice-cap station was glaciology program was initiated and maintained at approximately 75”28’N. carriedout by S. Ekman and R. M. and 83”W. at which meteorological rec- Koernerduring May, June, and July, cords and glaciological studieswere when Ekman was called home because made. Travel on the glacier and ice-cap of serious illness in his family. The pro- was by WEASEL, although a light air- gram was then continued during August craft (aPiper PA18 SuperCub) was and September by Koerner and Stewart, operated on skis and low-pressure tires following instructionsleft by Ekman. throughout the summerin support of Over 50 pits were cut on the “Sverdrup many of the operations. The plane, Glacier’’ and ice-cap to allow deter- flown by its owner, Ross Carswell, put minations of snow-depth, density, tem- in over 200 hours of flying without inci- perature, and hardness; vertical distri- dent in spite of numerous landings on bution of snow, firn, and ice-crystal difficult and unprepared areas,including structures; internal temperature varia- the glacier, the ice-cap, and open terrain tions to depths of 12 metres; accumula- after the snow had gone. tion and ablation. There was an attempt In the geophysics program numerous to determine the runoff of melt-water depth and volume determinations were into the sea, using water-level, current, made on the glacier and ice-cap, using and weir measurements. The glacier lost an electrical resistance measurement almost 2 meters of ice during 1961 and method described in more detail sepa- the firn line on the ice-cap reached its rately. Themaximum depth recorded on greatest height at about 1,700 metres. the glacier was 600 metres and the ice- Inthe fields of marine biology and 254 NEWS INSTITUTE
oceanography tidal observations were rings, meat caches, andseveral other made during acomplete lunar cycle. The structures.Preliminary surveys were development of algae as indicated by made on two other previously unre- chlorophyll concentrations in sea-ice ported sites located from the air and on was measured, and the chemical prop- the ground in thevicinity of Cape Hardy erties of the ice were examined. Routine and Cape Sparbo. Excavations were also measurements of photosynthesis in the made at a number of sites nearthe main coastal waters were made using three base, and 200 artifacts were recovered different methods; salinity, andthe from pre-Dorsetand Thule remains. concentration of chlorophyll, phos- Thematerial excavated covers about phates, and silicates were determined, 3,500 years of almost continuous occupa- and zooplankton and phytoplankton in- tion. Geological interpretations of the vestigated. Temperature andlight pene- sites were provided by Cowie and Or- trationmesaurements were also rec- miston, with controlled local surveys orded. Rather high chlorophyll concen- and levelling being done by Wyness and trationsand moderate photosynthesis Cress. Carswell located several new rates were observed, and there is some sites from the air between Cape Hardy evidence of limitation of nutrients. and Belcher Point, at Rigby Inlet, and Comparable observations in meteor- at Eskimo Inlet. ology were made at the base camp near Fossil collecting and mapping were sea-level andat an altitude of 1,500 carried out in the area, including lower metres at the ice-cap station They in- palaeozoic sediments at Cape Skogn, cluded standard synoptic observations Dundas Harbour, Crocker Bay, Burnett and measurementsof total and net short Inlet, Maxwell Bay, Sverdrup Inlet, and and long wave solar radiation. Work at Eskimo Inlet. The lithography and the ice-cap station also included a pro- stratigraphy of thenorth andsouth gram of micrometeorological observa- coasts were related to earlier work at tions of wind and temperature. Snow- Dundas Harbour and to thework of the depthand density observations were Geological Survey of Canada on the also made whenever possible. A third western end of the island during “Oper- station, onthe “Sverdrup Glacier,” ation Franklin”.The results permit provided continuous records of humidity correlation of the geology of Devon and temperature in aStevenson screen. Island with that of Cornwallis Island and The glaciologists weresupported in areasfarther to the west. Devonian the field by two surveyors who studied fossils were collected in Prince Alfred the horizontal and vertical motions of Bay and Arthur Fiord. theSverdrup Glacier using standard Five men will be continuing the surveying techniques to record motion meteorological and oceanographic pro- of 29 stakes arranged infive profiles.The gramsduring the winter of 1961-62. base-line for this survey was measured Investigations will bemade into the with tellurometerequipment by the mode of formation and the physical, Topographical Survey of Canada. The chemical and biological properties of glacier was found to move about twice sea-ice. Work will beresumed on all as farin August asin July, and the phases of the expedition in the summer maximum total movement observed of 1962, except for geology and archaeo- between June and the end of August was logy, but in meteorology more attention 9 metres. will be paid to solar radiation observa- The archaeological sites at Cape tions and micrometeorological processes Sparbo, which had been discovered in near the surface, using wiresonde bal- 1960, were excavated during the season loons and mast instruments. New in- by G. R. Lowther for the National Mu- strumentsfor this purpose, andfor seum of Canada, assisted by M. Tamplin. measuring light transmission through Over 350 artifacts were recovered and sea-ice will beprepared during the pre-Dorset, Dorset and Thule material winter. was found. The siteincludes houses, tent SPENCERAPOLLONIO INST ITUTE NEWS INSTITUTE 255
Geology Upper Cambrian fossils were identified. During the summer of 1961 an attempt Lower and Middle Ordovician rocks was made to trace thegeological succes- were measured and described in areas sion on eastern Devon Island upwards 3, 5, and 6, and faunas werecollected at from the metamorphic basement rocks a number of horizons. Some of these through any Precambrian sediments fossils may be new. into the Cambrian and Ordovician. All The objectives set for the season were fossil collections made were accurately achieved but work remains to be done located in measuredstratigraphical in order to understand 1) the metamor- order and came from rocks in situ. In phic basement complex and the associ- addition a detailed map of the central ated dykes, 2) the stratigraphyand part of Devon Island was begun andwill palaentology of the Upper Ordovician be completed from aerial photographs at and Silurian rocks in the western part a scale of 1:250,000. Complete coverage of the island. These strataand their at a large scale was not achieved because equivalents were studied during"Oper- of restriction inthe use of the expedition ation Franklin" of the Geological Survey aircraft. When the fossil faunashave of Canada, and also in considerable been identified in Bristol consideration detail by R. Thorsteinsson on Cornwallis will be given to the matter of faunal Island. Continuous air support would be provinces and palaeoecology. required to improve on that work. 3) The areas visited in Devon Island all Although thestructures in central lay between 82"W. and 89"W. and the Devon Island are very simple and would specific sites were located at:- not repay study for academic purposes, The coast near Cape Skogn thestructures in andnear Grinnell The large lake inland from Cape Peninsula are of interest, butagain work Newman Smith in that areawould require the use of an The western side of Dundas Har- aircraft on a continuous close support bour basis. A valley at the northeastern end J. W. COWIE of Croker Bay The head of Burnett Inlet Measurement of electrical resistivity A valley running northeast from Fellfoot Point(to the east of of ice Maxwell Bay) Measurements of the specific resist- The big valley runningsouth ance of the ice of the Devon Island ice- from Sverdrup Inlet cap (approx. 75'30'N. 83"W.) and of an The Precambrian metamorphic base- outlet glacier discharging intoJones ment complex was examined whenever Sound gave values of 50,000 to 100,000 time permitted. The foliation, strike, and ohm-metres andthey were about ten dip were recorded; and the basic dykes times higher for the underlying bedrock. sampled and faults traced.A petrological This large difference makes it possible sample wastaken from within each to measure accurately the thickness of measured section inthe sedimentary the ice-cap or of the glaciers by an cover. No Precambriansedimentary electrical method. rocks were recognized. The method consists of sending a d.c. The Cambrian rocks were measured current through theice by meansof two and described in areas1, 2,3,4,5, and 7. widely spaced electrodes and observing It was found possible to use lithology to the resulting electrical field. If the elec- correlate formations throughout the trodes are separated by a distance equal area, and the formational names set up to about twice the thickness of the ice by Kurtz et al. for the Dundas Harbour a large proportion of the current will area (1952) can be used over the whole flow through the underlying rock. The of central Devon Island. Fossil horizons large difference between the resistivities containing new faunas werefound in the of the two media results in a distortion Lower and Middle Cambrian but no of the field that can be measured easily 256 NEWS INSTITUTE
and from the amount of distortion the formed ice-masses alwayshave high thickness of the ice can be calculated. resistivity such findings would be of With an amountof equipment that can greatinterest to glaciologists, because be carried in three packs, thicknesses resistance measurements would provide exceeding 500 metres could be meas- an easy method for distinguishing be- ured, and on a glacier 30 km. long and tween accumulation and ablation zones. 2.5 km. wide 33 points were investigated Moreover, fullunderstanding of the in 1month. Most of this time was spent factors influencing the conductivity of on the often difficult travelling to the old and new ice could provide informa- points of measurement,whereas the tion about the climatic conditions pre- measuring itself required only 3 hours vailing atthe time of formation of at each point. ancient ice-masses. The geo-electrical measurements give K. VOECTLI information not only about the thickness 1Keller and Frischknecht. 1960. J. Res. and electrical properties of the ice- 64D439. masses but also about the conductivity of the bedrock. They indicate that the Glaciology major part of the glacier is resting on compact rock, but in one small area the During the summer of 1961 a program ice is underlain by thick sediments, of glaciological studies was initiated by which havea much lower resistivity S. Ekman of the University of Stock- than the ice-mass itself. holm, assisted by R. M. Koerner. When Several small areas of the glacier were Ekman was forced to return to Sweden of special interest because theyhad on July 5, Koerner carried on the work, higher resistivities of the order of 1 assisted by G. Stewart. The work in- megohm-metre. This kind of ice has a volved astudy of accumulation and small conductivity comparable to that ablation on the Devon ice-cap and on a of the ice-masses of the Alps and of the selected valley outlet glacier (the “Sver- Athabaska Glacier inAlbertal. There drup Glacier”), with particularatten- is as yet no explanation for the large tion to runoff on the glacier. difference in resistance between some Before the melting season began, the arctic ice and that known for glaciers in accumulation of the 1960-61 winter was middle latitudes. Devon Island has both examined at various altitudesby digging kinds of ice and provides an excellent a number of snow pits, and it was found opportunity for the study of this prob- that the accumulation varied very little lem. with altitude. Old firn was recognized Resistivity measurements on theice of above altitudes of about 1,700 metres. a meltwater-fed lakegave the same low Cores up to 12 metres long were ob- values as those generally found on the tained on the ice-cap. Ablation stakes glacier. Similar resultswere obtained were set in holes drilled in the surface by measuring of crevasses that had been of the valley outlet glacier, and along a filled with melt-water and frozen. These profile to the highest point on the ice- observations eliminate explanations cap at approximately 3-kilometre inter- based on the effects of recrystallization vals. or pressure. Analysis of different sam- When ablation had begun, synoptic ples of melt-water will supply evidence meteorological observations, snow tem- as to whether differences in salt content peratures, and temperatures at the ice arethe main cause. Measurements of surfacewere recorded every 2 hours, sea-ice nearthe coast gave values of together with the amount of ablation at only 40 ohm-metresand thus showed a campsite on the valley glacier. These that ice of high salinity can indeed have recordings were discontinued when the aquite low resistivity. On theother ice-surface temperature reached 0°C. hand ice resulting from recent falls of Runoff measurements were made in a rain orsnow had a veryhigh resistivity. melt-water stream on the valley glacier If it could be confirmed thatfreshly near the campsite at about 300 metres INSTITUTE NEWS INSTITUTE 257
above sea-level using an Ottcurrent winter. Generalobservations were made meter, and velocitya revolution counter. of factors indicating recession atthe The area of the catchment basin of the edge of the glacier, they suggest that stream was ascertained by triangulation terminal recession of the various small on the ground. ice-cap outlet glaciers is very small, but The period July 13 to 28 was spent in that vertical wastage is taking place. work on the ice-cap where pits were The program for the summer of 1962 dug into theprevious winter’s snow, and will be arranged after the study of the the progress of melting above and below 1961 results, but itwill definitely include the firn-line was examined. A line of a more detailed investigation of runoff snow pits at intervals of 2 miles was using more accurate methods, since the taken over the ice-cap and down to methods used and the results obtained below the firn-line on the southernside. during the past season were not com- Superimposed banded ice was found in pletely satisfactory. During the winter every pit. Densities, temperatures, and Koerner will obtain 10-metre cores at stratigraphy were studied in every pit, the ice-cap station, and 5-metre cores and ice samples were examined in order at 60 metre altitude intervals between to determinetheir crystallographic the ice-cap edge and the ice-cap station, structure. Work on the “Sverdrup Gla- examining all the cores for crystallo- cier” was resumed between July 29 and graphy and stratigraphy. Similar studies August 20. The scope of runoff meas- will be made of lake-ice and sea-ice in urements was extended to estimate the order to help the biological program. All runoff through all the streams on the glaciological studies will be extended glacier surface,relating them to the during 1962 to include the smaller sepa- runoff of the stream atthe campsite. The rate ice-caps and some areas of dead major part of the water was found to ice. flow at the sides of the glacier where the R. M. KOERNER streams actively erode the ice. The stream on the west side carries about four times as much wateras all the other Observations of glacial movements streams combined. Regular measure- Observations of the movement of an ments were made of the runoff past a outlet glacier of the Devon Island ice- lineat 300-metre altitudeand on one cap were made during the summer of occasion runoff through all streams was 1961 to help arrive at anestimate of the measured asclose to the glacier snout as amount of ice enteringJones Sound possible. All the main glacier surface each year. Theparticular glacier se- streams end in moulins at least 400 lected (tentatively named “Sverdrup metres from the snout. Glacier”) is one of the two largest out- Work was resumed on the ice-cap on lets west of Belcher Point and into Jones August 21 and continued there to the Sound proper. It flows10 miles from end of the summer field season on Sep- south to north from the point at which tember 6. The snow-pit studieswere it leaves the ice-cap, hasan average repeated between the ice-cap station width of 2 miles, and is 1.25 miles wide and the ice-cap summit. These revealed where it enters Jones Sound. It hastwo a profile of new firn lying on ice below major tributaries less than 1mile wide, the firn-line, indicating that themelting and two otherrelatively insignificant season on the ice-cap had been unusu- tributaries. The ice slopes steeply at the ally short.Five-metre ice-cores were edge of the glacier, with the perimeter made at 60-metre altitudeintervals rising vertically for 40 feetin most from a point 60 metres above the firn- places. A melt-waterstream 40 feet line to the ice-cap station at 1,400 me- wide flows along the western edge of the tres.The cores will be examined for glacier during themelting period. stratigraphy,density, gas bubble con- The surface movement of the glacier centration,and crystallography by was found by determining the position Koerner at the base camp during the of stakesset into the ice relative to 258 INSTITUTE NEWS
signals placed on the rock of the cliffs On July 18 the Topographical Survey overlooking the glacier valley. The of Canada measured a base-line using actual positions were found by resec- tellurometer equipment. The direct tion, using a Wild T1A theodolite, which reading was 5,285.1 feet (1610.90 metres). reads directly to 20 seconds, with angles The ends of the base-line were named being given to the nearest 5 seconds by N and S, and it was joined to the previ- the surveyor. Six signals were erected ous triangulation by a braced quadri- at distances of 2 miles apart in such a lateral; the greatest adjustment to any way thatat leastthree were visible angle was 3 seconds. Reciprocal vertical fromany one stake.The signals con- angles were taken at N and S, which sisted of 1-inch-diameter aluminum determined the height of S above N as poles, 6 feet in length, placed vertically 118.22 metres and 118 metres, the mean over a paint mark or rock and held in difference being 118.11 metres. Because position by a rock cairn. A tin can was thejourney between a signal anda placed on each signal as a reference stake took an average time of 5 hours, point for vertical angle observations. and because it was difficult to get on Twenty-nine stakes were placed in the or off the glacier at its perimeter, stake glacier along five profiles, at locations positions were fixed by resection. The previously selected from aerial photo- first round of observations on all stakes graphs. The profiles were approximately was begun and completed on July 2. 0.75 mile apart.The stakes were alu- The second serieson profile one only minum poles, 13 feet long and 1 inch was begun and completed on July 7. in diameter inserted into drill holes in Final observations on allstakes were the ice approximately 10 feet deep. begun on August 2 and completed on Profiles one, two, and three were linked August 15. At each stake four rounds by the placing of two stakes between of horizontal angles, vertical angles to each profile. One profile of seven stakes two signals weretaken, extra rounds was measured three times (at theoutset being taken when necessary to replace and after 30 and 60 days respectively) poorones. The greatest difference be- butthe remaining stakeswere fixed tween the mean and any single result twice only. Thevertical positions of was 15 seconds and averaged 6 seconds profiles one and two were fixed twice for the first observation on each stake. by levelling from a rock mark on a cliff There was still some snow left on side. The greatest closing error on any June 8, and there was very little melt- stake was .15 feet (4.5 centimetres). A ing so that little difficulty was expe- third round of levels was taken but not rienced in keeping the theodolite level related to the rock mark during observations. When the melting A minimum of five rounds was taken rate increased the tripod was set up on at each signal point. The greatest dif- thin flat rocks about 6 inches across, ference between the reduced angles for which were placed in holes cut 2 or 3 thetriangulation was 28 seconds, and inches down into the ice, with ice chip- averaged 17 seconds. The difference pings piled up aboutthe foot of the from the mean angle is approximately legs. In this way the theodolite was kept one-half that figure. Observations im- level during rounds. Therewere only proved during the season when the in- 4 days when the melting was sufficient- strument and signals became familiar. lyrapid to prevent work. Because of Thetriangulation was adjusted by a the ablation some of thealuminum braced quadrilateral and braced trian- poles had to be cut so that thetheodolite gles, thegreatest adjustment to any could be set up over the stake. one angle being 6 seconds. There was The centre stake of profile one moved sufficient light to observe at any time 3.49 metres in 35 days and 6.59 metres of the day or night from June 1 until duringthe next 37. The remaining theend of July, but thetemperature stakes in profile one confirmed the dropped to the freezing point by 1800 increased rate of movement during the hrs. every day. second period. The vertical movement INSTITUT E NEWS INSTITUTE 259 of the stake tops was small, being only ever no further arctic field workhad about 25 cm. for the greatest rise, but been done. The interest lies not so much the movement indicated thatthe ice in the actual resistivity, but rather in rose slightly at the edges and sank very the possibility that variation in resis- slightly in the centre. tivity of the ice can belinked with The profile stakesand the signals variations in the physical properties of were left in position. Thestakes will the ice and that an understanding of it be resurveyed in 1962 to determine the can be used as a glaciological tool. At amount of movement during the win- the present time there is no literature ter. The stakes will then be re-estab- on the subject in English and only a lished and the cycle of observations small amount in German. repeated using the same signal stations. We should like to acknowledge help Method given by Keith Arnold (Department of The basis of the method is measure- Mines and Technical Surveys) who gave ment of the difference in electrical po- advice in the preliminary planning, and tential (AV) between two points on the to Paul Atkinson, of the Topographical surface of a medium, caused by a cur- Survey of Canada, who measured our rent (I) passed through that medium. base-line. The potential field generated by a given P. CRESS currentdepends only on the resistivity ' R. WYNESS of the medium (measured in ohm- metres). If the medium is composed of Measurements of electrical resistivity layers of different resistivities, then each of these layers exerts an influence of ice-fvnnations according to its extentand its depth This is apreliminary report on the below the surface; in such a case meas- field work carried outby the geophysical urement of AV and I would give an party under Dr. Kurt Vogtli during the apparent value of the resistivity (p app) summer season of1961. A description that isa composite effect from all layers. of the techniques has been added since, Analysis of apparent resistivities can althoughthey are well-known geo- be used to determine the depth, extend, physical survey methods, prior to the and individual resistivity of the various last 2 or 3 years few attempts had been layers. made to adapt them to use with ice. The The potential field for any given si- recent development of a sensitive volt- tuation can be calculated theoretically meter with a very high input resistance by solving Laplace's equation v2v = 0 to replace the more cumbersome poten- with the appropriateboundary condi- tiometer is the most notable innovation tions. In a homogeneous isotropic me- in thepresent work. The final statement dium of resistivity p, bounded in one of results will come from Dr. Vogtli, plane by materiala of infinite re- who initiated, supervised, and took part sistivity (as, for example, the surface in the work. of the ground bounded by air),the The work had two main purposes. solution of the Laplace equation gives The first was to provide measurements the potential V at a point in the me- of the thickness of glacier ice using geo- dium distant r from the source of cur- electric means. The second was to rent I placed on the interface as being determine the resistivity of arctic land- V = (m.k.s. sytem). To illustrate, ice. Thelatter problem hasattracted 2 rr attention in the last 2 years since it had consider two electrodes connected to been found that ice from the Greenland opposite terminals of a battery and in- ice-cap has a resistivity of the order of serted in the earth so that a current I 10-3 times that of continental European flows between them. This is equivalent glaciers. The reasons for this difference to having two current sources +I and are generally thought to be a matter of "I separated by a distance d on the chemical composition and origin. How- air-earth interface. Assuming for the 260 INSTITUTE NEWS moment that the rock is uniform to awhere a, b and d are as shown in the considerable depth, thenthe potential figure. Inpractice the electrode con- anywhere is given by the sum of the figuration of either Wenner or Schlum- potentials due to the two sources. View- berger is used. These are shown in Fig. ed in a vertical plane through the elec- 2 together with the expressions for AV trodes the situation is as shown infor each. Fig. 1.
Fig. 1. Generalarrangement of electrodes.
If a voltmeter is now connected be- It is importantto realize thatthe tween any two points PI and P, then depth to which the current will pene- the values of I and VP, - VP, = AV trate increases with the distance be- can be used with the above formula to tween the electrodes. This is not obvious find p: but can be shown by plotting the lines of equipotential in each case. As a gen- eral rule, current I penetratesto a depth approximately equalto "a" in each
*V=PL 2 na I_ -D +I T -I
WENNER
SCHLUMBERGER Fig. 2. Arrangement of electrodes after Wenner and Schlumberger. INSTITUT E NEWS INSTITUTE 261 method, and effects produced atthat Effect of layers on current flow depth can be measured on the surface. When applied to depth finding in ice, When instead of a homogeneous me- the problem is essentially one of sepa- dium layers with different resistivities rating two layers, ice and bedrock. The are encountered p must be replaced by method of either Wenner or Schlum- p apD in the above equations. The cur- berger may be used but preferably both, rent flow will be bent toward the layers since the theoretical curves foreach are of lower resistivity and away from the available and thisgives two independent layers of higher resistivity. The potential results. Once the ice and bedrock resis- field is thus distortedfrom that produced tivities have been determined however, in the homogeneous medium andthe one method is sufficient, and Schlum- value AV at the surface is altered (Fig. berger’s being faster is preferable. 3). A single measurement cannot tell the Ina party of three, one man reads observer whether the observed resis- theinstruments while the other two tivity is thetrue resistivity of a ho- move the electrodes. The distance must mogeneous medium or theapparent first be measured out on either side of
Fig. 3. Distortion of electricalfield by layers of varying resistivity.
resistivity of a medium of several layers. the instrument set up, and the electrode However, by varying the distance “a” positions marked with flags. Distances the current can be made to penetrate of “a” are chosen so as to give an even to different depths at which it will ex- spacing on the logarithmic scale (i.e. perience various deflections depending an equal number of points in the ranges on the layers encountered. By plotting 1 to 10, 10 to lo2,lo2 to 103 metres). p apg against “a”a profile is obtained This is desirable since the theoretical that can be compared with similar pro- curves are plotted on “log-log” paper. files calculated theoretically for particu- Starting from maximum separation the lar situations. These theoretical curves electrodes are moved toward the centre have been found by solving Laplace’s with a reading being taken in each posi- equation with appropriateboundary tion. Between readings the cable must conditions for two or more layers with be rewound on the reels as the length various combinations of resistivities and of line required decreases; this precau- depths. If the plotted profile fits a theo- tion eliminates wear on the cables and retical curve, apicture of the resis- the inevitable “snagging” that occurs tivities and stratification below the when they are dragged across the ice. surface is obtained. Finally the flags are retrieved and the 262 INSTITUTE NEWS equipment is packed up. A profile with from each other. Since cable reels are “a” up to 1000 metres can be completed seldom well insulated they must be sus- in about 3 hours under good conditions. pended off the ice during readings if The data are plotted on “log-log” paper they are not actually beside the elec- and compared with theoretical curves. trodes. The instruments used were a Simpson (f) Telluric currents. Earth currents Model 269 Microammeter and aKeithley associated with ionospheric and mag- “300” Electrometer. The latter is a vac- netic storms can disrupt readings. If uum-tube voltmeter with a very high possible the profile should be made per- input resistance (lo14 ohms), necessary pendicular to the direction of such cur- when dealing with small currents and rents. high resistivities. The current electrodes The accuracy of a depth measurement were of iron, whereasthe potential elec- is almost entirelydetermined by the trodes used consisted of a copper rod in “fit” obtained betweenexperimental a solution of CuSO, all encased ina and theoretical curves. A single reading china pot with a porous lower section. can have an accuracy of between 30 and The solution is necessary since a spon- 10 per cent depending on the “fit”. Sub- taneous potential exists between the ice sequent readings in the same area im- and the electrodesthat must be the same prove the knowledge of resistivity layers and constant for both potential elec- involved, and in general an overall ac- trodes if the voltmeter is to read zero curacy within 15 per cent can be ob- before current flows. With metals alone tained. the difference is usually sufficient to Comparison with seismic methods of throw the voltmeter off scale on the sounding can be summarized briefly. more sensitive ranges, making readings The seismic method is more accurate, of AV impossible. more expensive, andthe equipment Other practical problems that arise necessary is more bulky. It is preferable can be summarized briefly as follows: for work on a thick ice-cap, being faster (a) Loss of current due to conduction at great depths and yielding more in- by meltwater streams. This effect is formation per shot. Geo-electric meth- minimized by appropriate positioning of ods can probably be used under a greater the profile. variety of conditions than seismic meth- (b) The difficulty in getting a current ods andare probably justas fast to to enter the ice. Current flow is gov- depths such as are found on a glacier. erned more by contact resistance be- Field work tween electrode and ice than by the The geophysical party and equipment path resistance in ice. Cold dry snow arrived on Devon Island on June 7. Be- makes a particularly bad contact. Salt fore going to the glacier on June 12, or brine improves all contacts consider- resistivity measurements were made on able and there should always be a sup- the ice of a lakeand on sea-ice near ply of one or the other. the base camp. Water samples of both (c)The resistance of bedrock. A were taken. knowledge of the possible rock types to The period from June 12 to July 9 be found below the ice is essential, was spent at the camp at the first fork especially if there are sediments whose of the glacier. During thistime, 24 meas- resistivity differs little from that of the urements were made, using the method ice. of Schlumberger. It was possible to (d) Crevasses, which affect measure- average only one sounding a day over mentsdirectly according totheir size a prolonged period because travel to and position relativeto the profile. and from the sounding points made it Areas of heavy crevassing cannot be difficult to complete two a day regularly. worked with this method. The ice in this general area had a value (e) Insulation losses. Great care must of resistivity throughout its depth of be taken to ensure complete insulation around 60,000 ohm-metres. Bedrock re- of the instruments from the ice and sistivity was of theorder of 400,000 INSTITUTE NEWS 263
ohm-metres, butat one point the ice Three areas of anomalously high re- was underlain by low-resistance silt in- sistivity were found on this part of the dicative of an old lake bed. A typical glacier. The first (with 2 megohm- curve on this section is shown in Fig. 4. metres) was on ice lying in an embay- On July 9 camp was moved to the upper ment of the cliff and not moving with
Fig. 4 Sample 10 I- I 0 , profile from lower part of 60.m Onm- Metres glacier. T4 I 0
n n n 4 1. I I I I I I I I I I II!!IIl 10 102 lo3
a ~ METRES
fork. On July 12 operations were moved the glacier, the other two occurred to the ice-cap station for a short time on ridges running near the glacier cen- and three soundings were made. These tre, at soundings 34 and 27. These ridges established a depth of ice of the order have almost vertical banding parallel to of 750 metres, and showed the order of their length. In both the effect occurred resistivity to be the same as that found in the top few metres. in Greenland (-60,000 ohm-meters). After Dr. Vogtli’s departure a week
Fig. 5. Sample profile from LAYER HIGH p upper part of OF glacier.
K GLACIER-ICE
I Io
2 9 1 I I I 10 102 lo3
a METRES
The period from July 19 to July 26, was spent at base camp during which when Dr. Vogtli left the island, was time resistance readings were taken in spent at the upper-fork camp complet- the bedrock under the base and on the ing soundings 25 to 36. The glacier in dolomitic sediments overlooking the this region has a characteristic layer of camp. The resistivity of the bedrock was iceof higher resistivity (of the order not typical, possibly because of sea- of 100,000 ohm-metres) from 10 to 40 water inflltration and high salt content. metres below the surface (Fig. 5). Its conductivity dropped off rapidly 264 NEWS INSTITUTE
with depth. The readings for the dolo- would be the existence of stress within mite should prove useful since dolo- the ice. This is backed up by two other mitic rocks underliethe ice-cap in observations; firstly, a plot of resistivity places and may underliethe glaciers across one ridge showed remarkable also. coincidence with the crack pattern; and On August 6 a team of U.S. Army secondly, a core taken at the point of Engineers took ice-depth readings at highest resistivity on this same ridge six points on the glacier using converted showed, according to the glaciologist, a radio altimeters. This method had been bubble patternclearly indicative of testedon ice-caps andhad not been stress. A water sample from this core previously used on glaciers. Complete has been sent to Dr. Vogtli for analysis results have not yet been obtained but along with his other samples. However, therewere difficulties involved in the not all ridges give a high value of p. No work, possibly because of the presence cores could be taken in other locations of water below the glacier surface. butthis should certainly be done in future. The time from August 7 to August 23 was spent at the upper-fork camp. The The tributary glacier examined in depth-finding program having been most detail was the “Piper Glacier”. A completed, an opportunity was taken to profile across the width of its mouth did investigate the regions of high resis- notreach bedrock, which was thus tivity discovered earlier. They were of shown to be deeper than 700 ft. More- interest for two reasons. Firstly, knowl- over, the top 100 metres showed a edge of the location and sources of high- resistivity averagingaround 200,000 resistivity ice could help understand its ohm-metres, dropping to thevalue of causes, and secondly, if such values are ordinary glacier ice below 200 metres. likely to arise in future depth-finding Aseries of readings taken along the interpretations, it would be useful to be centre line of the glacier to within 70 acquainted with the structures they are metres of the plateau at the top showed associated with and the area they will a great variation of resistivity ranging affect. Work was accordingly concen- from 600,000 ohm-metres at one point trated on the two ridges in which this to 60,000 ohm-metres at the same depth effect was found and on the glaciers atothers. No consistent pattern could from which it was suspected the ice be detected in the variation, using the might derive. limited number of points done. A Afairly detailed series of profiles water sample was taken from the area along a half-mile stretch of one such of highest resistivity; the core here did ridge together with readings spread out not show signs of stress as did the core along a four-mile stretch of the second, from the ridge. Although thisridge led to the following general observa- crossed directly in front of the glacier tions: mouth (and was in fact pushed up by thepressure of thetributary glacier (a) The high resistivity occurs in building up on the main stream of ice), stripsthat run parallel to the axis of no path of high-resistivity ice could be the ridges (and the glacier). The area found connecting areas of similar resis- of high resistivity lies withinthe top tivityon the two features. Chemical 10 metres of ice. analysis should yield evidence asto (b) The value of p in such a strip is whether or not thetwo phenomena stem not consistent along its length. If any- from the same cause. thing, the resistivity showed a tendency to increase down-glacier. An attempt to Two other pieces of work were done trace the strip containing sounding 27 for Dr. Vogtli. One was to confirm the back to a source or mother lode of some high value of p inthe embayment type was unsuccessful. mentioned earlier.The second was to (c) These facts tend to point toa check the isotropy of the banded ice phenomenon generated in these ridges that shows up on either side of the themselves, and the most obvious cause moraine on the upper section of the INSTITUTE NEWS 265 glacier. This involved doing Wenner an accuracy check for the geo-electrical andSchlumberger profiles at right work. angles and comparing the values ob- If possible 2 weeks will be spent at tained. the ice-cap summit, and near the firn- The period from August 24 to August line in mid-summer when conditions 28 was spenton the ice-cap. It was are more favourable. Firn can be dis- originally planned to do two long pro- tinguished from glacier-ice by seismic files at right angles witha view to methods but if geo-electric methods are obtaining the ice resistivities to a depth workable they would bequicker and of 1500 metres in an accumulation zone. much less expensive. It would be de- It was not expected thatthe bottom sirable in addition to co-ordinate resis- would be reached with the available tivitymeasurements on the glacier amount of cable. On reaching the sum- with the crystallographic work that is mit however, the CuSO, solution had being done. Being able to predict con- frozen, cracking two of the potential ditions below the surface by measure- electrodes; as a result copper electrodes ments on the surface would be a very had to be used. The instability of the valuable development of technique. metal electrodes on sensitive voltage Another field of interestis the ranges, coupled with insufficient current possibility of correlating the radiation due to the layerof cold dry snow allow- penetration in the upper layers of the ed reliable readings to betaken to a ice with the resistivity of the ice. It is depth of 100 metres only. For this, snow pure speculation as to whether or not had to be melted and asalt solution sucha relationship exists, butduring poured on the currentelectrodes at each the course of measurements this sum- point. The first 80 to 100 metres showed mer certain variations were noted in the an average value of 200,000 ohm-metres. upper metre or two of ice, which were If this value is typical of the firn, it possibly dueto excitation from solar should be easy to distinguish the fim radiation. from the older glacier ice underneath. During the winter the instruments However, since the firn value should and power supply are to be built into a merge graduallyinto the glacier-ice special lightweight and convenient unit value at depth, no clear-cut division or as a specialized instrument for this type two-layer interpretationcan be ex- of work. This equipment will include all pected. the refinements evolved during the 1961 season. There might be some benefit PIans for summer 1962 from carrying on a program of experi- Most field time in 1962 will be spent ments with radio altimetry equipment in depth sounding of the glaciers. It is and checking it against the geo-electric hoped that the ice-cap will be sounded and seismic work during the one oper- by seismic methods since it would be ation. much faster at the thickest parts, and a J. P. GREENHOUSE limited number of shots would provide University of British Columbia.