The Tectonics of the 'Great IceChasm', Filchner Ice Shelf Antarctica

by GILBERT WILSON

Received 1 May 1959; read 3 July 1959

ABSTRACT: The Commonwealth Trans-Antarctic Expedition, 1955-8, observed thatthe Filchner Ice Shelf was traversed by a straight east-west zone of brecciated ice-the'Great Ice Chasm'-about forty miles south of Shackleton. Analysis of the crevassepatterns at the eastern end of the Chasm indicates that the brecciated zone probablyresulted from a wrench- or tear-faulting movement, in which the northern side movedwestwards relative to the southern. The frictional drag caused by the strong westwardcurrent of the Weddell Sea on the base of the northern part of the floating ice-shelfis considered to have been largely responsible for the formation of the Chasm.

EVEN AS WE ARE often informed that truth is stranger than fiction, so is ituncommon for Nature to provide us with a clear, larger scale and morecomplete demonstration of some simple tectonic mechanism than we,ourselves, can produce by experiment. This, however, she seems to havedone at the eastern end of the 'Great Ice Chasm' which runs east and westacross the FiIchner Ice Shelf, about forty miles south of Shackleton,Antarctica (Fuchs & Hillary, 1958, Map 2, p. 100, and photograph No.6,between pp. 20--21). This Chasm, also referred to as the 'Grand Chasm',has a total length of about sixty miles (l00 km.), and a maximum width ofabout three miles (5 km.). 'The jumbled ice mass in the bottom lies 175 feet(54 m.) below the shelf level ... but the rupture must extend through theentire shelf. .. .' (Neuburg, Thiel, Walker, Behrendt & Aughenbaugh,1959, 114).

The photograph reproduced here in Plate 6A shows the relatively smoothsurface of the shelf broken by a long straight zone of intensely shatteredice which has a length, within the area shown, of at least fifteen miles(24 km.), It has an estimated width of 200 to 300 yards near its eastern end,and it gradually widens to a mile or more before it disappears beneath theclouds in the west. The near or eastern end of the Chasm is broken up intoa complex of crevasses which can be resolved into two main types orgroups. Those which continue the left or south margin of the Chasm arecurved and sigmoidal in shape; they are tight at the ends, and become wideror gaping in their central portions. Those to the right or north of theChasm form a system of straight-line fractures which show little variationin width along their lengths, except for a natural tapering as they die out.This system can in turn be divided into two sets: the more prominent

130

PRO C. GEOL. ASSO C., VOL. 71 (1960) PLAT E 6

A. Th e eastern end of the ' Grea t Ice Chasm' , looking west. Sigmoidal tension gas hes tothe so uth (left), and stra ight-line shea r fractures to the north (right). The length of theChasm shown is abo ut fifteen miles (Photo No . 2342)

B. The 'G rea t Ice Chasm', looking west. Str a ight-line shear fractures, loca lly curv inginto the direction of tension fra ctures, show as faint lines on the south (left) side of theCha sm. The SW.-NE. orien tat ion of the seracs in the Chasm can be seen in the fore­grou nd (Photo No . 2345)(Both photographs reproduced by kind permission of the Commonwealth Trans-Antarctic Expedition)

[To fa ce p . 130

THE 'GREAT ICE CHASM' 131

appears to trend roughly between east-north-east and north-east, and thesecond makes a large angle with it (Plate 6A). Parallel to the second set offractures are slight undulations or drifts which in places cross, or are cut,by other crevasses. Similar (1) drifts can be faintly seen on the surface ofthe Ice Shelf itself. Crevasses which are sub-parallel to the prominentfractures on the north side of the Chasm also occur on its south side, butas they approach the Chasm edge they curve in towards it. They show asfaint lines on Plate 6B. In addition the seracs within the Chasm itself canbe seen in the same photograph to have a general SW.-NE. trend. Thistrend is even more marked in other air-photographs taken by the Trans­Antarctic Expedition (Photo K.593 I).

The fracture pattern as a whole is monoclinic, and seems to conformclosely to the theoretical pattern that might be expected to develop if theChasm were a zone of wrench-faulting, on which the right-hand side(north) had moved to the west relative to the left-hand side (south) whichmoved to the east. This suggestion is primarily based on the impressionthat the sigmoidal crevasses represent tension gashes which had beeninitiated, opened up, and partially rotated by a continuous left-handedmovement along the main zone of fracture. The prominent straight-linecrevasses to the north of the Chasm would fit into the structural patternequally well as tensional fractures which have not been rotated, or asplanes of break in the ice shelf parallel to a direction of maximum shearingstress. Which is the more probable largely depends on the true angularrelationship between the crevasses and the line of the Chasm. It IS thereforenecessary briefly to consider the effect of foreshortening due to perspectiveand to the tilt of the camera, which together have yielded the apparentangles now seen between the crevasses and the Chasm in Plate 6, A and B.

If the position of the aeroplane relative to some fixed points on theground, its height, and the tilt of the camera were accurately known, itwould be possible to work out the true angles between the crevasses and theChasm from the apparent angles seen in the photographs. The height fromwhich the pictures were taken was about 1000 to 1500 feet above the ice­shelf, but the other necessary information is unfortunately unobtainable.Two possible lines of procedure present themselves. Firstly, one can make adiagrammatic plan of the Chasm with its crevasses shown at variousangles (Fig. 1B); then, from this plan one can construct a series of perspec­tive drawings corresponding to different positions of the aeroplane, andnote which drawing and which set of crevasse lines most closely accordwith the original photograph (Fig. 1A and Plate 6A). Alternatively, andmore easily, one can view the plan on the ground-glass focusing screen of aplate-camera whose height and angle of line of sight can be adjusted. Thismethod not only shows the plan in perspective, but enables one to makeallowance for the original tilting of the camera in the aeroplane. The

132 GILBERT WILSON

B.

Fig. 1. Showing diagrammatically the relationship of true angles to apparent anglesbetween the line of the Chasm and its crevasses, as seen in oblique air-views. All tracedfrom photographs.A. The eastern end of the Chasm with its crevasses, taken from Plate 6A. The dotted

line near the top indicates the general tilt of the horizon in the photographB. Diagrammatic plan of the Chasm with crevasse-lines drawn at different anglesC. and D. Oblique air-views of the plan B. In C the line of sight of the camera was

inclined at 25°. In D, it was at 35°, and the plan was tilted at T" to the left to give atilted horizon line as in A

model crevasse patterns can then be photographed. The set-up is verysimilar to that described by G. D. Hobson (1942).

Comparisons of the oblique air-views of the model with the actual planare instructive:

(i) The true angles between the lines representing the crevasses and the edgeof the Chasm are much smaller than the apparent angles seen in the obliqueair-views.(ii) The prominent crevasses in Plate 6A lie at an angle of much less than 45°to the Chasm, probably little over 20°. The corresponding straight-line

THE 'GREAT ICE CHASM' 133

crevasses to the south of the Chasm (Plate 6B) make an angle of not morethan 15° to the main zone of fracture.(iii) The apparent anglesin the obliqueair-view of a line having a true angleof 45° to the Chasm in the model, lie within a few degrees on either side of90°, as shown in Fig. 1, C and D.(iv) The short, less prominent straight-line crevasses in Plate 6A convergetowards a vanishingpoint far away to the left of the line of the Chasm. Theymust thereforestrike nearly perpendicularly to the Chasm itself,and at about70° to the more prominent set of crevasses.

The crevasse system to the north ofthe Chasm thus roughly corresponds,in its angular relationship to the main zone of fracture, to the 2nd Orderright and left wrench-faults which Moody & Hill (1956) find are associatedwith major transcurrent faults. However, the crevasse system as a whole,north and south combined, conforms even more closely to the variousfracture pattern which can be obtained in Riedel's well-known experiment(Riedel, 1929).

This experiment has been repeated and quoted by many workers (Hills,1940; Blyth, 1950; Shainin, 1950; Wilson, 1952; Cloos, 1955; de Sitter,1956; Kanungo, 1956, and others). It is simple to demonstrate, andconsists of two boards laid side by side and covered by a flat cake of wetclay. A shear is then imparted to the clay layer by moving one boardrelative to the other along their edges of contact. The stress system whichresults from this movement is shown in Fig. 2A. The moving boardimparts a shearing couple SCi, SCi, to the clay cake, in which the anti­clockwise torque of the couple must-if rotation does not occur-bebalanced by an induced secondary clockwise torque or couple acting atright-angles to the first, SC2, SC2. Such a stress system can be resolved intotwo principal stresses (Fig. 2B) of compression (PP) and tension (TT) atright angles to each other. Rupture of the clay cake may occur by tension,with the formation of gaping fractures (r) which strike at 45° to the mainline of slip and are normal to the tensional stress TT (Fig. 2 C). Alternatively(Fig. 2D), failure may occur along fractures more or less parallel to thedirections of maximum shearing stress-Sa and So; these theoreticallyshould lie symmetrically at angles of less than 45° on either side of thedirection of maximum compression PP. In Riedel's experiments failurealong both directions of shear did not develop, and the fractures formedonly at a small angle with the main line of slip and so corresponded to Sa.This angle rarely exceeds 20°. The shear-planes So, which should lie at amuch larger angle to the main line of slip than do the Sa-planes can seldombe developed. Both sets were however produced together by Ernst Cloos(1955), and in the Geological Laboratories of the University of Wisconsinwhen a block of limestone was fractured by rotational shear under con­siderable retaining pressure (Leith, 1923, fig. 11B; Wilson, 1946, fig. 52, c).Failure by both tension and shear together in the same experiment is

134 GILBERT WILSON

possible, and a good example of it has been illustrated by Cloos (1955,Plate 3).

The crevasse system at the eastern end of the Chasm suggests that failureby both tension and shear has occurred. The sigmoidal, gaping crevassesonthe left (south) side of the Chasm look like typical tension fractures, asillustrated by Shainin (1950), such as would be formed by a tensionalstress operating in the north-west and south-east quadrants. The two sets ofintersecting straight-line crevasses on the right-hand (north) side of themain zone of shattering show a fracture pattern that would conform to thedirections of maximum shearing stress if the ice were subjected to amaximum compression acting in a SW.-NE. direction, combined with a

~SC,

~ II

I II I,V <'

(J'"":.':.-:...-------- ---_"':.':.=~

VS Primary .... Plane lJ)I I

I II I

I {I

A.--- Sc, --,.. p

p

/

So

p

Sa

Fig . 2

A. The stress system developed in an element subjected to a couple SCI-ScI actin gcounterclockwise along a primary plane of slip

B. The resolution of the two couples SCI and Scs into principal stresses: T = Tension,P = compression

C. The development of ell echelon tension gashes (I ) at right angles to the tensionalstress (D and at 45° to the primary plane of slip

D. The theoretical direct ions of planes of shearing, Sa and So, at less than 45° to themaximum compressive stress, and their relat ionship to the primary plane of slip

THE 'GREAT ICE CHASM' 135

tensional stress acting in the opposite quadrants. The straight-line crevasseswhich lie to the south of the Chasm (Plate 6B) also conform to this stressdistribution. Hence, the orientations of the principal stresses on both sidesof the Chasm are in accord, even though the styles of fracturing of theshelf ice are different.

Fig. 3

A. The development of sigmoidal tension gashes; the sigmoidal form results from therotation of the inert material between the fractures while the ends of the fracturescontinue to develop at 450 to the plane of slip

B. The pronounced sigmoidal form of tension fractures developed partly by rotationand grading into planes of shear

The strong curvature of the sigmoidal crevasses is noteworthy. Shainin(1950)explained that the sigmoidal form of tension gashes is commonly theresult of a rotation, from the original 45° position, of the inert rock-matterlying between pairs of fractures while the fractures themselves are slowlydeveloping during a prolonged period of movement. It is only the materialbetween the actual fractures which is rotated and the tapering ends of thelatter continue to be extended at 45° to the primary direction of slip (Fig.3A). The tapering ends of curved crevasses of the Chasm, however, swingeven further around beyond the 45° angle, until they are in fact continuedin the straight-line crevasses which make a small angle with the line of the

136 GILBERT WILSON

Chasm on its southern side. It appears likely therefore that the strongcurvature of these sigmoidal crevasses has resulted from the combinedeffects of failure by tension accentuated by partial rotation and by failureby shear, the two appearing more or less simultaneously and grading oneinto the other (Fig. 3B and Plate 6B).

One gets the impression, from the photograph Plate 6A, that the centre­line of the zone of sigmoidal tension gashes is nearly colinear with thesouthern edge of the Chasm itself, and may mark its incipient continuation.This is further supported by the roughly SW.-NE. trends of the seracswithin the Chasm itself. These likewiseconform to the theoretical directionof tensional fracturing between the walls of the Chasm (Plate 6B, andFig.2e).

The remarkable straightness of the Chasm itself for some sixty miles,and the width of shattered ice which forms the feature, together with theevidence outlined above based on the monoclinic fracture patterns of thelocal crevasses, all indicate that the Great Ice Chasm, at least at itseastern end, is a tear or wrench fault, on which the movement was left­handed or sinistral. That is, the north side moved west relative to thesouth side.

The Trans-Antarctic Party'... believed this [the Chasm] to be causedby the movement of the ice shelf past an area of higher, snow-coveredground that had been observed from the air many miles to the west andsouth' (Fuchs & Hillary, 1958, 149) and the conclusions reached in thispaper support this suggestion. It was also observed that the pack ice of theWeddell Sea was continually moving westward despite the effectof variablewinds: 'It therefore seems that the current sweeping down the east coast ofthe Weddell Sea and along the edge of the ice shelf is strong enough toovercome the usual drift of the sea ice with the wind' (Fuchs & Hillary,1958, 42). Neuburgh et al. (1959) also noticed that this current jams thepack ice against the eastern coast of the Palmer Peninsula, which forms thewestern margin of the Filchner Ice Shelf, some 450 miles (720km.) west ofShackleton. This current, flowing towards the west, would by fractionaldrag on the base of the floating ice shelf simulate the moving board belowthe clay cake of Riedel's experiment, and the westward drift would tend toproduce a left-handed (sinistral or counterclockwise) couple in the icewhich is anchored somewhere in the south. The main concentration ofstress in the ice would be controlled partly by the location and shape of theanchorage in the south and west, and partly by the area of the underside ofthe floating shelf-ice exposed to the Weddell Sea currents. These arematters beyond the scope of the present paper. Nevertheless, the GreatChasm probably marks the zone which, in Riedel's experiment, correspondsto the line of junction of the two boards, though not necessarily to the'strandline' or boundary between the floating and grounded ice.

THE 'GREAT ICE CHASM' 137

For once it seems that in Tectonic Geology we are given definite evidenceof the direction and the nature of the motive force responsible for thestructure. It is of interest, not to say satisfaction, that the movement onthe Great Ice Chasm of the FiIchner Shelf, as deduced from the pattern ofits crevasses, corresponds in direction and sense to the movement of theWeddell Sea current, which was probably the cause of the effect that can beseen.

ACKNOWLEDGMENTS

The preparation of this paper has been greatly assisted by the interest andencouragement which I have received from members of the CommonwealthTrans-Antarctic Expedition. Sir Vivian Fuchs most kindly supplied thephotographs here reproduced in Plate 6. Both he and Dr. Harold Listerread over a preliminary draft of the paper and suggested its publication,while Dr. Jon Stephenson has discussed various points with me by air-mailfrom Lahore, Pakistan. To Dr. J. F. Nye, of Bristol I am most grateful forhelpful criticism and comments. I have also been encouraged and assistedby Professor H. H. Read, F.R.S., who is a member of the Committee ofManagement of the Commonwealth Trans-Antarctic Expedition, and bymy colleagues at the Imperial College.

DISCUSSIONSIR VIVIAN FUCHS: 'May I say how pleased I am to have heard Dr. Wilson's explanationof the formation of The Great Ice Chasm. Unfortunately we did not think of this whenwe were on the spot or we would have taken vertical air photographs and thereby madehis interpretation easier.

'There are two small points upon which I might comment. First, there is a considerablemass of ice pouring off the Touchdown Hills towards the west, which must exert agreat force upon the northern part of the Filchner Ice Shelf. Second, the ice shelf (1300feet thick) almost certainly has considerable relief on its undersurface which wouldallow the strong and constant current to exert a greater thrust to the west than if thesurface was smooth. Both these facts support Dr. Wilson's argument.'

DR. GILBERT WILSON, in a written reply, took the opportunity of again thanking SirVivian Fuchs for the interest he had shown in the paper. The fact that a large mass ofice was pouring off the Touchdown Hills, which lie to the north-east of the eastern endof the Chasm, would certainly tend to accentuate the shearing stress in the shelf-icein that area. If the crevasse systems pictured in the photographs were confined to theeastern end of the Chasm, it might even be considered that the moving ice from theTouchdown Hills was, itself, responsible for the fracture pattern on the north side ofthe Chasm. That this ice could be responsible for the Chasm as a whole seemed veryunlikely; and Dr. Wilson was glad to learn from Sir Vivian of the irregular relief of thelower surface of the Filchner Ice Shelf, as this supported the hypothesis outlined in thepaper.

REFERENCESBLYTH, F. G. H. 1950. The Sheared Porphyrite Dykes of South Galloway. Quart. J.

geol. Soc. Lond., 105, 393.CLOOS, ERNST. 1955. Experimental Analysis of Fracture Patterns. Bull. geol. Soc.

Amer. , 66, 241.

138 GILBERT WILSON

FUCHS, SIR VIVIAN & SIR EDMUND HILLARY. 1958. The Crossing of Antarctica, 338,London.

HILLS, E. S. 1940. Outlines of Structural Geology, London.HOBSON, G. D. 1942. A Method of Constructing Perspective Diagrams of Mine

Workings. Trans. Inst, Min. Met., 51, 311, London.KANUNGO, D. 1956. Structural Geology of the Torridonian, Lewisian and Moinian

Rocks of the Area between Plockton and Kyle of Lochalsh, ill Wester Ross,Scotland. Ph.D. Thesis, Univ. of London.

LEITH, C. K. 1923. Structural Geology, New York and London.MOODY, J. D. & M. J. HILL. 1956. Wrench-Fault Tectonics. Bull. geol, Soc. Amer., 67,

1207.NEUBURG, H. A. C., E. TmEL, P. T. WALKER, J. C. BEHRENDT, & N. B. AUGHENBAUGH.

1959. The Filchner Ice Shelf. Ann. Assoc. Amer. Geogr., 49, 110.RIEDEL, W. 1929. Zur Mechanik geologischer Brucherscheinungen. Zbl. Miner. Geol.

Paldont., Abt. B (1929), 354.SITIER,L. U. de. 1956. Structural Geology, New York and London.SHAININ, V. E. 1950. Conjugate Sets of en Echelon Tension Fractures in the Athens

Limestone at Riverton, Virginia. Bull. geol. Soc. Amer., 61, 509.WILSON, GILBERT. 1946. The Relationship of Slaty Cleavage and Kindred Structures to

Tectonics. Proc. Geol. Ass., Lond., 57, 263.---. 1950. A Quartz Vein System in the Moine Series near Melness, A'Mhoine,

North Sutherland, and its Tectonic Significance. Geol. Mag., 89, 141.