aerial survey of graham land* - asprs
TRANSCRIPT
Aerial Survey of Graham Land*
The Falkland Islands & DependenciesAerial Survey Expedition 1955-57
P. G. MOTT,
Technical Director, Hunting Aerosurveys Ltd.
ABSTRACT: The paper describes the recent expedition during the twoAntarctic summers of 1955/56 and 1956/57 carried 01tt by HuntingA erasurveys Ltd. under contract to the United Kingdom Government.Employing two Canso amphibean aircraft and a ship-based helicopter,the expedition completed full vertical air photo coverage of an area of35,000 square miles over the peninsula of Graham Land and adjoiningarchipelago in the British sector of the Antarctic.
Using a 950-ton freighter converted as a helicopter base, parties ofground surveyors were landed along the coast and islands of GrahamLand, where they completed over 200 miles of fully-observed triangulationto serve as ground control for the preparation of photogrammetric mapscovering the area photographed .. these are now under preparation by theDirectorate of Overseas Surveys.
The author, who organized and led the expedition, describes theequipment and methods employed for both photography and groundcontrol, with special reference to the photogrammetric problems involved.
l ITHE past quarter of a century airphotography has played an ever-in
creasing role in the exploration and mapping of the Polar regions, so much so thatit is a commonly held belief among thegeneral public that one has only to point acamera out of the side of an aeroplane flying over unknown territory to be able toclaim that it is mapped. As was laterproved in the case of some of the earlyflights across Graham Land, aerial photographic reconnaissance alone may lead toconclusions that result in major topographical inaccuracies in an area whereglacier and sea ice often merge togetherwhen viewed obliquely, and where successive features all encrusted in an over-all icesheet are so numerous and so alike as tohave a common resemblance. This statement is not, of course, to discredit the highspirit of enterprise and the remarkableachievements of Ellsworth and others whodid so much to pioneer the way for more detailed studies to follow.
The first serious attempt to carry out aphotogrammetric survey in the Arctic was,I believe, that of the Danish Government
in Greenland, where a regular programmeof air photography coupled wi th groundtriangulation first began in the early1930's and has since resulted in the production of a highly accurate series of1/250,000 maps covering nearly the entirecoastal strip of the world's largest island.
In the Antarctic the later expeditions ofAdmiral Byrd and the even more recentOperation Deep Freeze have carried outmainly trimetrogon photography coveringa very large part of the main continentalmass which, by itself, is probably thegreatest single contribution to our knowledge of the Antarctic that has yet beenmade. I believe I am correct in saying,however, that the great majority of thesephotographs are as yet entirely unrelatedto any form of ground control and musttherefore be considered of limited value formapping purposes until some form of control becomes available.
The Antarctic expedition which is thesubject of this paper was, I believe, thefirst attempt at one and the same time toprovide the desired elements of the photogrammetrist-vertical photographs tied to
* Presented at the Society's 24th Annual Meeting, Shoreham Hotel, Washington, D. c.,March 26, 1958.
415
416 PHOTOGRAMMETRIC ENGINEERING
P. G. MOTT
a pattern of identified ground points, theheights and positions of which were surveyed to wi thin plottable accuracy. I t alsohappened to be the first occasion in Antarctic history when a major survey of this
kind was placed on contract to a civiloperating company.
The Graham Land peninsula (or Palmerpeninsula, as you prefer to call it in theUnited States) extends for 800 miles fromthe main mass of the Antarctic continentand is divided from South America by 600miles of tempestuous sea known as DrakeStrait. (Figure 1). A central plateau of icebetween 4,000 and 8,000 feet above sealevel forms the backbone of the peninsula,which descends abruptly on either flank ina complex system of mountain ridges divided by innumerable glaciers and deeplyindented by fjords. The eastern coast ofGraham Land is permanently blocked bypack ice that is compacted under greatpressure by the prevailing easterly gales.During the summer months the westcoast on the other hand is usually navigable to ships as far as latitude 68°S. In 1908Britain formally laid claim to GrahamLand as part of the Falkland Islands Dependencies. In the course of her administration of the area Britain has set up anumber of permanent bases along the coastand surrounding islands for the purpose of
King:Gecrgel:--
Miles
S~iJ~~llF--
~~T~L=~~N~~I~-
A.]. Thornton
FIG. 1
AERIAL SURVEY OF GRAHAM LAND 417
FIG. 2. Canso in flight overGraham Land coast.
providing a meteorological reporting service and to carry ou t various forms of scientific research, including survey. Due to theextremely difficult nature of the terrainand the limited working season, the extentof the mapping carried out by these groundparties was necessarily slow and by groundmethods alone it would take a very longtime indeed to produce even a moderatelyaccurate map covering the whole area. Inthe early part of 1955 therefore the UnitedKingdom Colonial Office decided to putinto effect a plan that had been maturingfor some years, for an aerial survey of themore important parts of the area.
The combined programme of air photography and ground-control covering anarea of approximately 40,000 square mileswas put out to tender. In May 1955 thecontract was awarded to Hunting Aerosurveys Ltd.
Our plan was to establish a base for twoCanso amphibian flying boats at Deception Island, in the South Shetlands. The island is shaped like an atoll with a rim ofice-capped mountains containing an innerharbour, Port Foster, which is four milesacross and seven miles in length. Duringthe summer months this unique harbourbecomes free of ice and forms the safestnatural anchorage in the whole of Antarctica. Deception Tsland, which owes itsname to a single narrow entrance known asNeptune's Bellows, that can easily bemissed by a passing ship, is of volcanicorigin and is still subject to thermal activ-
ity in the forms of hot springs and occasional tremors.
Our choice of aircraft was limited by thefact that no airstrip existed anywhere inthe area from which a land plane with therequired endurance and capacity couldhave been operated. A pure flying boat, onthe other hand, would have been seriouslyendangered by ice floes drifting into theharbour from the open sea or from theterminal ice-cliffs of the glaciers descending from the mountains encircling PortFoster. By employing an amphibian wecould beach the aircraft ou t of danger ofice and at the same time greatly ease thework of servicing and refuelling. TheCanso has a range of 2,000 miles with anoperational ceili ng for photographic workof 13,500 feet; moreover, its rugged construction and proven reliability made itwell adapted for a project of this nature(Figure 2). As an insurance against loss ordamage, and in order to take the fullest advantage of the limited season of goodweather, we took two aircraft which werespecially modified in Canada by KentingAviation Ltd., and flown down via SouthAmerica and the Falkland Islands to theexpedition base, a transit of some 9,000miles.
For the trans-shipment of stores and expedition personnel (with the exception ofthe Canso crews, who flew down with theaircraft) we chartered a 950-ton Danishfreighter, the "Oluf Sven." Before departure the ship underwent extensive modifications in order to provide accommodationfor 22 passengers, in addition to her normal complement of 14 crew. A steel bulkhead was constructed separating the forepart of the ship, to enable her to carry60,000 gallons of high octane spiri t. Shehad also to be fitted with a prefabricatedhelicopter deck, which during the voyagesou th was carried as part of the cargo inher holds, and could be reconstructed onarrival at the expedition base after unloading was complete. In addition to hernormal navigational aids, the ship wasfitted with Decca radar, an Anschutzgyro-compass and echo-sounding equipment reading to a depth of 300 fathoms.On the initial outward voyage he carrieda total cargo of some 650 tons, j ncludi ngtwo crated helicopters, a 60-foot long prefabricated hut, a complete radio stationand literally hundreds of items of storesvarying from aero engines and a 4-ton
418 PHOTOGRAMMETRIC ENGINEERING
tractor to sacks of coal.Directly the base installations at De
ception Island were ready for the operationof the Cansos, we planned to use the shipas a mobile platform from which we couldfly off parties of surveyors from the deck onto selected points on shore, thereby overcoming the major obstacles of terrain andweather. We believed that by using helicopter transport it should be possible toprovide a close network of triangulatedcontrol connecting the mainland of Graham Land and the coastal islands. Although no similar operation had been attempted before in Antarctic conditions,helicopters had already been used for control surveys in the Canadian Arctic, andwe felt confident that the same techniqueswould prove equally successful in GrahamLand.
The area to be photographed coveredthat part of the Graham Land peninsulabetween 62° and 68° latitude S., includingthe South Shetland Islands and the PalmerArchipelago lying west of Graham Land.
The specification called for completevertical coverage over the whole area to besurveyed. In addition, a continuous seriesof fixed side obliques was to be taken withthe camera axis depressed 30° from thehorizon tal.
For the vertical cover we used a \iVilliamson Eagle IX camera fitted with theRoss 6-inch lens on a 9-inch format negative, and for the obliques the WilliamsonF.24 fitted with a lens of 3l-inch focallength on a 5-inch square format. The timeinterval for the obliques was set to give a10 per cent overlap in the foreground ofsuccessive pictures, while the wide-angleof the lens provided for a generous lateraloverlap between vertical and oblique,which incidentally made the identificationof the obliques far easier than it wouldotherwise have been. Only one obliquecamera was used at a ti me, the normalprocedure being to leave the camera ineither the port or starboard mounting, sothat when the aircraft was flying in anortherly direction the obliques would betaken say to the east, and on the adjacentstrip (flown south) the camera would bedirected to the west.
The purpose of the obliques was to assistin the interpretation and identification offeatures, to help in the study of ice conditions and sledging routes to the plateaufrom the coast, and with the possibility
FIG. 3. Eagle IX camera and mountingin Canso.
that in some areas they might be used tocarry out reconnaissance mapping beyondthe area of vertical cover. I n cases wherean area was already well covered with obliques, we did not therefore continue touse the oblique camera in subsequent vertical sorties.
The camera compartment was locatedimmediately aft of the step in the hull ofthe aircraft, where it was normally free ofcontact with the water. The mounting ofthe Eagle IX was fixed to a sliding carriage, so that it could be moved forwardleaving the opening in the hull to be filledby a hinged port which, when lowered andbolted down, provided a water-tight hatch(Figure 3). The design and manufacture ofthis installation was carried out by Kenting Aviation in Toronto and proved highlysatisfactory. For navigation the nose compartment was fitted with the bomb-aimer'sglass window, and a special hatch with aremovable wa ter- tigh t cover was providedto allow a prismatic drift sight to be projected below the hull during flight.
Considering we had only five months inwhich to organize and equip the expeditionfrom scratch, it took some pretty hardwork by all concerned to ensure that nothing was left behind. It was only throughthe whole-hearted assistance of Government and Service departments, the combined resources of the Canadian andBritish companies of the Hunting Group,and the practical help and advice of a verylarge number of manufacturing firms, consultants and friends, that this preparatory
AERIAL SURVEY OF GRAHAM LAND 419
FIG. 4. Landing steel plate forconstruction of slipway.
work got done in time. We were also exceptionally fortunate in being able to offload the whole of the highly complex taskof radio communications including radioand radar aids to Messrs. rnternationalAeradio Ltd., while responsibility for thesupply and operation of the helicopterswas undertaken by Messrs. Autair Ltd.,and the extensive refitting of the ship bythe Mercantile Dry Dock Company ofJarrow.
The expedition ship "Oluf Sven" finallysailed from London Docks on 21st October1955, and after calling at Montevideo tobunker and load our total supplies of aviation fuels reached Port Stanley in theFalkland Islands on 27th November. Fromthere she sailed to Deception Island and inthe early hours of Saturday, December 3rd,we dropped anchor in \iVhalers' Bay, opposite the wreckage of the old whaling factory, which was abandoned a quarter of acentury ago on the introduction of pelagicmethods of whaling. In the centre of thebay the ice-covered slopes forming part ofthe main mountain ridge of the island giveway to a broad, gently sloping beach ofvolcanic ash. Although too soft to bear theweight of a heavy aircraft, we were laterable to lay some 20 tons weight of perforated steel plates to form a metal slipwayextending well below low water and providing a complete taxying circuit on landfor the two aircraft (Figures 4 and 5).
On our arrival the only habitable building left of the whaling station was thelarge hut occupied by a small party of the
Falkland Islands Dependencies Surveys,who are permanently stationed therethroughout the year. Through their assistance our shore party were able to findsleeping accommodation in the loft of theF.LD.S. hut and we were also able to takeover another partly dilapidated hut nearthe shore, which we renovated and used asa radio station. The new hut which we hadbrought down with us on the ship was designed to provide a generator room,kitchen and canteen at one end, while theremainder was divided into laboratory accommodation for processing the aerialfilms.
There was, of course, no proper jetty orfacility of any kind for landing our equipment. Some 350 tons of stores and equipment had therefore to be landed by meansof a military type of pontoon raft towed bya motor-boat. Individual loads up to sixtons including a large tractor, aero engines, hut parts, aviation fuels and oils,and several thousand different items weresuccessfully ferried ashore usi ng thisequipment, without loss or damage. Whilethe unloading continued the rest of theparty was engaged in the construction ofthe prefabricated hut, in setting up theradio station and in providing the essentialinstallations in readiness to receive theaircraf t.
Deception Island has a reputation (wellearned!) for orographic cloud which hangsround the mountain rim and, when thetemperature drops, sinks into the central
FIG. 5. Canso aircraft on slipway atDeception Island.
420 PI-IOTOGRAMMETRIC ENGINEERING
"crater" of the island, descending at timesto sea level. Since the southern limit of thearea to be photographed was 400 miles distant from our base and there was no alternative haven, it could happen that an aircraft, having taken off in good weather,might return from a long sortie to find thewhole landing area smothered in cloud. Tooffset this danger we had to provide twoforms of approach aid; the first a radio MFbeacon, and the second a radar deviceknown as Rebecca-Eureka which gives thepilot both bearing and distance during approach. For long-range communicationsbetween the base, the aircraft and the shipwe installed an HF transmitter of 300watts output. Finally, to provide localcommunication between ship, motorlaunch, aircraft, field parties, helicopterand control-base, a 50-watt VHF systemwas installed at the base with a IS-wattset on the ship.
A land line from the radio hut was laidfor two miles up to the highest poin t of thecliffs commanding "The Bellows," theentrance to Port Foster, in order to ensureunrestricted response to the Rebecca unitsof an approaching aircraft in thickweather.
All this preliminary work occupied thewhole of December and it was not untilJanuary 7th that we were at last able tosignal the two aircraft waiting in the Falkland Islands that we were ready to receivethem. By this time, however, the briefseason of good weather had already passedand almost to the day with the arrival ofthe planes there began three weeks of continuous high winds, cloud and snow. Itwas not, therefore, until early in Februarythat the first productive flights took place,when several islands of the South ShetlandGroup and some coastal strips along theGraham Land peninsula, including thewhole of Trinity Island, were sucessfullycovered.
On these early flights we learned a gooddeal, which proved of untold value in thefollowing season. The fact that the ceilingof the aircraft was limited to 13,500 feetover terrain which often rose with appalling abruptness from sea level to 7,000 feetand over, in a distance of only one or twomiles, made it necessary at times to increase the lateral overlap at the start of arun to 60 per cent, and even then it waspossible to get a gap only a few frameslater. Forward overlap in such areas gave
similar difficulty and required a very highconcentration on the part of the cameraoperator, the interval dropping as low as14 seconds over high ground, comparedwith 35 seconds over average terrain.
Navigation was, of course, the greatestproblem, particularly over the centralplateau of Graham Land, which in placesis a featureless expanse of snow brokenonly here and there by a line of crevassesor an occasional line of nunataks. As wewere working at the end of the operationalseason, we had to take the fullest advantage of the few remaining days, and severalof the islands were flown in one sortie without resorting to navigational strips. Alongthe mainland we reverted to the system offlying long control strips up to 120 miles inlength employing a gyro-controlled D.R.compass combined with computed drift tomain tain course. On return to base, thesecontrol strips were immediately processedand printed overnight, every alternateprint being stapled to the next in a continuous roll which was subsequently usedby the navigator to track the adjacentstrips on either side. By this means therewas no need to rely on the navigator'smemory, and even in areas of limited detail straight flying and lateral overlapcould be maintained.
Another matter on which we had to gainexperience was exposure and development.All photography was carried out on IlfordHyperpan film using a minus blue filter(comparable to Wratten G). Stop F.llwas used throughout with a normal exposure time of 1/200 second, though in thelatter part of the season this was increasedto as slow as 1/50 second, and also in someof the long sorties carried out in the secondseason, when photography was continueduntil 8 o'clock in the evening.
Over the high plateau, the glaciers, theshelf-ice and ice-domed islands, the unrelieved sameness of texture and tone require the most exacting assessment of development, without which variations in relief become indistinguishable, whereas inthe areas of high relief there is an excessiveamount of contrast between the brillianceor reflected light off the snow comparedwith vertical faces of exposed rock and thedeep shadows cast by the mountains. Allalong the coast lie literally thousands ofrocky islets washed by the sea, which require entirely different treatment to theadjacent ice-cliffs and glaciers. For de-
AERIAL SURVEY OF GRAHAM LAND 421
With the assistance of a teleprinter, thefullest meteorological reports available
21
1956/5712
63
Nil
24
DecemberJanuaryFebruaryMarch
neers managed to fabricate new parts toreplace those damaged, which enabled theCansos to be flown out. 'liVe had our lesson,however, that any attempt at photographic flying beyond the end of Februaryis not only a waste of effort but mighteasily result in the aircraft being unable towithdraw before winter sets in.
I n the following season the "01 uf Sven"returned to Deception on 26th November.Compared with the enormous labour involved in setting up the shore installationsthe previous year, the re-opening of thebase in November 1956 was a relativelysimple matter, and was completed inunder a week.
On their way south the two Cansos spentsix weeks in the Falkland Islands, wherethey managed to obtain stereo coverage at1/25,000 scale over the whole of the islandgroup, an area of some 4,500 sq uare miles.The weather in the Falklands is notoriously cloudy, and the area subject to almost continuous winds of very high velocity. The angle of drift on occasionsreached 29° and in a single run could varyfrom 10 degrees starboard to 10 degreesport, while the time interval changedfrom 30 seconds to 90 seconds betweensuccessive runs. Despite these difficulties,a remarkably high standard of flying wasachieved, and in the only eight days ofclear weather over 5,000 effective negatives were exposed.
Following this good start, the aircraftflew the remaining 800 miles to Deceptionon 9th December. On the way they encountered some severe icing, and a largechunk of ice thrown off the propeller of oneaircraft narrowly missed one of the crewwhen it tore an 8-inch hole in the perspexroof above the navigator.
The weather in the second season followed very closely that of the precedingyear, the number of days suitable forphotographic flying showing an almostexact parallel, as can be seen from thefollowing table:-
1955/561257
Nil
velopment of the first Eagle IX filmKodak 19B was used, but was found togive such high contrast that detail in thesnow areas was lost. By trial and experiment it was found that 076 gave a muchsofter negative and by varying development times between 20 and 40 minuteseven the most contrasting detail could berecorded on the negative. The obliqueswere found to require entirely differenttreatment to the verticals. 076 wouldbarely produce an image on the obliques,even after 50 minutes of development, andthe best results were obtained using D19Bat quarter strength with an average development time of 15 minutes.
The facilities provided for processing thefilms in the expedition hut could scarcelyhave been bettered. The darkroom wasdivided into two compartments, one halfbeing used for development of the filmsusing standard Morse units, and the otherfor contact printing. A third room contained two rotary film drying units andone large print dryer with shelving and alarge table for sorting and checking cover.The power required for working theequipment was provided by two Enfield6.5 k.w. diesel generators working on independent circuits. Water, of which we often required 500 gallons a day, waspumped from an existing catchment "well"into a series of old aircraft fuel tanks builtin between the rafters of the hut, wherethe water was stored at maximum headand was prevented from freezing by thewarmth of the rooms below. Luckily, during our whole stay at Deception the socalled "well" never dried or froze up, possibly due to the influence of thermal heat.At the warmest period of the summer wewere also able to use the local melt streamsfor washing the rough prints.
Due to the lateness of the season, from12th January until 25th March when theaircraft left, only eight days occurredwhich were in any way suitable for verticalphotography, and of these only one daywas sufficiently clear to permit both aircraft to carry out a productive sortie. ByMarch, conditions both as regards weatherand photographic light had rapidly deteriorated. In a violent gale on the nightof 13th March, both aircraft receivedsevere damage to their ailerons whilepicketed on shore. Using the workshop onboard the "Oluf Sven" and some iron salvaged from the whaling factory, the engi-
422 PHOTOGRAMMETRIC ENGINEERING
from South America, the Falkland Islandsand F.LD.S. stations along the GrahamLand coast were clu.-ing the pe.-iod of flyingoperations received daily by DeceptionRadio (as it became known), and plottedby the expedition met. off-iccr on synopticcharts.
Within two days of their arrival bothCansos were airborne on their first productive sortie, and in the following six weeksa total of 238 hours was flown by the twoaircraft. On one particularly good day(26th December) each aircraft was airborne for 11 hours, the total then coveredamounting to 3,000 square miles withnine full magazines exposed.
The visibility on days of still clearweather was quite exceptional. I will always remember accompanying the aircrafton one of the long flights to the southernmost limit of the area. As we flew down theeast coast of Adelaide Island the massivepeaks of Alexander Land some 200 milesfurther south stood out clearly against acloudless sky, and it was difficult to resistthe temptation to fly on instead of usingup our spare fuel in tracking back and forthalong the photographic strips. Seen fromthe air it was difficul t to appreciate thetremendous size of the landscape thatextended below us for hundreds of mileswith a brilliance so dazzling that one couldscarcely look at it with the naked eye.Fifty miles distant to the east the central plateau of the peninsula at a heightof 8,000 feet above sea level dropped in oneviolent step towards the Weddell Sea thatmust have been a sheer cliff face of 3,000feet in depth, but appeared no more thana spade-cut in the snow. Many of the fjordswere remarkably free of ice and mirroredthe images of the surrounding peaks.Lallemand Fjord, to the south of DarbelBay, was however completely blocked withclose pack, as was the strait separatingHanussa Bay from Lambeuf Fjord. Thelimit of the main continental pack wasclearly visible, lying at approximately lat.69°S. We estimated on this flight that theheight of Mt. Gandry, the highest summitof Adelaide, which had been shown at7,600 ft. on some existing maps, was infact over 10,000 feet; this makes it thehighest peak in the whole area.
By the end of ] anuary the bulk of theair photography had been completed. Inthe meantime we had been honoured by avisit from the Royal Yacht "Britannia,"
bringing H.R.H. the Duke of Edinburgh,who carried out an inspection of the aircraft and installations at our Deceptionbase, and took a deep and penetratinginterest in all the aspects of our work.
The photographic operations were finally closed down on 28th February and theaircraft departed for Port Stanley on 4thMarch.
The areas photographed by verticalphotography with full stereo coverage ata mean scale of approximately 1/20,000included all the South Shetland Islandsincluding Elephant and Clarence Islands,]oinville and D'Urville Islands and thewhole of Trinity Peninsula as far as 63°S.The whole of the Palmer, Danco, Grahamand Loubet Coasts, the whole of the Palmer Archipelago and Biscoe Island groups,the eastern half of Adelaide Island andPourquoi Pas Island were also photographed. Only the western half of Adelaide Island and the high plateau lyingeast of the Danco, Graham and LoubetCoasts still remain uncovered by verticalpictures, but with the assistance of theobliques it should be possible to carry outprecise photogrammetric mapping of 95per cent of the specified area.
A total of 35,000 square miles of verticalcover was flown in the two months ofoperations. 113 vertical and 23 obliquefilms, totalling 12,000 effective negatives,were processed by our laboratory unit oftwo photographers, with what additionalassistance they could muster. In additionthey produced some 17,000 contact prints.
GROUND-CONTROL
Almost the entire coastline of GrahamLand and surrounding islands is fringedwith ice-cliffs 50-150 feet high, formedeither by glacier snouts or what is termedPiedmont ice, which clings to the steeprock faces (Figures 6, 7 and 8). Inland,except for al most perpendicular cliffs,there is virtually no ice-free land and thereis practically nowhere one can go which isfree from the danger of hidden crevasses(Figure 9). The outer islands are in mostcases as steep and unapproachable as themainland and are moreover protected byuncharted reefs and treacherous currents.The weather is seldom fine and always unpredictable, while the wind can changefrom calm to hurricane force in the space ofand hour. Generally speaking, the furthernorth you are on the Graham Land coast
AERIAL SURVEY OF GRAHAM LAND 423
FIG. 6. Rock buttress on Liege Island withmainland of Graham Land beyond.
the less settled is the weather, and in thenorthern areas the sea ice cannot be reliedon for sledge travel, even in the depths ofwinter. In the past here and there a partyhad been landed by boat and observed ahurried astra-fix, and there had been one ortwo pockets of triangulation carried out inthe vicinity of the F.LD.S. bases.
By using a ship-based helicopter, wewere able to overcome practically all theobstacles which had prevented previousparties from gaining a foothold, and at thesame time to make use of every scrap ofgood weather, even if it lasted only acouple of hours.
In the first season we took out two 5.51Sikorsky helicopters stripped down intotheir main component parts and carriedboxed in the hold on the outward voyage.It was not until the aircraft operations
FIG. 7. Glaciers of Graham Coast. Fixedoblique photo taken with F.24 Williamsoncamera. 3J:" f.l on 5" X 5" forma t.
FIG. 8. Ice plateau of Graham Land withglaciers descending towards west coast. Fixedoblique taken with Williamson F.24 Rose 3t"lens from 13,500 ft.
were under way, and all cargo had beendischarged from the ship, that we wereable to reconstruct the flight deck andassemble the helicopter in readiness forflight. In the meantime the surveyorsmeasured a precise base-line of 3,000 feetalong a straight volcanic beach on thewest shore of Whalers' Bay. Although thebase terminals were 10 feet above highwater a considerable length of the base laybelow the tide line, and measurement hadtherefore to be carried ou t wi thi n theperiod of 4~ hours between ebb and flow.
FIG. 9. Creva<;se patterns on glacier-PalmerCoast (W. Graham Land). Eagle IX camera,Rose 6" lens.
424 PHOTOGRAMMETHIC ENGINEERING
For the alignment of the base we drovein pegs at approximately 100 ft. intervalswith a hardboard slat nailed to the topof each peg, the pegs being lined up, ofcourse, by theodolite. Prior to staking outthe line, the bulldozer was used to clearthe beach of whalebone and flotsam andthe line of measurement carefully graded.Final measurement was made three times,using a 100-ft. Invar tape on the flat underthe calibrated tension of 20 lbs. The workwas carried out with a team of six men,two at either end of the tape, one. on thetheodolite and the sixth man standing inthe middle of the tape with a whistle. Theends of each tape length were simultaneously scratched with a sharp knife on thehardboard, the recorders being given theusual warnings, ending up with the whistle.After applying slope and temperature corrections, the agreement between three independent measurements of the base cameou tat 1 in 225,000.
From this initial base a well-bracedtriangulation net was extended to coverthe whole of Deception Island, incl udingthe two highest peaks of the island, Mt.Kirkwood and Mt. Pond, and then acro~s
to points on Snow and Livingston Islands.The helicopter was used to lift a number ofempty 44-gallon fuel drums, which wereslung on a rope below the fuselage anddropped on the survey point. Using thedrums we were able to build temporarybeacons up to 9 ft. high and 6 ft. thick,which were clearly visible at a distance of30 miles away (Figure 10). We had hopedthat these beacons might have become encrusted with ice and drift and become apermanent feature of the landscape, butas a result of successive thaws and veryhigh winds they were usually carried awayin the course of a few weeks, and will haveended up half-way down the mountainside at the bottom of a crevasse. Nevertheless, they served their purpose very wellindeed and though we tried other types ofbeacon, the barrels proved the most reliable.
The early sorties with the helicopterwere an illuminating, if at times a rathersobering experience, for those taking part.A climb that had previously taken usseveral hours to make on foot would beaccomplished in the same number of minutes by this remarkable form of transport.Not only is it possible by this means forthe surveyor to seize every opportunity
FIG. 10. 5.51 helicopter landing barrels on amountain top to act as a signal for surveyors.
offered by short breaks in the weather, butit enables him to arrive at his destinationunfatigued and free of the chilling effectresulting from getting overheated on theclimb. By the same means it was possibleto reconnoiter in a 2-hour f1igh t an area of400 square miles, or to beacon in one dayof good weather 3 or 4 stations separated10 to 20 miles apart.
Using the air photos, wherever available,the points to be occupied were selected toconform to the following five requirements:-
1. Between 10 and 25 miles apart andinter-visible with other connectedpoints in the scheme.
2. Not higher than 1,500 feet abovesea level, in order to be below theaverage height of stratus cloud whichpredominates in the area.
3. On or near a satisfactory helicopterlanding site.
4. Having a practicable route down tothe coast in case of emergency in theevent of failure of the helicopter.
5. Identifiable on the air photos (thishad to be waived on several occasionsbut could be overcome by artificialbeaconing followed by low-level airphotography when necessary).
The ship would proceed to the area for avisual reconnaissance first from the deck,followed by a flight with the helicopter ifcondi tions were sui table. The normalduration of flight from deck take-off tolanding at the trig. station was usually
AERIAL SURVEY OF GRAHAM LAND 425
FIG. 11. Bell 47 D helicopter landing topick up surveyor.
about three minutes, one surveyor beingcarried at a time (Figure 11). On the firstjourney would be taken essential survivalkit including 10 days' food, a tent, campand climbing equipment and a short rangeradio; the survey instru men ts followingwith the second flight; the reverse procedure was followed on return of the partyto the ship. In this way no one was everleft on shore without adequate protectionand means of survival.
Landing a helicopter on a snow surfacerequires very great judgement and skill onthe part of the pilot. Without some form ofartificial marker there is often no means ofgauging height above the ground. Initiallywe used a red flag attached to a stickwhich was intended to stick upright inthe snow, but more often than not thestick fell over or missed the target. A better form of marker is a weighted boxpainted red and of known dimensions, atleast two being carried in case the firstmisses the target and is carried down themountainside. A solid object of this kindwill also assist the pilot in judging theangle of slope, and give some indicationof the nature of the surface. For the firstfew drops the helicopter never alightedat all and we soon effected a techniquewith the 5.51 for clambering in and outat different hovering heights not exceedingfive feet above ground. For convenience ofloading and unloading equipment, however, it is desirable that the aircraft shouldland whenever possible, and with increased experience, plus the addition of
snow shoes fitted to each of the wheels, thisbecame the normal procedure.
By the middle of March conditionswere rapidly deteriorating, and we wereforced to abandon allY further survey untilthe following year. Despite having beenforced to operate at the tail end of theseason under condi tions of al most continuous cloud and high winds, we hadnevertheless made a promising start andgained a great deal of experience that "'asto prove of tremendous value in the future.
The most important outcome of thisexperience concerned the housing of thehelicopter on the ship. Since in the firstseason all available space on the ship wastaken up on the outward voyage withstores and aviation fuel, it was impossibleto make any provision for a hangar. Consequently, during the whole period of survey operations the helicopter had to remain on deck, its rotor blades supported onwooden stanchions standing on platformscantilevered outside the ship's side andwith its fuselage protected only by waterproof covers. In severe gales when the windsometimes reached up to 80 miles an hour,we expected to see the aircraft carried offthe side of the ship, and on one occasionfollowing a four-day blizzard it took 36hours to clear the fuselage of ice before itcould be used. Maintenance on deck wasat best a wretched job for the engineers andcould only be carried out in conditions ofcomparatively mild weather when wecould often have been flying.
In the second season we had no aviationfuel to transport and were therefore ableto use the forward hold of the ship as ahangar. The forward section of the deckwas replaced by a sliding steel (McGregor)hatch, giving access to the full opening ofthe hold. A lift operated by two handwinches was built to raise and lower thehelicopter from deck level. By this means,the helicopter, this time a Bell 47D, wastaken out from England fully erected andcould be serviced in periods of bad weather,during which the engineers were able towork in conditions of relative comfort inthe heated hangar below decks. The ship,free of the encumbrance of the helicopteron deck, was more easily maneuverableand the helicopter secure at all times whenon the ship. All survey, field rations andcamp equipment was also kept in thehangar, sorted into individual loads sothat at a moment's notice they could be
426 PHOTOGRAMMETRIC ENGINEERING
loaded onto the helicopter in readinessfor take-off.
Unfortunately, due to budget limitations beyond our control, only one helicopter could be taken and as luck wouldhave it, this crash-landed on the secondday of operations and became a totalwreck. While the ship returned to Montevideo to collect a replacement, a party offive surveyors was landed on King GeorgeIsland, the largest of the South ShetlandGroup. The terrain on this island is lessrugged than many, and the surveyorswere to travel on ski, manhandling sledgesand using small boats to cross AdmiraltyBay, which divides the island almost intwo. A base-line was measured on the icecap by similar means to those used for theDeception base, and despite meetingfrequent conditions of high wind, blizzardand cloud, the triangulation of half theisland was complete by the time the"Oluf Sven" returned. On the arrival ofthe ship, Bancroft and the team with theuse of the helicopter were able to doublethe area already covered by them in fiveweeks of extremely hard travel. Thisdirect comparison between the employment of air and ground travel is a measureof the vital role played by the helicopterin this type of terrain.
The ship returned with a new helicopterat the end of January and sailed immediately for the Palmer Coast where a freshbase-line was measured and the surveybegun in earnest. Our object was to extenda fully observed triangulation from CapeSterneck in latitude 64° southwards toincl ude as many as possible of the coastalislands and reaching as far south as wecould in the remaining six weeks of theseason (Figure 12). Already the briefAntarctic summer was over and the weather during this period gave us very littlerelief. Almost every day brought eitherlow cloud, high winds, or blizzard; oftenall three together. Nevertheless there werevery few days on which no work was done,owing to the extraordinary speed andmaneuverability of the helicopter, whichenabled us to make use of even the shortestbreaks in the weather when the windmight drop for an hour or two, and thecloud base lift sufficiently to give visibilitybetween the observed stations. Often whenconditions were impossible in one area wewere able to continue working by movingthe ship into the lee of an island or coastal
FIG. 12. Vertical photo Wieneke Island withbrash and icebergs. Williamson Eagle IX camera, Rose 6 inch lens, flying height 13,500 feet.
ridge. One had to be continuously on thealert during the hours of daylight watching wind, cloud and sea, which in the spaceof an hour could change with unpredictableand provoking abruptness.
The usual day's routine began at 6 A.M.,
when there would be a discussion betweenthe Captain of the ship, the senior pilotand the expedition leader. If it was decided that conditions were suitable tooperate, the ship would move to within2-3 miles of the point to be occupied.While the surveyors were breakfasting andpacking their personal kit, the radio officer tested the field sets, the steel hatchover the hangar was opened by the crew,and the engineers made ready the helicopter.
All hands were then mustered to ~an thewinches which operated the lift bringingthe helicopter on deck. By 8 A.M. thehelicopter was airborne on its first sortieand in less than ten minutes would beback to collect the second surveyor. Thecomplete operation of landing two surveyors with 600 lbs. of equipment seldomtook longer than 20 minutes and theirevacuation required less time even thanthis. Before the helicopter returned fromthe second flight the ship would often be onits way to the next point so that a minimum of time was lost. We soon found,however, that it did not pay to land morethan two parties at a time. On the fewoccasions when we broke this rule, we invariably finished up with one or more
AERIAL SURVEY OF GRAHAM LAND 427
parties left on shore, as a result of a suddenchange in the weather; several days mightpass before they could be relieved, thusdelayi ng the whole progress of the work.
In this manner the survey advancedsteadily and by using every break in theweather, however short, we completed inthe six-week period a belt of fully observed triangulation over a distance of 130miles. A total of 84 control points, of which24 were occupied stations, was establishedto an accuracy which is comparable tothat of normal secondary-control. Thetheodolites used were 'vVild T2 and CookeTroughton & Sims "Tavistock" readingdirect to 1 second of arc. The averagetriangular closure was in the region of 5-10seconds. The network was tied to a 16 st-arposition-line fix carried out with great careand the base-measurement checked to 1in 50,000. I do not think we can expectbetter in this typc of country, and I believeit can be claimed that it is the most accurate survey yet made in the Antarctic.Had we been equipped with the Tellurometer it is possible that even more rapidprogress might have been made, but unfortunately this instrument arrived justtoo late to be of use.
HYDROGRAPHIC WORK
As a by-product of ground controloperations, the navigating officer of the"Oluf Sven" carried out a valuable hydrographic survey. Using the air photo cover,radar ranges and bearings to points, incl uding those fixed by the triangulation,he replotted the coastline and from some900 miles of soundings recorded on theship's echo-sounder an accurate chart was
produced covering the areas of HughesBay and the southern Gerlache Strait,neither of which had been previouslysounded. When it is appreciated that manyof the most prominent coastal featureswere found to be 10 miles out of position,let alone the hundreds of coastal reefsand islands which showed up clearly onthe air photos but did not exist on anyprevious map, the reason why this expedition was necessary is not difficult tounderstand.
CONCLUSION
The measure of our succcss will perhapsbest be judged by the eventual publicationof the new series of mosaics and mapswhich are now undcr preparation by theDirectorate of Overseas Surveys, bascd onthe results of the expedition's work. Contact prin ts of the air photos are in themeantime already being widely used bythe field parties of the DependenciesSurveys, for planning sledging routes andfor geological, glaciological and otherstudies.
In these and other ways, it is to be hopedthat all the money and effort expended onthis first British photogrammetric surveyin the Antarctic will have been fullyjustified. \iVhatever else we achieved, I amcertain that our expedition has demonstrated beyond doubt that the long-rangephotographic aircraft with helicoptertransport for the land surveyor are an indispensable combination in overcomingthe tremendous obstacles of terrain andclimate, if this last remaining unmappedcontinent is to be surveyed within a foreseeable period of time.