Operation Morning Light was the code name used for the joint Canada-US operation to locate and clean up the radioactive debris associated with the re-entry and self-destruction of the Soviet nuclear-powered Cosmos 954 satellite in northern Canada in January 1978. That which follows is a personal account of that operation by the principal architect of the GSC’s Airborne Gamma-Ray Spectrometry system.
(GSC, retired 1995)
© Copyright 1995, Q. Bristow
The Geological Survey of Canada’s airborne gamma-ray spectrometry program was begun by Arthur Darnley in 1966 when he arrived from the U.K. to join the GSC. One of his first actions was to contact the Commercial Products Division of Atomic Energy of Canada Ltd., (AECL) with a proposal for a collaborative project to develop a very sensitive airborne gamma-ray spectrometer, that would be able to detect and provide quantitative measurements of terrestrial radiation from the three naturally occurring radioelements (potassium, uranium & thorium). Darnley reasoned that if the necessary sensitivity could be realised, the data would provide considerable additional information to aid geological mappers in the identification of rock types. This project, which began in 1967, was ultimately very successful and spawned a whole new technique called Airborne Gamma-Ray Spectrometry (AGRS), that continues to provide a substantial economic benefit to the Canadian geophysical survey and instrumentation industry to this day. I was the AECL project leader for this development and subsequently joined the GSC where I became head of a Section with a mandate to develop all sorts of instrumentation for geochemical & geophysical applications, including airborne instruments. The Section included Jacques Parker and Yves Blanchard. One of the projects was the development of a second-generation airborne gamma-ray spectrometer in 1976, to replace the original AECL version which was becoming obsolescent.
The new version used a NOVA minicomputer, manufactured by the Data General Corporation. This company was a spin-off from the Digital Equipment Corporation (DEC), formed by the process which was the nightmare scenario of hi-tech companies of the sixties and seventies - the splinter group syndrome.
The NOVA machine was similar to other minicomputers of the 60’s and 70’s that proliferated from many manufacturers. Each one had its own assembly language which had to be learned if one was ever to do anything useful with it. After much blood, sweat, toil and tears, our team produced a really sophisticated second-generation airborne system to replace the original AECL model. At that time the standard method of communicating with minicomputers was via a data terminal, a device with a keyboard, and a thermal paper printer like a Fax machine. This was not considered adequate for an instrument where it was crucial to be able to see the different peaks in the gamma-ray spectra as they were being acquired and recorded on magnetic tape at one second intervals. We ended up designing and building an interactive graphics display which was unique for any sort of airborne scientific instrument at that time. It was one of the most daunting tasks I had ever undertaken, involving as it did some incredibly complicated ’real-time’ software to make the minicomputer keep track of servicing all the gadgets connected to it without tripping over itself.
To put things in historical perspective, it is worth noting that at that time, the biggest recent impact of technology on the average consumer had been the basic four-function pocket calculator, with its hard-to-read and power-consuming red LED display. The era of personal computers with CD-ROMs and megabytes of memory was still many years in the future. The NOVA minicomputer for example had just 64 kilobytes (0.064 Mb!) of memory to accommodate all the necessary programming, not just for the data acquisition and recording, but also for all of the operator interaction and display, all without benefit of any underlying operating system at all.
Because the business of airborne gamma-ray spectrometry was still evolving, all sorts of options had been included in the new system that would probably never be needed for routine use, but just might be useful in experimental work where the exception is always the norm. One of these enabled the operator to program a variety of moderately complicated functions which could be plotted on the six-channel strip chart recorder in real time. The standard channels were simply the intensities of the potassium, uranium & thorium signals as they varied along a flight line. These additional functions allowed the sums, products or ratios of any of these signals, or indeed of the signals from any other regions in the spectra (recorded at one second intervals) to be plotted on the chart recorder.
It was late in the autumn of 1977 before this very complex machine had been built and was ready for a test flight. The first concern was that it would stand up to the vibration and electrically noisy environment of an aircraft, but everything looked good. I was particularly pleased with the interactive graphics display and felt that it had indeed been worth all the extra effort, (there were times when I was afraid that I had side-tracked the team into putting a lot of development time into something that was of marginal value in the overall project). That first test was more of a shakedown flight than any thing else. Some more rigorous tests would certainly be needed before it could be declared operational. However, the weather closed in and there were no further flights. That was of no particular concern because there would be plenty of time in the spring of 1978 for some more methodical and exhaustive testing.
In January 1978, a news item surfaced about a Soviet nuclear-powered satellite, Cosmos 954, that had become unstable and was gradually descending from its normal orbit. At first there was only minimal interest, but as the days went by, news of the impending re-entry of this satellite and the fate of the nuclear reactor which it carried, moved onto the front pages. The Soviets had conceded that it was indeed powered by a nuclear reactor, something had gone wrong and they no longer had any control over it. By this time it was apparent that the scenario had all the makings of a major nuclear catastrophe; a satellite carrying a lethal nuclear payload circling the earth every ninety minutes or so and coming ever closer to the point where it would suddenly enter the atmosphere and plummet down to the surface - somewhere. This was literally Russian Roulette with a vengeance and if that ’somewhere’ turned out to be a major population centre, then all hell would break loose.
In the early morning hours of January the 24th, as it crossed the northern Pacific from west to east, it finally disappeared from the radar screens that had been monitoring its progress. Shortly after that, there were reports from various locations of a ball of fire streaking across the morning sky over the Northwest Territories. When these reports along with other evidence were pieced together, it became apparent that the satellite had almost certainly re-entered the earth’s atmosphere somewhere over the Queen Charlotte Islands (several hundred miles north of Vancouver) along a trajectory that would have taken it north-east across Great Slave Lake towards Baker Lake. If this turned out to be correct, then there would be debris from the impact, wherever it finally landed, strewn across this area. Furthermore a lot of it would probably be dangerously radioactive if the nuclear reactor had survived the re-entry more or less intact and had then broken apart on impact.
This area was by then in the deep-freeze of the Canadian sub-Arctic winter, when temperatures range down to -40C (-40F) degrees or lower. At that point the Canadian government quickly agreed to accept an offer of American technical help in locating any radioactive debris that there might be. It was clear that any such operation would require massive logistical support on a scale which only the military could mount at short notice. Working on the premise that the impact trajectory was as described above, it was decided to make the Canadian Forces Base at Namao (a suburb of Edmonton, Alberta) the headquarters for the people who would be involved in the search for the satellite. The operation itself was code named ’Operation Morning Light’, because of the sightings in the early morning which had provided the only specific evidence that could be linked to the re-entry.
Bob Grasty, who was one of the scientists in the Radiation Methods Section involved with the GSC airborne gamma-ray spectrometry program, went to the Canadian Forces Base at Namao (CFB Namao) to see what was going on, the day after the satellite ’Cosmos 954’ went missing from the skies. What he found was that the place had been virtually taken over by the Americans. They had a standing task force, ’Nuclear Emergency Search Team’ (NEST) stationed at Las Vegas, Nevada, whose mission was to fly helicopter surveys to monitor the area for radiation leaking from the surface following each and every underground atomic weapons test that was conducted at the Nevada Test Site. They were well equipped, and on virtually permanent standby to be deployed as a self contained and self-sufficient unit to almost anywhere in the world at a few hours notice, to deal with just such an emergency as this appeared to be. They had arrived complete with all their gear in five C-141 military cargo aircraft, which had brought all the paraphernalia and the 120 people that came with it. The C-141 cargo plane was a behemoth about the size of a 747 airliner. The equipment that it disgorged included not only a complete mobile communications centre, with powerful relay transmitters to communicate with their home base, but two helicopters with gamma-ray spectrometry systems already installed, as well as several other spectrometry systems ready to be installed in any available aircraft. Also included were two complete data processing units in what came to be known as the ’bread vans’, because they were the type of small delivery vans typically used by bakeries.
Grasty quickly tuned in to the situation and discovered that no one was aware that the Geological Survey of Canada had been conducting airborne gamma-ray spectrometry surveys for ten years and had a great deal of experience in that field. He told the people in charge that the GSC now had a new and sophisticated system that was tailor made for the sort of airborne search that was about to begin. As a result, an urgent request was made for assistance from the GSC. I got the message at about ten o’clock on the morning of the 25th of January and spent the rest of the day preparing to leave for CFB Namao that night. The preparations included stripping the place of just about every piece of potentially useful equipment that wasn’t nailed down and arranging for it to be boxed up for shipment. A requirement had arisen earlier for a new borehole geophysical logging system to measure gamma radiation (and other variables). It had been designed to be based on the NOVA minicomputer and hence was compatible with the new airborne system as far as possible. That decision was now paying off under circumstances that could not possibly have been foreseen. The complete borehole logging system was of course at the top of the list to be included in the shipment as a spare to back up the airborne system.
By seven o’clock that evening I was ready to go, along with another member of the Radiation Methods Section, Peter Holman. Peter was the field operator who for the previous ten years had been conducting the airborne surveys with the GSC Skyvan twin-engined turboprop aircraft every summer as party chief. He was responsible for acquiring the data and processing it into maps showing the concentrations of the three radioelements. These maps, published by the GSC, were eagerly snapped up by mining and exploration companies who by then had come to recognise their value in mineral exploration. Peter had a great deal of hands-on experience in navigating closely spaced flight lines with nothing more than a topographic map or a mosaic of aerial photos on his knee, and in dealing with the vagaries of airborne geophysical surveys in general.
A military Cosmopolitan aircraft had been assigned to take us from Ottawa to Edmonton and it was being prepared for departure as we made our way to the operations sector of Ottawa airport where the GSC had a hangar for the Skyvan aircraft. The spectrometer was large and heavy. The electronic units (NOVA minicomputer, magnetic tape drive, graphic display, keyboard/printer terminal, six-channel strip chart recorder, and other modules) were contained in an instrument rack about four feet wide, which was bolted on to a sheet metal box containing the twelve large scintillation detectors, cushioned in polyethylene foam to provide thermal and mechanical insulation. This detector box was about six feet long by two feet high, giving an overall length with the instrument rack attached, of about nine feet. This integral structure was mounted on small wheels, so the whole unit could be moved easily into and out of the aircraft from a special dolly, the same height as the aircraft floor, with rubber tired wheels.
As fate would have it, there was a heavy snow storm blowing outside and the spectrometer had to be taken somehow from the GSC hangar to the waiting aircraft, which was in a military hangar about half a mile away at the other side of the airport. The dolly was designed to be towed around the hangar and the apron outside by a small van, but it had no springs and no protective cover against rain or snow. Peter found some plastic sheet and managed to tape it over the instrument rack portion and then some air was let out of the tires of the dolly to provide some minimal springing. After coupling it up to the van we opened the hangar door and were met with a wall of snow blowing in. We looked at each other and I said "Well, we don’t have a hell of a lot of choice do we?" Peter agreed and we got in the van and gingerly started to pull the hundred-thousand dollar spectrometer out into a blinding snow storm and sub-zero temperatures, protected only by the plastic sheet which fitted where it touched. I knew that the various modules, including the computer would be none the worse for getting a bit chilly, provided they were allowed to warm up before being turned on, but like most things electronic, they would take a very dim view of getting wet. The scintillation crystals were the most fragile components, easily fractured by either mechanical or thermal shock, which is why such care had been taken in the design to have them snugly cushioned with lots and lots of insulation. This turned out to be the first of many occasions when that paid off.
We crawled across the network of runways through the snow storm, with Peter driving the truck gingerly with its expensive load trundling behind, for what was one of the longest half-hours of our lives, before arriving at the military hangar. There we saw the Cosmopolitan: it was a twin-engine turbo-prop, about half the size of the original Viscount turbo-prop airliner in wide use until the early 70’s. It looked huge in the hangar and the cargo loading bay was somewhere up in the rafters. The military chaps were a bit nonplussed when they saw the size of our contribution to the satellite search effort, but nothing daunted them and they used a fork lift to hoist the spectrometer, dolly and all, on to a loading gantry which in turn raised it up to be level with the cargo bay. The entrance was in the side of the fuselage, unlike the GSC Skyvan, which had a rear door entry giving unimpeded access to the interior of the aircraft. This meant that the spectrometer, all 600 kilograms of it, had to be pushed in and turned at the same time, because it was too long to go in broadside. They had a difficult time shoe-horning it in and Peter and I had an even more difficult time watching them do it. Several times one of the straps being used to hoist it, slipped off, and it dropped heavily a few inches onto one corner or another. Absolutely none of the equipment in it liked to be dropped and by the time that it was safely in the aircraft I had really begun to despair that any part of it would ever function again.
The trip to Edmonton was uneventful and we landed at CFB Namao at about midnight local time. There we met up with Grasty and had a chance to see at first hand the circus that the base had become. By this time the American team had made several sorties with their systems but had found nothing real. By this time also, international interest was at fever pitch with hordes of media people running round all over the place. The temperature in the search area, which went across the frozen surface of Great Slave Lake, was about -40C (-40F) and that presented a real problem for the American helicopter-borne gamma-ray spectrometers. Their scintillation detector packages were carried on pods attached to the helicopter skids and had minimal insulation. This meant that their crystals were unable to withstand the thermal shock of the brutally cold temperatures. After all, their systems were all designed to operate in the warm sunny climate of Nevada. This problem was soon taken care of by the military who made available several huge Hercules C-130 cargo aircraft to carry the American spectrometry systems, and later on some military Chinook helicopters. The Hercules is a four engined turbo-prop machine, perhaps twice the size (or more) of the Cosmopolitan. It can fly for more than twelve hours without refuelling and can carry an awesome amount of cargo. At one point later on in this operation a Hercules was used to ferry a full-sized bulldozer to a destination where it was needed to construct a runway for the Hercules to land with cargo to establish a base camp.
Holman and I met the base commander Colonel Dave Garland, and some of his people. He was just the man that was needed to handle the potentially chaotic situation that was developing at such an alarming rate. He was a cool customer and practically unflappable. He needed to be; in the space of twenty four hours his base had been overwhelmed by the invasion of not only the very large, very high powered and very high-tech American NEST people, but also by an international media entourage which was growing by the hour. Reporters from Japan and Europe were pouring in, eager to file the ’firstest-with-the-mostest’ stories of a nuclear disaster descended from the skies like some sort of divine retribution. Garland had to set up press conferences to take care of all of that, and more importantly, keep abreast of (and control over) the mammoth logistics problems which were looming up, as briefings outlining the magnitude of the problem and what it would take to deal with it, filtered through from the Americans.
There was much mis-information and general confusion reigning during the first few hours. On one of the sorties some sort of blip had been recorded by an American spectrometer and they were practically sure that it was something real. Word of the probable ’hit’ was passed on to Ottawa, where the Minister for National Defence Barney Danson, enthusiastically proclaimed to the world that proof-positive had been found that the nuclear-powered satellite had indeed landed on Canadian soil. "It’s either a piece of radioactive debris, or the greatest uranium mine in the world" , said Danson. Unfortunately the hit turned out to be some sort of electrical burst of interference, probably from the aircraft radio. When this was verified, the Chief of the Defence Staff, Admiral Falls, gave a second news conference in which he confirmed that the report had been false, adding: "I said before that I didn’t believe that there was the remotest possibility of anything landing on earth. I still feel it unlikely that anything has landed". It was against this backdrop that the GSC contingent entered the fray.
The first order of business was to get the spectrometer out of the Cosmopolitan aircraft and into a Hercules. This was accomplished with less difficulty than had been encountered while getting it in, but it was still a bit of a cliff-hanger. The Hercules was an oversize version of our ’little’ Skyvan survey aircraft, with a rear door entry and a flat floor, designed of course for all sorts of cargo. Compared to the Skyvan, the inside seemed like a football field, with the spectrometer (which practically filled the available space in the Skyvan), sitting like a little box in the middle of the vast interior. In the aviation business, the authorities are very fussy about approvals for the installation of any kind of airborne instrument packages, for obvious reasons, related to surviving a crash. The mechanical design has to be such that the package is able to survive an impact force equivalent to ten times its weight in a forward direction and four times its weight in a downward direction, without disintegrating and demolishing everything in the interior cabin space. This of course includes particularly the method of anchoring the package to the fuselage. Peter and his colleagues had had to go through much paperwork and bureaucracy to get the necessary certification for our Skyvan installation and I wondered how that would be handled vis a vis the Hercules. Absolutely no problem at all, as it turned out - just set it down where it was most convenient and tie it down to some of the many anchor points in the floor with about half a dozen huge nylon-web cargo straps. Those straps had mean and powerful ratchet buckles capable of crushing the sheet metal box and I was a little nervous as a couple of corporals began to reef them down. I muttered something about certification for such an ’installation’ and the answer came back "...no problem...it’s cargo isn’t it?" and that was the end of that.
After all that it was time to fire up the spectrometer and find out whether all of its many parts had survived the passage. I loaded the magnetic tape containing the program with some trepidation and was infinitely relieved when the proper display popped up on the little graphics display screen. At least the computer was working. Then it was time to check out the scintillation detectors and all twelve of those were also in good health. By then it was about two o’clock in the morning and Holman and I were told that the plane was scheduled to take off for a twelve hour flight in the search area at dawn (six o’clock) and could we tell the air crew who would be going to operate the spectrometer. At that point I was really the only one who knew how to operate it, because Holman’s only contact with the new spectrometer had been on the one and only test flight which had been flown many months previously. I said I would go and went off to get at least a couple of hours of badly needed sleep, after being on the go for almost twenty four hours (there is a two hour time difference from Ottawa to Edmonton).
The giant machine took off at about 7:30 AM and headed north over the winter landscape of Alberta to Great Slave Lake. The ferry flight took about two hours at an altitude which was far too high for the spectrometer to detect anything from the ground. When it reached the search area it descended to about 1500 feet and began a long survey, criss-crossing the frozen surface of a portion of the western section of the lake. I had my eyes glued to the strip chart recorder, one channel of which had a special ratio function programmed into it to recognise man-made as opposed to natural radiation. This I had done on the advice of the chief data processor of the American team, a delightful chap by the name of Thane Hendricks. The American spectrometers had minimal display capabilities as their equipment was designed to operate in small helicopters where every extra pound mattered. The capability which the GSC system had to display things like running ratios on the strip chart recorder subsequently turned out to be a vitally important feature for real-time recognition of radioactive debris.
The aircraft flew for about eight hours without any tell-tale blips on the chart recorder, and then turned south and climbed up to an altitude of ten thousand feet or so for the return flight to CFB Namao. It arrived back somewhere about 8:00 PM. There was even more of a circus atmosphere prevailing by then, than when we had left twelve hours previously. In the interim, the word had gotten around that a super-sensitive Canadian spectrometer had joined the search, the only made-in-Canada high-tech scientific instrument to be part of the operation at that point. Grasty, Holman and I found ourselves dogged by reporters anxious to know more about the ’Canadian system’ as the spectrometer had now been dubbed. In the middle of all this, I was desperately trying to make a coherent set of notes for Grasty and Holman, to give them some guidance on how to operate the system and what to look for, as they were scheduled to go on the next sortie.
By this time it had been decided to include some members of the Canadian press on a Hercules flight with the ’Canadian system’. This was done to try and dispel the growing impression in the media that Canadians were just helpless by-standers in a dramatic scientific detective game, being played out in their country by an imported team of American Scientists and Engineers. This flight, with the press on board and Grasty and Holman as operators of the spectrometer, took off at about 10:00 PM, just as soon as the Hercules had been refuelled, following the return of my flight with it. I went to catch up on some desperately needed sleep, because I was going to be on deck again for yet another twelve hour stint immediately following the return of the press flight. This was in order to keep to the plan to have that Hercules searching around the clock. I slept the sleep of the practically dead for about twelve hours and had to hurry to make it onto the plane again without causing any delays. As a result, I did not make contact with Grasty and Holman before taking off. After searching another segment of the survey area the plane returned to base sometime around 2:30 AM.
I was greeted by a frantic crowd of reporters and members of the American team. It transpired that after inspection of the chart record following the press flight, a sizeable blip had been found on the ratio channel which corresponded with a location somewhere near Snowdrift, a small Indian hamlet on an island at the eastern end of Great Slave Lake in McLeod Bay. This literally set the world abuzz, because up to that point nothing had been found, other than the original false hit, after almost three days of continuous searching with three Hercules aircraft. Speculation by the media entourage had reached a fever pitch, which increased as successive press conferences failed to report any positive evidence that the satellite and its potentially deadly nuclear cargo had landed anywhere. At that point however there was a glimmer of something, but it could not be confirmed until the gamma-ray spectrum of whatever it was that caused the blip had been examined. That could not be done because the only one who knew the format for the recording of the spectra and other information on the magnetic tape was me, and I was incommunicado for the next twelve hours on another flight, or at least in no position to get into nitty gritty details over the radio without my files of notes.
It turned out by a quite extraordinary and serendipitous coincidence that the Americans were also using NOVA minicomputers for their systems, although as was indicated earlier their systems did not have the sort of interactive capability with a graphics display that the GSC one did. They relied on post-processing of their data, which is why they had brought their ’bread vans’ as mobile data processing centres. It meant, among other things, that the GSC magnetic tape data could be read by their equipment and that they could produce hard copies of individual spectra for examination, but only if they knew how the data had been recorded. I escaped from the press with Thane Hendricks and we closeted ourselves in one of the bread vans with a couple of the other American physicists. I was able to provide Hendricks with the vitally important details and he was then able to use one of the utility programs in his voluminous library of software to read the tape and search for the spectra which had caused the blip. For one moment, I was gripped by an awful misgiving, I could not remember actually having processed a tape from the new system to verify that the data were in fact being recorded properly. Fortunately the data coming from the processing program quickly confirmed that the spectra were indeed all there, it was just a question of combing through hours of data on a one-second by one-second basis to find what we were looking for. The utility program was designed to do that in short order once certain criteria had been entered on the keyboard and it was not very long before it found the spectral data that had caused the blip and had ’stacked’ them (added them together) to make one very clear well-defined spectrum and print it out.
We looked at it with bated breath and then looked at each other, we were all familiar with the major peaks in the spectra from nuclear reactor isotopes and what was staring us in the face right there was the peak of one of them, Lanthanum-140. There could only be one explanation for the presence of that nuclear fission product on the frozen surface of Great Slave Lake, one of the most remote and desolate regions of North America. That single gamma-ray spectrum provided the incontrovertible proof-positive that Cosmos-954 had landed and that there was indeed cause to be concerned about radioactive debris along the impact trajectory. Once that news broke, the place went from being a circus to something more like the trading floor of a stock exchange as frantic media people fought over any and every available phone to file their stories. The ’Canadian spectrometer’ was suddenly the star of the show and the GSC contingent was amazed to see the spectrometer graphics display staring out from the T.V. screen during the eleven PM CBC-TV newscast that evening. Someone must have shot some footage of it during the press flight. The success was played up for all it was worth by the Canadian media, anxious probably to make the point that some ’made-in-Canada’ science and engineering had been responsible for finding the first piece of radioactive debris (thereby focusing the search), and not the high-powered American team with their almost limitless resources.
At that point, reinforcements from the GSC in Ottawa were urgently needed to help Peter Holman take over the search missions of the Hercules with the GSC spectrometer, which would now become routine and literally round the clock until further notice. Arthur Darnley, (now in a senior management position) arranged this and within a couple of days we had a staff of people from the GSC to keep things going. These included Jacques Parker, Yves Blanchard, Bill Hyatt and George Cameron (then both in the Radiation Methods Section) and Keith Richardson, the Head of that Section. Keith was a geologist who had specialised in the geology of the naturally occurring radio-elements and was at that time in charge of the GSC airborne gamma-ray spectrometry project. He had spent the 1960’s with NASA in Houston, training astronauts in basic geology in preparation for the manned Lunar mission. His experience there was very apropos, because the American organisation was very much along the lines of a NASA project. As a result he was able to get a handle on setting the GSC contingent up as an integrated group within the overall operation. By that time the Atomic Energy Control Board (AECB), the nuclear regulatory agency in Ottawa, had become involved and had sent a team of people to organise the actual recovery of the radioactive bits and pieces as they were located. One of the key problems was navigation. The search area was in a region of the Northwest Territories where even the few landmarks that could be seen in summer weather, were now under a blanket of snow stretching to the horizon in all directions. This was long before the days of the satellite Global Positioning System (GPS), and the only way to conduct a search like that was to use a radio navigation system. The Americans used a Microwave Ranging System (MRS). It required two transmitting beacons to be set up in advance which would allow coverage of an area about 80 km square. They had brought all the gear with them and it was subsequently adopted for use in the search. The transmitter beacons were not however designed to operate in Arctic conditions, which meant that they would have to be ’winterised’. Col. Garland and his people organised that, with special insulated boxes being built at the workshops on the base to house the beacons and their batteries.
The system worked well and enabled the flight path of the aircraft to be plotted after-the-fact using the beacon positioning data that were recorded along with the spectrometry data on the magnetic tape. Military helicopters were used to set the beacons down and to move them from one sector of the search area to another as the operation slowly covered the entire impact trajectory of Cosmos-954. There were some problems however. On one occasion, the aircraft arrived at the search area, only to discover that one of the MRS beacons was dead. That was a real disaster because there was absolutely no way to find the wretched thing. If both beacons were transmitting normally, there was no problem because there was a procedure for zeroing in on either transmitter, much like homing in on a rescue beacon (which is how the helicopters found them to service or move them). The better part of six hours was spent flying circles at an altitude of about one thousand feet around the beacon that was working, to find the beacon that wasn’t, with the navigator using binoculars.
As the search settled down to a routine operation, the media interest quickly waned and things became rather less dramatic and more business-like. From the end of January until early April, GSC team members Holman, Blanchard, Parker, Hyatt and Cameron put in hundreds of hours in gruelling twelve-hour sorties over the remote and desolate impact-trajectory of the satellite, all of them flown at very low altitude. In addition they had to keep up with processing the truly enormous amounts of data that were produced. Another key member of the team was Dominique Boucher, who processed data back at the Camp while the aircraft were out flying. This was surely one of the best examples that is on the books anywhere, of versatile GSC people turning their hands to practically anything and everything in dealing with a protracted national emergency.
As was indicated earlier, there had been several communications gaffes committed by the powers that be in Ottawa (who were supposed to be in charge of everything), because of the difficulty of keeping up with the pace of the fast moving events at Namao. As soon as there was a breathing space, a delegation of general staff officers and senior civilian officials descended on Namao for a strategy meeting to make sure that there would be no more embarrassing incidents. Darnley, Richardson and I attended as part of the on-scene Canadian scientific contingent. It was a fairly formal affair with translation into French being recorded by a translator sitting at one end of the conference table. He spoke into a microphone contained in a rubber mask covering his nose and mouth, which prevented the sound of his voice from causing any distraction to the rest of the participants.
One member of the delegation was a rather colourful Brigadier General who had been appointed as the senior public relations and press officer in Ottawa for everything to do with Operation Morning Light. He had the rather thankless task, among other things, of briefing the press every morning on the latest developments. He had been caught off-guard several times (to his considerable discomfort) by pointed and probing media questions, and he was now adamant that every scrap of information must be transmitted to Ottawa HQ as soon as it was available on a daily basis. That brought dissent among some of the on-scene participants, who argued that the available resources were just too limited to be tied up with writing comprehensive daily situation reports (SITREPS) to Ottawa, simply to keep the press at bay. At that point the visibly exasperated General lost his cool completely and shouted, his voice rising with indignation; "It’s all very well for you people - you’re not the ones who have to get up every morning and face those bloody news hounds in that (expletive deleted) bear pit". The effect on the translator was comical to behold, we saw his eyes widen above the rubber mask, while the rest of his face froze as he struggled to find an appropriate way to handle the situation. Before he could utter a word, the General swung round with his finger pointing at him and said "don’t you dare translate that!" The meeting dissolved into helpless mirth for about a minute, after which, as is so often the case when something breaks the tension, a sensible compromise was reached.
Initially the Americans were processing both their own data tapes and those recorded by the GSC spectrometer, because there was no GSC software to do that at the time. It was not long before a major bottleneck developed and the analysis of the taped data for the ’hits’ from a flight (locations of bits of radioactive debris detected by the spectrometry systems) were lagging badly. It became clear that the GSC team needed to have their own independent means of finding the hits on their tapes, but it was also clear that there was no time to embark on a major software development project to accomplish it. The tools that came with the NOVA for software development were primitive in the extreme, as they were for most minicomputers in the 1970’s. To begin with there was no monitor screen, only a data terminal with thermal print paper, which allowed you to type in your program line by line. Editing was incredibly tedious because it was not possible to see a cursor blinking against the line or word that you wanted to change. You had to identify the line number, make the change blind and then print it out to verify it. After that the program had to be assembled and linked, a two-pass operation involving the loading of a special assembler program from a magnetic tape, changing that tape for the one containing the text of the program which had been typed in, and finally (after yet another tape rewind and change) all the individual modules of the program were linked together into the final version. In the course of the assembly operation it would flag the inevitable errors that one made, either in the syntax (missing commas or spaces), which meant starting all over again. It took about half a day (if all went well) to produce an updated version of a program, complete with the all-important printout, showing the numbers of the locations in memory where the various instructions and key pieces of data lived.
In the course of developing the large and complicated software package, which enabled the NOVA to run all its peripheral gadgets and keep track of everything, I had been through this frustrating and tedious performance a good many times and had developed a way around it. The so called ’machine language’ cut and paste approach. If for example one needed to change a bunch of instructions in the program sequence, then it was possible to go to the memory locations where they were and key in the binary code (ones and zeros) corresponding to the desired instructions (the codes for all the instructions were listed in the manual), through the sixteen switches on the front panel of the NOVA. This of course was a daunting prospect but could be made a lot easier if the numbers could be typed in. I had written a utility program to do just that which I had used extensively for patching changes to the software during the development phase. As a result I knew most of the instruction codes by heart and it was not too difficult to make a lot of progress quite quickly without ever having to resort to the dreaded edit-assemble-link business. The price was to have reams of notes detailing the changes that had been made to the original version, so that a new version could be produced at some point incorporating the changes and giving a printout that actually agreed with what was in the machine.
I realised that it would probably be possible to turn the operating program more or less inside out, so that it would read the tapes that it had produced during the flights, generating exactly the same strip-chart record and graphics display of the spectra, that it normally did while it was receiving the data from the sensors in real-time. All that had to be done was to put the gamma-ray spectra from the data tapes into the same place in memory that they normally went when they were coming directly from the detectors, and do the same thing for the navigational data. The data processing part of the program would not know the difference and would not need any major modification.
It took me about three days of uninterrupted work to make the modifications required and at the end of it the GSC team had a utility program tailored to detecting ’hits’ which could sort through a tape at fairly high speed. It could detect any man-made radiation using Thane Hendrick’s ratio technique; stop the tape; back it up a few records before the event and accumulate the normal or ’background’ spectra; then do the same for a few seconds worth of data after the event. It then subtracted the background from the ’hit’ spectra and printed them out automatically.
The spare system which had been brought at the beginning, was then pressed into service as the data processing system back at CFB Namao. The resulting spectra, processed in that way, revealed very clearly which man-made radioactive isotopes were present in the debris lying below the aircraft when it was detected, and helped to identify certain parts of the satellite. For example there were a number of beryllium rods (used as neutron shields and reflectors in the nuclear reactor), which invariably had Cobalt-60 contamination. The characteristic spectrum of Cobalt-60 was easily spotted in the hit records, and the information that the hits were probably due to beryllium rods, would then be passed on to the helicopters which were going out to recover the pieces at the specified location.
Identification and recovery of the debris involved military helicopters equipped with radiation detection systems which were less sensitive than the ones used in the Hercules. They would relocate hits which had been identified by the Hercules flying the search grids and mark the positions, so that the recovery helicopters (with suitable lead-lined boxes and people carrying hand-held radiation monitors) could go right to the spot. The problem of marking the sites in such a way that they would be visible from the air was solved with typical military pragmatism. An order was placed for several thousand condoms and a production line was set up with people filling them with coloured dye. When these were dropped from the helicopters over the site, they burst open on contact with the frozen snow making a highly visible splash of colour which could be seen by the recovery teams.
The Russians (or Soviets as they were known then) were initially extremely coy when contacted via diplomatic channels for technical information about the design of the nuclear reactor aboard Cosmos-954. They refused to give any details other than that it had been designed to self-destruct before re-entry in the event of failure of the satellite control system. They did not concede that it had come down over Canada but phrased everything in a sort of hypothetical manner, obviously to avoid any possibility of liability later on. The general paraphrase was along the lines of:
Eventually they were more forthcoming and provided more technical data which was useful in trying to work out how much of it there had been, how much we had recovered and the likely fate of the rest of it.
At one point a large piece of the satellite known as the ’antlers’ was discovered quite inadvertently by two young explorers who were part of a team of six adventurers making a voyage in the Thelon River Game Sanctuary, some 400 miles to the east of the eastern tip of Great Slave Lake. They were travelling by dog-sled with two teams of dogs, following the route that had been taken by the English Naturalist John Hornby in the 1920’s. The Thelon River area is one of the coldest in the western hemisphere and surviving a winter there is a major challenge. Two of the party came upon the metal object, surrounded by evidence of melted ice, at a place called Warden’s Grove and having heard of the Cosmos-954 episode on their radios, suspected that it might be a part of the debris. How right they were. They contacted Edmonton and passed on the information, following which the military flew in to investigate and brought the party out, leaving personnel to feed the dogs. There was some concern that the two who had found the ’antlers’ might have suffered serious contamination, but after extensive tests, it was concluded that nothing serious had befallen them. Initially it was suspected that the antlers were just the visible portion of a much larger chunk of the satellite, possibly the reactor, submerged beneath the ice. Subsequent investigation however showed that nothing else was there and that the ’antlers’ themselves were not a health hazard from a radiation point of view.
This incident highlighted the problems that the media created with their relentless pressure for ’human interest’ stories. The radio messages from the two explorers were relayed via an operator in Yellowknife and were picked up by the media. Some reporters decided to try and get into Warden’s Grove by chartering planes from Yellowknife to get first hand pictures for their stories, despite the fact that a directive had been issued prohibiting any civilian flights into or over the area. At that point no one knew whether or not the main segment of the reactor core was there, or what the radiation levels might be. It was therefore imperative to head off any attempt by the press to go rushing in where experts feared to tread, thereby putting themselves at risk and further complicating an already difficult situation. Fortunately the charter operator notified Namao about the proposed press expedition into Warden’s Grove and Garland ordered paratroops to be dropped in to secure the whole area. Strictly speaking he needed the blessing of the Solicitor General in Ottawa to do this because the guarding of a non-military site came under the jurisdiction of the RCMP. However the man who was the Solicitor General in Trudeau’s government at the time, Frances Fox, had just resigned as a result of an unrelated scandal. Some sort of official approval was nevertheless obtained in time for the operation to proceed without any delays.
When it became clear that most of the debris was likely to be found immediately to the northeast of the eastern end of Great Slave Lake, it was decided to set up a field camp there. The idea being that parties could operate from the camp to recover debris, rather than having to spend more than three hours flying time to get there and back from Yellowknife, which was the nearest centre having the necessary facilities from which to operate the military Twin Huey helicopters. The camp was named by the ’inmates’ as ’Camp Garland’, and the lake on whose shores it was established was called ’COSMOS Lake’. Setting it up was a major challenge, but one which the Canadian Military was equipped to undertake. The first order of business was to get a bulldozer in to clear the snow off the lake to make a landing strip for smaller aircraft to land with supplies and equipment. At first sight one might be forgiven for thinking that this was a chicken-and-egg problem. After all the only aircraft big enough to bring in a bulldozer in the first instance was a Hercules, which would need a sizeable landing strip cleared of snow to deliver it.
No problem, the bulldozer would be set down by the Hercules without it having to land, in a death-defying manoeuvre developed by the Canadian Military known as ’Low Altitude Parachute Extraction System’ (LAPES). The bulldozer was trussed up on a massive wooden pallet with several parachutes attached to it. The Hercules approached the lake with its rear cargo bay open and swooped down to within a few feet of the surface. As it almost touched the ice, the parachutes were released into the slipstream and the lynch-pins anchoring the wooden pallet were withdrawn. As the parachutes deployed, the enormous drag that they exerted yanked the pallet out of the aircraft and onto the snow covered surface of the lake. The pilot then opened up all four engines to full throttle and roared off the end of the lake, climbing as fast as he could to get back up to a safe altitude. It was a very very tricky thing to do. The payload of any aircraft has a drastic effect on the position of its centre of gravity, which in turn affects its aerodynamics and the way it handles. That is no problem if it doesn’t change, but in this manoeuvre the sudden loss of a several tons of cargo (in this case the bulldozer) occurs at the worst possible moment - when the aircraft is practically touching the ground. It requires a massive and finely tuned correction of the controls by the pilot (with split-second timing) to compensate for the sudden upward pitch of the nose of the aircraft which results. Over compensation could cause the aircraft to go nose-down and plough into the ground. There have been some tragic accidents involving the LAPES manoeuvre since that time.
Yves Blanchard was on the flight that day and saw it work flawlessly. The bulldozer came out right-side-up and undamaged. An operator was parachuted in to get it going and it did what was required of it, which was to clear the snow and make a landing strip. After that there were many more LAPES operations to supply other equipment that was needed before a landing strip capable of accommodating the huge Hercules aircraft was fully operational. In the interim there was a regularly scheduled series of helicopter supply flights into Cosmos Lake and the camp was quickly established. It was maintained there for many weeks, supported by a shuttle service of helicopters and Hercules aircraft. In the event the site that was chosen turned out not to be the centre of the main debris area, but no one could fault the decision to set up a camp there, given the paucity of information available at the time, and the pressure to get something going which was coming from all levels.
Towards the end of Operation Morning Light the GSC Hercules went on a mission to check out some uncertain hit that had been registered. Because of the shortage of aircraft, the flight had to be combined with a stop-off at Yellowknife to pick up supplies for Camp Garland. It landed at Yellowknife and a whole lot of gear was loaded on, including some gasoline heaters for the tents. It took off again and set a course for Cosmos Lake, but it wasn’t long before there was a strong smell of gasoline all through the cabin. One of the heaters was leaking the stuff all over everywhere, which is just what you do not need, by and large, in an aeroplane. It returned to Yellowknife where everything was cleaned up and finally made it to Cosmos Lake. At this time (late March), the airstrip was still intact, but only just. By then the temperatures had moderated substantially, with the spring thaw not far off. The weight of accumulated snow on each side of the ice strip (which was still being regularly cleared by bulldozers), was causing the ice to bow up slightly in the middle, creating stress-fractures and weakening it. The big aircraft landed without incident, but didn’t stay any longer than necessary, and two of the engines were kept running ready for an immediate full-throttle takeoff if the need arose.
By early April there were only a few of people left to wrap up the search, which had been continuing on a daily basis since the end of January, although with a much lower priority after all the initial furore had subsided. The original orders from cabinet level in Ottawa were to "continue the search with all available resources until there is no further risk to public health and safety". Easy for the politicians to say when they hadn’t a clue what constituted a ’risk’. Garland and I both felt that all that could reasonably be done had been done, given the difficult circumstances. Were there any more bits of hot stuff out there? Probably, but the team could have spent the rest of their lives looking for them and still not be certain of having found all of them in an area as vast and desolate as the one covered by the footprint of the satellite re-entry. Garland suggested that a memo be drafted to ’Mount Olympus’ advising the powers that be that it was time to call it off. The memo was a committee effort by Garland, me (as the GSC representative at that time), the representative from the Atomic Energy Control Board, and one of the last remaining Americans. It laid out the rationale for ending the search with appropriate recommendations for a more detailed search in the summer, when it would be a more practical proposition. Somewhat to everyone’s surprise it was accepted, which meant that all the participants could stand down and return to their normal lives.
The non-radioactive fallout from the affair, political journalistic and bureaucratic, continued for many months. There were all sorts of articles published about it from all sorts of perspectives, including substantial articles in Time and Readers Digest. Some months later a rather lurid and melodramatic account of the whole episode was published as a hard-cover book entitled: ’Operation Morning Light, Terror in our Skies, the True Story of Cosmos 954’ , written by Canadian author Leo Heaps. He interviewed all the people he could, and then went away and produced his pot-boiler. It made a good read, but accuracy and balance were not front and centre.
Some time later, in a rather bizarre twist, a Canadian company, McPhar Geophysics, got a contract worth nearly two million dollars to sell airborne gamma-ray spectrometry equipment based on the GSC design, to the Soviet Union for monitoring their own nuclear facilities. That gave all concerned considerable satisfaction, because it meant that all of the effort which had gone into that project was finally reaping handsome dividends for Canada, which is what is supposed to result from government research and development programs of the sort that we were doing.
One of the more comic opera aspects of the demobilisation was that one of the NOVA minicomputers belonging to the GSC somehow went missing in the great trek back from Edmonton. This was not discovered until the GSC team had been back for a little while and we had begun to pick up where we had left off. We searched high and low for the missing computer in various military hangars, aircraft and warehouses to no avail. Finally I rang up Thane Hendricks in Las Vegas and asked him if by any chance he had an extra NOVA kicking around. They had so many that one more would hardly be noticed. Sure enough, it was there and he was profuse in his apologies. They must have scooped it up along with all the others and stuffed into the maw of their giant C-141 cargo aircraft for the return journey without realising it. They sent it back post haste and soon the GSC was notified that it was being held in customs in Ottawa. When Holman went to clear it, they insisted that it was being imported and the GSC had to pay duty on the thing which really annoyed everyone. It was just another example of how big government can trip itself up by one agency slavishly following the rules and red tape to the detriment of all concerned.
The powers that be, both military and civilian, wanted joint reports on the whole affair, which was no surprise. I went, as the GSC representative, with Garland and some of his people and representatives of the Atomic Energy Control Board, to Las Vegas, to meet with the American team and thrash out some of the details. We were given the royal tour of the Nevada Test Site by helicopter, where the atmospheric atomic bomb tests of the 1950’s were conducted, but where by then (spring 1979) only underground tests were done. We saw some of the results of ’Project Ploughshare’, the attempts to demonstrate that it was possible to use carefully planned underground nuclear explosions to do the huge excavations necessary for canals and other megaprojects. The whole visit provided a fascinating (but mostly classified) vignette of their operations.
Various official reports were produced of course and one of them, (very much for public consumption) had on the cover the imaginative and evocative logo designed fairly early on in the Operation by Connie Cho, a graphics artist on the American team. It became a sort of mascot for everyone connected with the operation and was incorporated into T-shirts and other items. The key element was a small but visually effective modification of the second letter ’G’ in ’Morning Light’, to depict the Soviet hammer and sickle emblem. It was unmistakable and instantly recognisable and neatly highlighted the Russian connection.
A properly documented history of Operation Morning Light was finally put together five years later in 1983 by Dick Morrison, one of the air force officers who was himself a helicopter pilot and one who had participated in the operation. He had been subsequently attached to the Directorate of Military History in Ottawa, where he was given the resources to produce a hard-cover book entitled ’Voyage into the Unknown’. Although not a tale of gripping drama in the style of Heaps’ book, it represented a valuable and coherent record of what happened where and when, and the problems that came close to derailing the operation on several occasions.
One of the points he made concerned the recommendations which I had included in the brief that Col. Garland had sent to Ottawa. One of these was for the creation of a ’paper team’ of experts by the Federal Government, assembled from various disciplines with the necessary resources assigned to it, which could be activated at short notice to deal with any similar emergency involving the accidental release of radioactive material. That recommendation was not acted upon at the time, possibly because (as Morrison suggests) it would have involved cutting across separate Federal and Provincial jurisdictional areas, or more likely because the probability of such another similar accident seemed comfortably remote at that time.
The Cosmos-954 satellite released an enormous amount of radioactivity across the frozen barren lands of Canada’s Northwest Territories in 1978. It did not result in a major catastrophe for two reasons. The first is the obvious one that the region is sparsely populated, and the second one was that almost fifty percent of the area is covered by lakes and rivers, which eventually absorbed most of the radioactive debris with minimal harm to the environment. What was probably not appreciated was that had the re-entry happened a little less than three orbits earlier, the impact trajectory would have strewn radioactive debris over a footprint extending from the Gulf of Mexico and passing close to the major population centres of Detroit and Toronto. Two and-a-bit orbits later and it would have re-entered over Hudsons Bay, putting Quebec City directly on the impact trajectory. In either case there would have been absolute mayhem. It took the near melt-down of the reactor at Three Mile Island in Pennsylvania a year later in 1979, to convince people that the idea of a Nuclear Emergency Response Team was one whose time had come with a vengeance.
Since then (incredibly), two more of the nuclear-powered Soviet Cosmos series spy satellites have re-entered out of control. Fortunately both landed harmlessly in the oceans. Meanwhile the 1986 Chernobyl affair dwarfed all other incidents of accidental release of radioactivity. The really bizarre aspect of that was the fact that the huge radioactive cloud which drifted over western Europe was not detected until a worker entering an atomic research plant in Sweden at the beginning of the day, tripped a radiation alarm that was in place to monitor workers for radioactive contamination on leaving the plant! Since then of course governments the world over have set up Nuclear Emergency Response Teams and have invested in radiation monitoring equipment both for ground stations and for airborne use in emergency situations.