Impetus

Construction of the Pinetree Line had only just started when air planners started to have concerns about its capabilities and siting. By the time it detected a potential attack by jet-powered aircraft, there would be little time to do anything before the attack reached Canadian or northern U.S. cities.^[1] Additionally, the Pinetree systems used pulsed radars that were fairly easy to jam and were unable to detect targets close to the ground due to "clutter." Although expensive in terms of fuel use, it would be possible for Soviet bombers to evade detection by flying lower and plotting a course between the stations.

Bennett Lewis, head of the AECL Chalk River Laboratories and former Chief Superintendent of the UK Telecommunications Research Establishment (TRE) had proposed to the Defence Research Board (DRB) a system that avoided both of these problems.^[2] Known today as a forward scatter bistatic radar, it used two antennas, a transmitter and receiver, separated by some distance. The antennas were positioned and aimed so that the signal from the transmitter filled the space above the line between the two stations. An aircraft flying into this region would reflect some signal back towards the receiver, allowing detection at altitudes as great as 65,000 ft.^[1]

A major advantage of the system is that it requires much less power to operate effectively. In a conventional radar, the radio signal has to travel to the target and back again. As each leg of the journey is subject to the inverse square law, the resulting radar equation contains a fourth-power dependence. In contrast, a forward-scatter radar signal always travels about the same total distance, from the transmitter to the receiver, modified only by the altitude of the target. This means it is dependent on the square root of range and not the fourth root, and thus delivers considerably more energy onto the receiver than a conventional radar over the same range. Also, unlike a conventional "monostatic" radar, the transmitter did not have to turn off to allow the receiver to listen for the signal. Since the total amount of energy received at the receiver is a function of both the peak power and the length of the pulse, using a continual signal means the same total energy will be deposited using much lower peak transmitter power. As a result, Lewis' system would require smaller sites and much less power than conventional radars like those of the Pinetree Line.^[3]

The major disadvantage of the system is that it did not indicate the aircraft's location within the beam, unlike a pulsed system where pulse timing can be used to determine range. This means the forward-scatter concept is useful for making a "radar fence" or "trip wire"^[1] that indicates that something is approaching, but not exactly where it is. To help address locating the target to a degree, the proposal was to build two interlinked fences, so that each pair of stations was perhaps 30 kilometres (19 mi) apart, a short enough distance that the radar on an interceptor aircraft would be able to find the target within that area. Using two overlapping sets also allowed one pair to cover the dead zone directly above the towers of the other.

Lewis' initial concept was to place the transmitters and receivers on telephone poles and electric power transmission towers, which provided both a convenient location as well as the small amount of power needed to run the electronics. In the case of the telephone poles, the lines would also be used to send the data back to the tracking stations. This concept generated a considerable amount of interest, although it was abandoned for reasons that are not entirely clear. Willis and Griffiths speculate it might be the need for 1,000 such radars,^[4] but it is also likely that the desired to locate the line further north than the heavily settled areas in southern Canada was likely significant as well. In any event, the simplicity of the concept helped bring it to the attention of air planners.^[4]

Spider Web

The DRB decided to pursue Lewis’ idea in 1950–51 by directing a research contract to the Eaton Electronics Research Laboratories of McGill University, headed by Professor Garfield Woonton. Lewis suggested to DRB and Woonton that he put the project in the hands of associate professor, J. Rennie Whitehead as project leader, a former colleague of his from the TRE days in the UK who had recently taken a position at the Lab. Some preliminary tests were made in 1952 with breadboard hardware built by a graduate student, Hugh Hamilton, in order to confirm the validity of the idea.^[2]

In the meantime RCA Victor had been brought in by the DRB to design and produce the receivers, transmitters and antennas for tests on a substantial scale. The testing was performed in the summer of 1953, when Whitehead and his team of RCA Victor and RCAF personnel installed and operated a string of seven stations stretching from Ottawa to Mattawa^{[lower-alpha 1]} along the Ottawa River valley. Known under the code name of "Spider Web" at the suggestion of Hamilton, the tests were made with aircraft from CFB St. Hubert, near Montreal. All observations were transmitted to and made in the line HQ, which was set up in the equipment hut of one of the seven stations, located in Deep River.^[2]

Flight Lieutenant Andrew Matthews of the 104 Communications Flight at RCAF St. Hubert arranged for a series of different aircraft to fly through the network, including an Auster light aircraft, a T-33 Shooting Star, an Avro Lancaster bomber and even a recently acquired de Havilland Comet jet transport. The tests revealed a great deal about the spectral ‘signatures’ of aircraft crossing the line at different points, and demonstrated the capability to detect all sizes of aircraft from 100 ft to over 40,000 ft in altitude. During this time Dr. Ross Warren of RCA Victor and Dr. Whitehead jointly developed the theoretical background for the work in a major report to DRB.^[2]

The Spider Web trials were followed in 1954 by intensive tests on a single 30 miles (48 km) wide link, built in the Eastern Townships by Bell Canada, who had by this time been given the go-ahead for the implementation of the Mid-Canada Line. When Whitehead inquired why RCA had not been given the contract, a colleague replied "Who do you think runs Canada?"^{[2][lower-alpha 2]} The trials on this prototype link were also conducted by Whitehead and a small team in collaboration with Air Defence Command, St. Hubert, this time on behalf of Bell. The trials involved the flyover of numbers of B-52 bombers by arrangement with Strategic Air Command and a local bombplot unit. They also had full-time use of an Avro Lancaster from CFB Greenwood for the important low-level tests.^[2]

Deployment studies

In February 1953 the Canada-U.S. Military Study Group (MSG) was asked "to study those aspects of the North American Air Defence System in general, and the early warning system in particular, which are of mutual concern to the two countries."^[6] The MSG then asked the air defence commanders of Canada and the United States to prepare independent briefs on the subject. By July 1953, RCAF Air Defense Command had completed its brief, followed shortly thereafter by its USAF counterpart. Both reports suggested building a Doppler fence farther north, along the 55th parallel, roughly at the entrance of James Bay into Hudson Bay.^[6]

In October 1953 the MSG recommended to both governments "that there be established at the earliest practicable date, an early warning line located generally along the 55th parallel between Alaska and Newfoundland",^[6] and outlined their minimum operational requirements. By the end of November 1953, the Mid-Canada Line had been approved in principle. Unlike the jointly-operated Pinetree line and future DEW line, the Mid-Canada line would be funded and operated entirely by the RCAF. The DRB estimated that the system would cost about $69,700,000, while an independent RCAF report placed it at $85,000,000, equivalent to $219,200,000 in 2021.^[6]

In December an effort started to try to understand what sort of problems would be encountered during construction. Several "trains" consisting of tractor-pulled sleighs set out cross-country. One, manned by the RCAF, set out eastward from Fort Nelson, BC in order to link up with a second moving west from Flin Flon, Manitoba, while a third crewed by the Army left Lake Nipigon near Thunder Bay, Ontario for Lansdowne House about 200 kilometres (120 mi) further north. The missions proved that it was possible to build the new line, but only during the winter when the muskeg was frozen solid.^[6] These missions also inspired the U.S. Army to invest in purpose-built overland trains which they experimented with in the 1960s but never put into production.^[7]

While that was taking place, efforts were underway to start primary siting studies. It quickly became clear that the areas in question, at least in eastern Canada, were so remote that there was no really accurate topographical information. A huge effort to map the area in a 15-mile wide strip across the entire country was started by Transport Command almost immediately, and ended by the spring of 1954. With this information in hand a construction division was set up, the Systems Engineering Group (SEG), in February 1954, tasked with producing a final report on the system to be submitted on June 1.^[6]

Finalizing the plans

As the experiments continued it became clear that by using taller masts, 350 feet (110 m) tall, the radar stations could be located further apart, up to 90 kilometres (56 mi). This reduced the number of stations required. Nevertheless, the price rose, now estimated at about $120,000,000. Although their final report was not yet ready, the SEG put in an interim report in June and it was approved by cabinet by the end of the month.^[6]

In their report they outlined the system that would be built almost exactly. It called for eight major Sector Control Centres, numbered from 200 to 900, each of which control up to thirty unmanned radar sites for a total of 90 radar stations.^[6] Each of the radar stations consisted of a single tall mast with a number of small dishes in fixed positions on top (typically four, two pointed in either direction), with power and electronics located in a building beside the base of the mast.

The sector control centres were linked using an advanced microwave communications system developed in part by CARDE, which scattered off the troposphere for long-distance communications. The southernmost site along the eastern portion of the line, at Cape Henrietta Maria on Hudson Bay, was used as the main communications point, with three additional repeater stations transferring data from the line southward to the NORAD command center in North Bay, Ontario. The easternmost station at Hopedale, Labrador was co-located with an existing Pinetree Line station in order to save construction costs.^[6]

All aircraft transiting the line would have to file a flight plan through the Mid Identification Zone, or MIDIZ, centred on the fence. The plans also called for the construction of several airbases known as Line Clearance Aerodromes just to the north of the line, where interceptor aircraft could operate in times of heightened alert.^[6]

At about this time another huge civil engineering project was underway in Canada, the construction of a cross-Canada microwave relay telephone system. Since many of the logistics problems were similar, the construction group, led by Bell Canada, was selected as a major contractor for the base construction. Detailed site selection started in 1955, with a major surveying effort running across Canada at the 55th parallel. The sites were so remote that the RCAF had to form up its first all-helicopter squadron in order to provide flight support for the survey teams.^[6]