GCHQ’s sigint ship project was still going strong in 1967, and there was a constant stream of new ideas. In their desperate efforts to find new ways of listening in to the Soviets, American scientists had come up with a bizarre project called ‘Moonbounce’. This exploited the possibility of picking up faint echoes of Soviet communications that were inadvertently reflected back towards earth from the moon. The special technical equipment on the sigint ship would be ideal for Moon-bounce, and was to work in tandem with an American satellite installation at RAF Oakhanger. The Americans had hoped to build this in Turkey, but the government in Ankara had been troublesome and the plan fell through. The British had then rushed forward, explaining that a site could quickly be made available at Oakhanger. With the Americans meeting the capital costs, this was considered ‘a good bargain’.12
Moon-bounce was a vast programme for the Americans, who had already begun what was innocuously called the ‘Naval Radio Research Station Program’ at Sugar Grove in West Virginia. Six years of studies had revealed that the moon reflected all kinds of radio energy from central Siberia with ‘minimum distortion’. This was hailed as facilitating a radically new approach to intelligence-gathering that would provide ‘unique daily intercept’ from Soviet radars, communications nets and other electronic emitting devices. The Sugar Grove facility would work as a genuine scientific research station for radio astronomy when the moon was out of view, but as a massive sigint collection station when it was accessible. However, by the late 1960s satellites were proving to be a better way of collecting these signals, and were making Moon-bounce redundant.13 Moreover, back in Britain there were growing arguments over money, with GCHQ and the Navy attempting to offload most of the cost of the sigint ship on each other.14 Both GCHQ and NSA were encountering serious financial problems, and were having to choose between projects.15
The final death knell for Britain’s sigint supership was sounded elsewhere. On 8 June 1967, one of NSA’s larger sigint ships, the USS Liberty, was tracking developments during the Six-Day War between Israel and Egypt in the Sinai Desert. The Liberty was busy intercepting both Egyptian and Israeli traffic while anchored twelve miles off the Sinai coast near El Arish. One of its tasks was to try to pick up any evidence of Soviet advisers assisting the Egyptians. Planners were aware that the ship was in a dangerous spot, and ordered the commander to move further away from the coast, but these instructions were not received. During the afternoon the Israelis, who had been keeping the ship under aerial surveillance for several hours, launched a ferocious attack by air and sea. The first was conducted by Mirage jet fighters, and the second by torpedo boats. By the time the torpedo boats attacked, the Liberty’s crew had hoisted an enormous flag that measured no less than seven feet by thirteen, declaring her to be an American ship.16
Thirty-four of the Liberty’s crew died and 170 were wounded. The ship, with eight hundred holes from prolonged rocket and cannon fire, managed to limp away to Malta. In July 1967, the President’s Foreign Intelligence Advisory Board, with full access to sigint intercepts of Israeli radio traffic, concluded that although Israeli aircraft had correctly identified the Liberty as an American ship, the subsequent attack was the result of incompetence and miscommunication.17 However, James Bamford, whose sigint history was written with the cooperation of NSA, asserts that the attack was deliberate, and that the Liberty was prominently displaying the American flag. He claims that the reason for the attack was that the Israelis feared that the Liberty had been eavesdropping on a massacre of Egyptian prisoners of war. Lieutenant General Marshall Carter, the Director of NSA, also continued to insist that it had been a deliberate attack. The issue of the Liberty remains a matter of bitter controversy.18
Months later there was a further disaster. In January 1968 the USS Pueblo, a smaller NSA vessel, was captured by North Korea while eavesdropping off its coast, leading to an eleven-month stand-off. The crew were poorly trained, and the facilities to destroy secret equipment and documents if threatened were close to non-existent. Accordingly, the haul of cryptological materials secured by the North Koreans and their Soviet advisers was considerable.19 Some have argued that the real purpose of the capture of the Pueblo was acquiring an American KL-47 cypher machine. This would have been invaluable, since the KGB had already received the keys to this cypher from one of its top agents in the United States, John Walker, an American naval communications expert.20 Sigint ships, which had looked so attractive a few years before, were now considered a liability. The Liberty and Pueblo disasters had alarmed Cheltenham; however, the main problem was overambitious plans colliding with shrinking budgets. In the autumn of 1968, soon after Dick White was appointed as Cabinet Office Intelligence Coordinator, he was asked to conduct a further inquiry into the alarming issue of the spiralling sigint budget. GCHQ’s supership was cancelled. New sigint aircraft were now catching the attention of planners, while what remained for naval sigint was redirected towards the latest submarines.21
Hitherto, much of Britain’s seaborne sigint effort had consisted of a partnership between GCHQ and the Navy’s venerable Super-T submarines. In 1960 the intrepid submarine commander John Coote had taken HMS Totem on several listening missions around the Soviet Union’s Arctic coast, but he confessed that his vessel was now ‘well past…her “sell-by” date’.22 Surprisingly, instead of being scrapped, both the Totem and the Turpin, two veteran spy subs, were purchased by the Israeli Sea Corps, along with HMS Truncheon. In 1965 the ageing submarines were refitted, while a substantial number of Israeli personnel were trained by the British at HMS Dolphin. The Truncheon and the Turpin then transferred to the main Israeli naval base at Haifa. However, when the Totem, now renamed the Dakar, Hebrew for ‘Swordfish’, prepared to set sail on 9 January 1968, those watching on the quayside were alarmed. The Super-Ts were designed to take a crew of sixty, but in order to get all the trainees home, she embarked a total of sixty-nine people. Her new modifications included a special ‘wet and dry’ air lock alongside the conning tower for the despatch of special forces – which added to her weight.23 Dockers observing her final preparations on the quayside were also astonished to see many crates of contraband – mostly whisky – also being loaded, followed by dismantled motorcycles. Grossly overloaded, she set sail around midnight.24
A week later the Dakar arrived safely at Gibraltar, then set out across the Mediterranean. Her last reported position was somewhere east of Crete on 24 January. Nothing more was heard of her. Her wreck was finally discovered in 1999, south-west of Cyprus, at a depth of over nine thousand feet. She was found by the Nauticus Corporation, the same salvage team that located the Titanic. Something had caused the submarine to dive below her maximum pressure depth, and she had suffered a catastrophic implosion of her hull. The Israelis salvaged the conning tower, which is now on display at Israel’s Naval Museum in Haifa. However, few of those passing this striking memorial are aware of the secret sigint past of the Dakar and its sister ships, or of their eventful missions inside the Arctic Circle.
Meanwhile, in Britain, the baton of special submarine operations passed to the new ‘O’ class submarines, two of which were fitted out for sigint activities. For their commanders, the biggest anxiety was being forced to the surface by Soviet depth-charging or technical failure, and then captured. Alfie Roake, one of Britain’s most intrepid spy-sub commanders, recalls that the question they continually asked themselves through the 1970s was how to avoid becoming another Pueblo.25 In December 1968, British naval officers explained the division of labour: ‘Some strategic Comint gathering is undertaken for GCHQ, but they carry out the tasking and evaluate the results.’ The Navy itself was more interested in elint, and helped to run a cell at Cheltenham which analysed radars that fed into a Comprehensive Comparative Radar Library. Thereafter, shipborne analysis computers drew on this library of taped intelligence to identify particular Soviet ships.26 In the late 1960s attention was shifting to an exciting new task: the tracking of the first Soviet submarines capable
of carrying ballistic missiles, Moscow’s equivalent of Polaris. They first entered service in 1967, and there were normally four parked off the coast of the United States at any one time.
Britain’s submarine operations were increasingly carried out in cooperation with her Commonwealth allies, who queued up to buy the new ‘O’ class submarine, and enjoyed the thrill of replicating her special missions. Australia acquired its first ‘O’ boats in 1967, and senior British officers and technicians transferred to the Royal Australian Navy with them. The Australian sigint target was the burgeoning Soviet naval power in the Pacific. The ‘mystery boats’ were always crowded, carrying their standard crew of sixty-two together with up to a dozen civilian sigint specialists. Australia’s specialism was the ‘underwater look’, a perilous manoeuvre that permitted very close reconnaissance of the propellers, propulsion sounds, sonar fit and electronic signature of the enemy ship. Admiral Peter Clarke, who served first in the Royal Navy and then in the Royal Australian Navy, recalls that the ideal position was just outside a harbour, where ships slowed to five knots. The submarine would close on its quarry at depth, and then gradually rise just ahead of the ship. Skilled captains would position their periscope about six feet below the vessel. Special lights and cameras would then scan the underside, while hydrophones and receivers recorded its emissions. A good captain might make two passes. ‘But it was a very full-on thing,’ adds Clarke. ‘You were driving several thousand tons of submarine to within feet of a vessel that you could not see.’27
In the 1970s the ships and submarines working for GCHQ were outshone by a new British sigint aircraft called the Nimrod R1. This state-of-the-art spy plane was the prestige intelligence project of the decade, entering service in 1974. Although GCHQ could not afford to join the Americans in the expensive game of satellite collection, a fleet of dedicated sigint aircraft was something no other European country had. Rather like nuclear weapons, it was something that marked Britain out as special.28 The main advantage of airborne sigint was that it allowed detection of signals at great range, often from within the airspace of a neighbouring friendly country. This in turn meant that a great volume of signals could be collected. Aircraft could also collect in areas where the creation of ground sites was not possible, and could move around rapidly to catch targets of opportunity, typically a major military exercise. The downside was the huge density of signals that were being received by a small number of personnel in one aircraft, who often found themselves overwhelmed by the volume.29
Discreet discussions on replacement of the existing Comet sigint aircraft had begun more than a decade earlier, perhaps as early as 1961.30 A highly secretive group called the Technical Committee of London Signals Intelligence Committee, headed by Ken Perrin, began work on the future direction of elint research in November 1961.31 Perrin’s group had been a key centre for the development of Peter Wright’s ‘Airborne Rafter’ programme that had hunted for KGB spies using a special aircraft. They knew NSA was devoting huge effort to strategic elint, so GCHQ decided to stay away from this area, ‘with the exception of special purpose equipment in quiet bands’. GCHQ’s plan was to focus upon niche areas, such as tactical collection along the border with the Eastern Bloc.32 Plessey was awarded a development contract for what was called ‘an experimental sideways-looking elint system’ in 1962, which was effectively the beginning of the Nimrod R1 programme.33 This new equipment revealed the ever-increasing complexity of sigint. The existing Comet aircraft depended on teams of human operators wearing headphones who undertook the reception and analysis manually, using narrow-band receivers. However, the growing density and complexity of electronic signals meant that they were simply being overwhelmed. It was ‘impossible for the operator to sort out and examine all the active transmissions’ in the limited time that an aircraft spent over the search area. This meant that the most interesting material, the unusual signals that might mean new enemy equipment, was being lost. Plessey’s new system was designed to do much of the work of the operator, and store what it detected for leisurely analysis after the aircraft had returned from its mission.34
The key decisions on the Nimrod were taken in March 1964 by the London Signals Intelligence Committee, which oversaw all of British sigint. It agreed that although the cost of replacing the Comets ‘represents a significant proportion of the UK expenditure on Sigint’, it was nevertheless essential. It was not just that some signals were otherwise inaccessible, but also that problems over bases meant a continual reduction in ground stations in areas such as Africa, causing GCHQ ‘the greatest difficulty’. The committee noted that stations in Iran and Turkey, which were the source of much intelligence on new Soviet radars, ‘could be denied to us’ at any time, adding ‘this had already nearly happened in Turkey’. It then compared the costs of possible aircraft, including the Boeing 707, but the Nimrod R1, which was a specialist variant of the RAF’s maritime patrol aircraft, was by far the cheapest.35
The Nimrod had strong backing from the Chiefs of Staff.36 It bolstered the much-prized special Anglo–American intelligence relationship by making a bigger contribution to shared sigint.37 As Air Vice Marshal Harold Maguire, the Deputy Chief of Defence Staff for Intelligence, explained, there was a limited choice of partners to share burdens with, for although the French ran airborne elint missions, there was no exchange with them. Other NATO partners, such as Norway and Turkey, offered full cooperation, but mostly used ground-based stations, and ‘in no case do they receive the full benefits of US resources as we do’. Britain had bilateral elint agreements with other European countries, such as Sweden, offering some of the benefits of airborne elint in exchange for ‘basic useful data collected by their ground stations in their own areas’.38 The Germans had begun to acquire an airborne capability, and were heavily advised by the British, but their aircraft were small.39 Overall, there was a clear hierarchy. The British were subordinate to the Americans, but enjoyed special status as a sophisticated collector. The Anglo–American partnership in Europe was especially close because of ‘common tasking through joint monthly meetings’, which allowed the British to focus on searching for new signals or low-level air defence analysis.40
Britain was the only NATO country receiving raw American sigint. Accordingly, Maguire explained, it was the only country with the capability to make its ‘own assessments in our area of interest and, where necessary, challenge US assessments’. This had been ‘critical’ when discussions on future NATO weapon-system requirements had occurred. He added: ‘We know that our relatively small airborne ELINT programme is appreciated by the Americans as a sharing of the collection task, particularly as their resources are stretched because of worldwide commitments.’ Because of the American need to focus on Vietnam in the late 1960s, the RAF was ‘covering areas that they leave entirely to us’. Some argued that Washington had come to expect the British to do much of the airborne sigint in Europe, ‘and a failure would threaten the massive help they give us in the whole Sigint area’.41 All this reflected bad relations between NSA and the US Air Force during the Vietnam War, which resulted in a lot of duplication of effort and a drain on American resources.42
No one questioned the Nimrod rationale, but hiding the costs almost amounted to a secret operation in its own right.43 The total cost of £14 million could not be accommodated within the already tight sigint budget. On 26 July 1967, Burke Trend concluded that it was unacceptable for the Nimrods to gobble up all the available funds in that budget; but since they were an essential purchase they were declared to be an integral part of Britain’s nuclear strategic weapons programme, as their intelligence supported targeting. Hence the cost of the Nimrods was not only kept out of GCHQ’s budget, it was even kept out of the larger and more elusive overall ‘Cost of Sigint’ spending. Instead it became part of what the RAF called ‘our overall contribution to the hidden SIGINT costings’.44 This meant that all of the RAF’s signals activity was swallowed by the more prestigious RAF Strike Command to tie sigint in with nucle
ar weapons.45
As a result, GCHQ began to give more focus to the Nimrods’ wartime role.46 One of their tasks in a future war would be to compensate for the loss of the sigint units in Berlin, which commanders expected to be sabotaged by the KGB on the eve of war and then ‘quickly overrun by Soviet forces’.47 The Nimrods also helped Britain to have a say in nuclear war planning for Europe, which was mostly conducted on an Anglo–American basis.48 Increasing emphasis on wartime use was reflected in the new ground facilities at RAF Wyton, which included rapid comint transcription facilities and a full analysis-assessment cell. The idea was that ‘hot intelligence reports’ could be sent quickly to field commanders, while deeper analysis would proceed at a more leisurely pace in GCHQ.49 In its comint role, planners expected to be able to cram seventeen voice operators and four supervisors inside the Nimrod.50
In October 1968, GCHQ’s Director of Plans, Ken Perrin, and the RAF sigint managers embarked on a prolonged visit to Germany to grapple with the issue of support for forward commanders. GCHQ was still agonising over the release of sigint to NATO allies ‘in the period of extreme tension leading to war’.51 It also had to decide on the time spent training for wartime tactical sigint, set against what Perrin called ‘strategic supplementary tasks’, in other words the Nimrods’ contribution to GCHQ’s work in peacetime.52
GCHQ Page 29