Indeed, rudimentary passive and active sonar techniques had already been used in World War I to search for submarines, but these earliest systems, at relatively high frequencies, achieved detection ranges of only several thousand yards under favorable conditions – and World War II sonars seldom did much better. The basic physical phenomena subsequently exploited in SOSUS to achieve longer-range submarine tracking were only discovered in the late 1930s and not adequately understood until mid-way through the 1939-1945 war.

An important early step in developing more effective sonar systems – and SOSUS in particular – was the invention of the sonic depth finder (SDF) in the early 1920s as a direct outgrowth of the rudimentary active sonars used in World War I. Not only did the SDF advance the state-of-the-art in acoustic technology, but it also facilitated detailed depth and ocean-bottom surveys with a speed and accuracy never before available using lead-line techniques. This, in turn, led to growing interest in marine geology and the adaptation of seismic methods developed for use on land to geological exploration of the sea floor. It was in this context in 1937 that Lehigh University scientist Maurice Ewing made a seminal observation while doing seismic refraction experiments in three-mile-deep water in the North Atlantic. Using underwater explosive charges as sound sources, Ewing noted that a chain of impulsive echoes – generated by repeated reflections between the ocean bottom and the sea surface – was clearly perceivable onboard his research vessel. From this result, Ewing reasoned that even allowing for a significant loss of sound intensity at each bottom and surface encounter, the sound signal of the charge – particularly at the lower frequencies – was capable of traveling great distances underwater with only limited attenuation. He further postulated that if there were horizontal sound propagation paths in the deep ocean that avoided surface and bottom reflections – a so-called “deep sound channel” – acoustic signals could travel hundreds, or even thousands, of miles and still be detectable by judiciously located hydrophones.

Almost simultaneously, another crucial element appeared with the invention and refinement of the bathythermograph by scientists at the Massachusetts Institute of Technology (MIT) and the Woods Hole Oceanographic Institution. For the first time, the bathythermograph made possible the continuous measurement of ocean temperature with depth – and most importantly, the determination in detail
of how underwater sound speed varied with distance below the surface – the sound velocity profile (SVP). In conjunction with numerous in situ measurements of the SVP, growing theoretical understanding of how underwater sound “rays” are refracted – or bent – by vertical variations in the sound velocity provided the analytical tools to support Ewing’s hypothesis of long-range propagation paths under certain conditions of ocean temperature and depth.

In general, in warmer waters near the ocean surface, the sound speed is relatively high. At greater depths, where the water is increasingly cooler, the sound velocity decreases toward a minimum. At that point, pressure effects take over, and the sound speed begins to rise again as depth continues to increase. The deep sound channel is found at the depth where the sound velocity is a minimum. Because sound “rays” always tend to bend away from regions of higher sound velocity, a wave directed upwards from the sound channel axis will be refracted back down again – and a wave directed downwards will be bent upwards. Thus, sound paths from sources in the deep sound channel weave back and forth across the channel axis and – because they become “trapped” in a deep ocean layer away from the surface or bottom – can travel long distances with minimum attenuation. Moreover, if there exist propagation mechanisms available to bring near-surface sound down to the depth of the sound channel, those signals will also become trapped and traverse long distances with minimal loss. The sound channel axis is normally found at a depth of several thousands of feet, depending on thermal conditions, and because of the unusually warm waters of the Gulf Stream and Sargasso Sea, it lies more deeply in the Atlantic than in the Pacific.

The Deep Sound Channel – Growing ASW Potential

The first SOSUS stations – NAVFACs – were sited from Barbados to Nova Scotia on a huge semi-circle that opened onto the deepwater abyss west of the Mid-Atlantic Ridge. Later, additional Atlantic-area stations were established at Argentia, Newfoundland, Keflavik, Iceland, and Brawdy, Wales.

As part of the upsurge of ocean-acoustic research that accompanied the coming of World War II, Ewing and his colleagues performed a variety of at-sea experiments that further confirmed sound propagation in the deep sound channel, while also discovering the phenomenon of the near-surface convergence zone.¹ On the basis of these experiments, Ewing proposed in 1943 that the Navy develop a system for communicating over long ranges by detonating time-coded explosive charges in the sound channel itself. Accordingly, during the spring of 1944, he supervised a major sea test in which USS Buckley (DE-51) steamed outward from a stationary receiving ship, periodically dropping small explosive charges fused for various depths. These explosions were still clearly discernable until Buckley had to break off the trial at a distance of 900 miles. By the end of the war, the Buckley experiments had led to a subsequent effort to develop an air-sea rescue system known as SOFAR – for Sound Fixing and Ranging. In the SOFAR concept, downed pilots would drop small explosive charges to the depth of the deep sound channel, where their sound output could be expected to travel for thousands of miles to deep, bottom-mounted hydrophones and triangulated to locate the survivors. At the time, however, exploiting the SOFAR channel for submarine detection at long range seems not to have been suggested, although by mid-war, the U.S. Navy was already using ray-tracing methods tactically for sonar performance prediction.

Even after World War II ended in mid-1945, the Navy continued to support a strong research program in underwater acoustics and, in particular, made enough additional progress in understanding the deep sound channel to establish – with the Army Air Force – major SOFAR networks in both the Pacific and Atlantic.² However, with the onset of the Cold War and the growing danger of a Soviet submarine force based on the best of German World War II technology, the application of underwater sound specifically to anti-submarine warfare (ASW) became a top priority. By early 1950, the Navy had come to believe that Soviet submarines posed the greatest threat to America’s security and approached the Committee on Undersea Warfare (CUW), an academic advisory group empanelled in late 1946, for suggestions on studying the problem. The result was Project Hartwell, a series of MIT-organized technical meetings attended by top-level scientists and naval officers during the first half of the year. Not unexpectedly, long-range submarine detection was among a variety of undersea warfare topics discussed by the Hartwell participants. In this regard, physicist Frederick Hunt, former head of Harvard’s Underwater Sound Laboratory, electrified the gathering with a convincing argument that Ewing’s SOFAR channel could support long-range propagation modes sufficient for detecting submarines passively at distances of hundreds of miles. Moreover, frequencies below 500 Hz would penetrate readily to the deep sound channel from virtually any source depth. This insight – not universally accepted at the time – formed the scientific basis for SOSUS and made possible long-range undersea surveillance surprisingly early in the post-war era.

SOSUS – Early Engineering Development

As a key result of the Project Hartwell findings, the Office of Naval Research (ONR) in late 1950 funded a contract with the American Telephone and Telegraph company (AT&T) and its manufacturing arm, Western Electric, to develop an undersea surveillance system based on long-range sound propagation. Under this aegis, Bell Telephone Laboratories initiated a series of experimental trials by installing undersea listening arrays off Sandy Hook, New Jersey and Eleuthera in the Bahamas. Additionally, AT&T adapted its sound spectrograph, which had recently been invented as a tool for analyzing speech sounds, into a similar device called LOFAR – for Low Frequency Analysis and Recording – designed to analyze low-frequency underwater signals in near-real time. Both LOFAR and the spectrograph generated a frequency-versus-time representation of an incoming sound “bite” on which the time history of its spectral content was indicated by the blackening of specially-sensitized paper by an electrostatic stylus that swept repeatedly along the frequency axis. In this way, the presence of distinctive submarine sound signatures – comprising both broadband noise and discrete frequency components (“tonals”) – could be discerned against the ocean background in the composite signal picked up by an array. This body of work, largely at AT&T, was code-named Project Jezebel and placed under the direction of CAPT Joseph Kelly at the Bureau of Ships.

Meanwhile, the Navy continued to support Maurice Ewing, by then at Columbia University’s Hudson Laboratory, to study the general phenomenology of low-frequency underwater sound. This effort, augmented by additional work at Woods Hole and the Scripps Institution of Oceanography in California, was focused on establishing a solid understanding of long-range sound transmission and denoted Project Michael. When the findings of Projects Jezebel and Michael were brought together for the purpose of designing, engineering, and deploying the broad-area surveillance system envisioned by Hartwell’s Frederick Hunt, the resulting effort – with the highly classified acronym, SOSUS – was eventually given the unclassified designation, Project Caesar.

The first prototype of a full-size SOSUS installation – a 1,000-foot-long line array of 40 hydrophone elements in 240 fathoms of water – was deployed on the bottom off Eleuthera by a British cable layer in January 1952. After a series of successful detection trials with a U.S. submarine, the Navy decided by mid-year to install similar arrays along the entire U.S. East Coast – and then opted two years later to extend the system to the West Coast and Hawaii as well. These early SOSUS line arrays were positioned on the sea floor at locations that accessed the deep sound channel and oriented at right angles to the expected threat axis. Their individual hydrophone outputs were transmitted to shore processing stations called “Naval Facilities” – or NAVFACs – on multi-conductor armored cables.

At the NAVFACs, the acoustic signals were processed to create a fan of horizontal “beams,” each of which represented the composite sound signal from a small angular sector – on the order of two to five degrees wide – oriented in a particular azimuthal direction. Narrow-band time-frequency analysis in the spectral region was performed on these multiple beam outputs simultaneously using the LOFAR technique described above. The ability of narrow-band frequency analysis not only to discriminate against broadband ocean noise but also to identify characteristic frequencies associated with rotating machinery was key to detecting and classifying targets. A LOFAR analyzer was associated with each beam of each array served by a NAVFAC, and typically, the large watch floors were filled with hundreds of these “gram-writers” busily turning out LOFARgrams on “smoky paper” 24 hours a day. These records were scrutinized continually by specially-trained personnel looking for the distinctive submarine “signatures” which gave indication of a possible target along a given bearing line. Then, if simultaneous contacts were gained on multiple arrays in separated locations, the target’s position could be estimated by triangulation.

First-Generation Installations and Initial Operational Experience

The first NAVFAC built by the Caesar program was commissioned in September 1954 at Ramey Air Force Base in northwestern Puerto Rico.³ Before the end of the year, similar stations were in operation at Grand Turks and San Salvador in the Bahamas, and by late 1957, additional NAVFACs had been established at Bermuda, Shelburne (Nova Scotia), Nantucket, Cape May⁴, Cape Hatteras, Antiqua, Eleuthera, and Barbados. A glance at the map of the eastern North Atlantic makes clear the rationale for siting these first-generation listening facilities. They form a huge semicircle from Barbados to Nova Scotia, opening toward the deepwater abyss west of the mid-Atlantic Ridge. This provided both excellent coverage of the deep ocean basin off the eastern seaboard and the opportunity for contact correlation among arrays with widely separated vantage points. For optimum acoustic coupling with the deep sound channel, the arrays “looked” outward from the edge of the continental shelf, and because cable lengths were limited to somewhat less than 150 miles, the NAVFACS had to be located at coastal sites where the shelf break came closest to land. Two years later, this concept of operations was expanded to incorporate a SOSUS station at Argentia, Newfoundland to process the outputs of
a number of shallow-water arrays south of the Grand Banks.

The year 1957 also saw the extension of SOSUS to the Eastern Pacific, with the installation of NAVFACs and associated arrays at – from south to north – San Nicholas Island, Point Sur, and Centerville Beach, California; Coos Bay, Oregon; and Pacific Beach, Washington. Still later, additional arrays would be terminated at Guam, Midway, Adak (in the Aleutians), and Barber’s Point near Honolulu.

Operationally, the Navy intended SOSUS to provide early warning of hostile submarines entering the North Atlantic or Eastern Pacific, as well as generating “cueing” information for area ASW forces. By combining bearing data from separated arrays “holding” the same contact, geographic “probability areas” of target position could be calculated and passed to patrol aircraft, surface ships, or submarines to facilitate reacquisition of the target for fine localization and prosecution. This concept of operations necessitated the establishment of regional SOSUS Evaluation Centers – later called Naval Oceanographic Processing Facilities (NOPFs), with the first two at Norfolk and New York – that correlated contact information from multiple NAVFACs with other intelligence sources, such as radio direction-finding. The NOPFs then forwarded the resulting target position estimates and probability areas to local and regional ASW commands.

The primary threat against which SOSUS was originally designed was snorkeling Soviet diesel submarines at the surface, and the system’s key technical characteristics – such as frequency coverage – were established accordingly. Fortunately, the resulting capability proved even more effective against deep-running Soviet nuclear-powered submarines when the first of these went operational in 1958. In a 1961 demonstration of the capabilities of the system, SOSUS tracked the USS George Washington (SSBN-598) across the North Atlantic on her first transit from the United States to the United Kingdom. Then, in June 1962, NAVFAC Cape Hatteras achieved the first SOSUS contact on a Soviet diesel submarine, to be followed a month later with the first detection of a Soviet nuclear boat west of Norway by NAVFAC Barbados. Later that year, during the Cuban Missile Crisis, the first positive correlation with a visual sighting was made, when a patrol aircraft confirmed the presence of a Russian FOXTROT-class submarine that had already been detected by NAVFAC Grand Turks. In 1968, NAVFAC Keflavik made the first SOSUS detections of Soviet CHARLIE- and VICTOR-class nuclear submarines, and that same year, SOSUS played a key role in locating the wreckage of USS Scorpion (SSN-589), lost near the Azores in May. Moreover, SOSUS data from March 1968 facilitated the discovery and clandestine retrieval years later of parts of a Soviet GOLF-class submarine that foundered that month north of Hawaii.

Subsequently, as increasing numbers of Soviet submarines from bases in the Barents and White Seas achieved access to the North Atlantic by rounding northern Norway and steering south through the Greenland-Iceland-United Kingdom (GIUK) gap, the decision was taken to extend SOSUS into more northerly waters, and new NAVFACS were established at Keflavik, Iceland in 1966 and Brawdy, Wales in 1974. Also, better processing and cable technology allowed siting arrays farther from
shore and using “split array” techniques in which a single line array was divided into segments whose outputs were processed separately and then re-combined electronically to achieve narrower beams and greater directivity. In 1974, Keflavik was the first NAVFAC to detect a DELTA-class Soviet SSBN as it moved down into the North Atlantic.
This is a typical LOFARgram, in which frequency is portrayed along the horizontal axis and time along the vertical axis. The presence of spectral energy is indicated by a darkening of the paper. In this representation, a long, narrow, vertical line indicates a persistent narrowband component centered on a single frequency – perhaps a“tonal” caused by rotating machinery – along the bearing line of interest. Wider gray areas result from broadband noise sources or the ambient background.

Cold War Effectiveness – and Decline

As the Cold War deepened, and both the size and capability of the Soviet submarine fleet continued to grow, SOSUS became “the secret weapon” that enabled U.S. ASW forces to keep close track of virtually all potentially hostile submarines operating in the deepwater regions off both coasts. This capability was facilitated by geographical constraints that forced Soviet submarines into predictable deployment patterns and the rudimentary state of Russian acoustic quieting, which left their submarines some 30 dB noisier than U.S. counterparts – and hence easily detectable from thousands of miles away. In the mid 1980s, the network of fixed SOSUS arrays was augmented by a small fleet of civilian-manned, ocean-going, acoustic surveillance ships deploying the Surveillance Towed Array Sensor System (SURTASS), a towed line array over 8,000 feet long.⁵ By means of satellite communication links, contact information developed by the SURTASS ships at sea was passed to the SOSUS Evaluation Centers ashore and melded with data from the fixed arrays to establish position estimates for likely targets. In time, the totality of fixed arrays, shore processing facilities, and SURTASS ships became known as the Integrated Undersea Surveillance System (IUSS).

Eighteen first-generation SURTASS ships, of which USNS Indomitable (T-AGOS-7) is an example, were commissioned between 1984 and 1988. Largely civilian-manned, they displaced 2,260 tons on a length of 224 feet and deployed a towed line array more than 8,000 feet long on deep-ocean ASW patrols of 60–90 days duration.

Eventually, with the help of key information supplied by the Walker-Whitworth espionage ring⁶, Soviet intelligence learned of the existence of SOSUS and its remarkable success in tracking Soviet submarines at long range. Thus, beginning shortly after John Walker’s first treasonous revelations in 1968, the Russian navy embarked belatedly on a rapid submarine quieting program, and within five years, the radiated noise levels of their first-line boats had begun to drop recipitously. By the end of the Cold War in the late 1980s, Russian submarines were much closer to their U.S. equivalents, and the ability of IUSS to detect and track them at long range had deteriorated significantly. In an attempt to regain some of the acoustic advantage lost to Soviet quieting, IUSS system developers turned from long line arrays and their fans of pre-formed beams to large fields of simpler, “upward-looking” hydrophones densely distributed on the ocean floor, each capable of detecting submarines only in its immediate vicinity. Thus, detection and localization were subsumed into a single process, and the first “Fixed Distributed System” built in accordance with this strategy was deployed in 1985.

Moreover, with steady improvements in acoustic signal processing through the 1970s and 1980s, the first generation of shore-processing hardware, which turned out hundreds of single-beam LOFARgrams – 24 hours a day, every day – was gradually replaced by computer-based workstations that could analyze the incoming acoustic data digitally and display it on multiple computer screens. Additionally, to reduce manpower requirements and achieve other efficiencies, most of the original arrays were re-terminated at alternative shore sites or “remoted” to central processing facilities, which led to a steady reduction in the number of operational NAVFACs. These transitions were completed in 1997 and 1998, but by that time – ironically – quieter submarines, major changes in Soviet operating patterns, and finally, the end of the Cold War had already eliminated much of the justification for maintaining IUSS at its full capability.⁷

(above) The Soviets built more Delta-class SSBNs – 47 in four variants – than any other class of ballistic missile submarine. The first Delta I appeaered in 1972; the last Delta IV was completed in 1992. The earlier boats displaced 11,750 tons submerged and carried 12 strategic missiles. The later variants added approximately 2,000 tons and four more missiles. SOSUS first detected a Delta at sea in 1974.
	(left) Between 1958 and 1982, the Soviet Union built approximately 80 Foxtrot-Class conventionally-powered submarines, of which nearly 20 were intended for export. Displacing 2,400 tons submereged on a length of 300 feet, the Foxtrots became the workhorse of Russia’s diesel-powered submarine force during the height of the Cold War. Snorkeling conventional submarines were SOSUS’s original targets, and the first of these was detected in 1962.

IUSS Today – and Future Prospects

Today, while the Navy maintains a number of SOSUS arrays in either operational or standby status, only three shore facilities at Dam Neck, Virginia, Whidbey Island, Washington, and St. Mawgan, United Kingdom⁸ – remain to process their dwindling output. With few possibly-hostile, nuclear-powered submarines still operating at sea and modern, quiet, diesel-electric boats essentially undetectable at long range by passive means, there’s not a lot to listen for anymore, and targets of potential interest are rare. However, several existing arrays have achieved well-publicized successes in peacetime pursuits such as tracking migrating whales and detecting illegal driftnet fishing on the high seas. Moreover, as the Navy explores the use of low-frequency active (LFA) acoustics for detecting and tracking quiet submarines in the future, both the fixed arrays and the remaining SURTASS ships may well play an important role as adjunct or bi-static receiving sites.

When it was first suggested over 50 years ago as a means of exploiting contemporary oceanographic findings and state-of-the-art technology for wide-area undersea surveillance, SOSUS was an audacious concept, and its successful implementation was one of the most impressive engineering feats of the early Cold War. Later, during the most dangerous phases of that simmering conflict, IUSS gave the United States an unprecedented capability for long-range submarine detection and strategic early warning that we can only envy today in this new era of asymmetric threats.

Dr. Whitman is the former Senior Editor of UNDERSEA WARFARE Magazine. Much of his long Navy technical career was involved with SOSUS and undersea surveillance. Earlier, as a graduate student at MIT’s Speech Communications Laboratory, he was responsible for the maintenance and calibration of the sound spectrographs that only a few years earlier had evolved into SOSUS’s LOFAR signal processors.

Bibliography

Cote, Owen R. Jr., “The Third Battle: Innovation in the U.S. Navy’s Silent Cold War Struggle with Soviet Submarines,” MIT Security Studies Program, March 2000 (available at http://chinfo.navy.mil/navpalib/cno/n87/history/cold-war-asw.html [1])

Tyler, Gordon D., Jr., “The Emergence of Low-Frequency Active Acoustics as a Critical Antisubmarine Warfare Technology,” Johns Hopkins APL Technical Digest, Vol. 13, No. 1 (1992)

Urick, Robert J., Principles of Underwater Sound, 2nd Ed., McGraw-Hill, New York, 1975

Weir, Gary E., An Ocean in Common, Texas A&M, College Station, 2001

Additionally, two websites have provided useful information and general background: http://www.globalsecurity.org/intell/systems/sosus.htm
[2]
http://www.iusscaa.org [3]
(Home page of the IUSS-Caesar Alumni Association)

1 In deep-water sound propagation, a convergence zone (CZ) is a circular ring of anomalously-low transmission loss that is caused at predictable distances from a submerged sound source by the near-surface bunching of sound rays that originally left the source over a narrow range of vertical angles. Typically, CZs appear at multiples of 30 to 35 miles from the source, but the specifics are highly dependent on the SVP and local bathymetry. Generally, much higher-than-normal target-detection probabilities are observed in the CZs.

2 By 1950, a Navy SOFAR explosion had been heard over 3,500 miles away. Moreover, several rudimentary experiments in tracking submarines with the net had shown favorable results.

3 When Ramey AFB was disestablished in 1973, the NAVFAC remained in operation but was renamed Punta Borinquen.

4 After NAVFAC Cape May was destroyed by a hurricane, it was replaced in 1962 by a new facility at Lewes, Delaware.

5 The initial first-generation SURTASS ship was USNS Stalwart (T-AGOS-1), commissioned in April 1984. The last of the 18 ships of that class entered service four years later.

6 John Walker, Jr. was a U.S. Navy warrant officer and career submarine communications expert who over a period of 15 years sold countless naval messages and the keys to decipher them to the Soviets, thus revealing a vast amount of highly sensitive information about U.S. naval operations and capabilities. Later, Walker recruited another Navy communications specialist, Jerry Whitworth, his brother, Arthur, and even his own son, Michael Walker, before he was turned in by his wife. All four men were arrested in 1985 and subsequently prosecuted, but by then, enormous damage to U.S. security had occurred.

7 By 1991, the existence and missions of SOSUS and IUSS had been formally declassified.

8 NOPF Dam Neck was established in 1980 as part of a major consolidation of the Western Atlantic arrays. NOPF Whidbey Island was created in 1987 to perform a similar function for the Pacific. The Joint Maritime Facility (JMF) St. Mawgan, operated in conjunction with the United Kingdom, replaced the former NAVFAC at Brawdy, Wales, in 1995.

---

This article was originally published in Undersea Warfare, the magazine of the US Navy submarine force, and is reprinted here in full with no edits.