Over-the-horizon radar, or OTH (sometimes also beyond the horizon, or BTH), is a type of radar system with the ability to detect targets at very long ranges, typically hundreds to thousands of kilometres, beyond the radar horizon, which is the distance limit for ordinary radar. Several OTH radar systems were deployed starting in the 1950s and 1960s as part of early warning radar systems, but these have generally been replaced by airborne early warning systems instead. OTH radars have recently been making a comeback, as the need for accurate long-range tracking becomes less important with the ending of the Cold War, and less-expensive ground based radars are once again being considered for roles such as maritime reconnaissance and drug enforcement.
Radio waves, a form of electromagnetic radiation, tend to travel in straight lines. This generally limits the detection range of radar systems to objects on their horizon due to the curvature of the Earth. For example, a radar mounted on top of a 10 m (33 ft) mast has a range to the horizon of about 13 kilometres (8.1 mi), taking into account atmospheric refraction effects. If the target is above the surface, this range will be increased accordingly, so a target 10 m (33 ft) high can be detected by the same radar at 26 km (16 mi).
In general it is impractical to build radar systems with line-of-sight ranges beyond a few hundred kilometres. OTH radars use various techniques to see beyond that limit. Two techniques are most commonly used; shortwave systems that reflect their signals off the ionosphere for very long-range detection, and surface wave systems which use low-frequency radio waves that due to refraction follow the curvature of the Earth to reach beyond the horizon. These systems achieve detection ranges of the order of a hundred kilometres from small, conventional radar installations.
The most common type of OTH radar uses ionospheric reflection. Given certain conditions in the atmosphere, radio signals broadcast up towards the ionosphere will be reflected back towards the ground. After reflection off the atmosphere, a small amount of the signal will reflect off the ground back towards the sky, and a small proportion of that will reflect back towards the broadcaster. Only one range of frequencies regularly exhibits this behaviour: the high frequency (HF) or shortwave part of the spectrum from 3 to 30 MHz. The "correct" frequency to use depends on the current conditions of the atmosphere and the sunspot cycle, so systems using ionospheric reflection typically employ real-time monitoring of the reception of backscattered signals to continuously adjust the frequency of the transmitted signal.
The resolution of any radar depends on the width of the beam and the range to the target. For example; a radar with 1 degree beam width and a target at 120 km (75 mi) range will show the target as 2 km (1.2 mi) wide. To produce a 1 degree beam at the most common frequencies, an antenna 1.5 kilometres (0.93 mi) wide is required. Due to the physics of the reflection process, actual accuracy is even lower, with range resolution on the order of 20 to 40 kilometres (12–25 mi) and bearing accuracy of 2 to 4 kilometres (1.2–2.5 mi) being suggested. Even a 2 km accuracy is useful only for early warning, not for weapons fire.
Another problem is that the reflection process is highly dependent on the angle between the signal and the ionosphere, and is generally limited to about 2 to 4 degrees off the local horizon. Making a beam at this angle generally requires enormous antenna arrays and highly reflective ground along the path the signal is being sent, often enhanced by the installation of wire mesh mats extending as much as 3 kilometres (1.9 mi) in front of the antenna. OTH systems are thus very expensive to build, and essentially immobile.
Given the losses at each reflection, this "backscatter" signal is extremely small, which is one reason why OTH radars were not practical until the 1960s, when extremely low-noise amplifiers were first being designed. Since the signal reflected from the ground, or sea, will be very large compared to the signal reflected from a "target", some system needs to be used to distinguish the targets from the background noise. The easiest way to do this is to use the Doppler effect, which uses frequency shift created by moving objects to measure their velocity. By filtering out all the backscatter signal close to the original transmitted frequency, moving targets become visible. Even a small amount of movement can be seen using this process, speeds as low as 1.5 knots (2.8 km/h).
This basic concept is used in almost all modern radars, but in the case of OTH systems it becomes considerably more complex due to similar effects introduced by movement of the ionosphere. Most systems used a second transmitter broadcasting directly up at the ionosphere to measure its movement and adjust the returns of the main radar in real-time. Doing so required the use of computers, another reason OTH systems did not become truly practical until the 1960s, with the introduction of solid-state high-performance systems.
A second type of OTH radar uses much lower frequencies, ones that will diffract around the surface of the earth, and especially over the sea. Like the ionospheric high-frequency systems, the received signal from these groundwave systems is very low, and demands extremely sensitive electronics. Because these signals travel close to the surface and lower frequencies produce lower resolutions, low-frequency systems are generally used for tracking ships, rather than aircraft. However, the use of bistatic techniques and computer processing can produce higher resolutions, and has been used as of 1990s.
Engineers in the Soviet Union are known to have developed what appears to be the first operational OTH system in 1949, called "Veyer". However, little information on this system is available in western sources, and no details of its operation are known. It is known that no further research was carried out by Soviet teams until the 1960s and 70s.
Much of the early research into effective OTH systems was carried out under the direction of Dr. William J. Thaler at the Naval Research Laboratory. The work was dubbed "Project Teepee" (for "Thaler's Project"). Their first experimental system, MUSIC (Multiple Storage, Integration, and Correlation), became operational in 1955 and was able to detect rocket launches 600 miles (970 km) away at Cape Canaveral, and nuclear explosions in Nevada at 1,700 miles (2,700 km). A greatly improved system, a testbed for an operational radar, was later built in 1961 as MADRE (Magnetic-Drum Radar Equipment) at Chesapeake Bay. It demonstrated its ability to detect aircraft as far as 3,000 kilometres (1,900 mi) using as little as 50 kW of broadcast energy.[N 1]
As the names imply, both of the NRL systems relied on the comparison of returned signals stored on magnetic drums. In an attempt to remove clutter from radar displays, many late-war and post-war radar systems added an acoustic delay line that stored the received signal for exactly the amount of time needed for the next signal pulse to arrive. By adding the newly arrived signal to an inverted version of the signals stored in the delay line, the output signal included only the changes from one pulse to the next. This removed any static reflections like nearby hills or other objects, leaving only the moving objects like aircraft. This basic concept would work for a long-range radar as well, but had the problem that a delay line has to be mechanically sized to the pulse repetition frequency of the radar, or PRF. For long-range use, the PRF was very long to start, and deliberately changed in order to make different ranges come into view. For this role the delay line was not usable, and the magnetic drum, recently introduced, provided a convenient and easily controlled variable-delay system.
Another early shortwave OTH system was built in Australia in the early 1960s. This consisted of several antennas positioned to be four wavelengths apart, allowing the system to use phase-shift beam forming to steer the direction of sensitivity and adjust it to cover Singapore, Calcutta and the UK. This system consumed 25 miles (40 km) of electrical cable in the antenna array.
The first truly operational development was an Anglo-American system known as Cobra Mist. Built starting in the late 1960s, Cobra Mist used an enormous 10 MW transmitter and could detect aircraft over the western Soviet Union from its location in Suffolk. When system testing started in 1972, however, an unexpected source of noise rendered it largely unusable. They eventually abandoned the site in 1973, the source of the noise never having been identified.
U.S. Air Force
The United States Air Force Rome Laboratory had the first complete success with their AN/FPS-118 OTH-B. A prototype with a 1 MW transmitter and a separate receiver was installed in Maine, offering coverage over a 60 degree arc between 900 and 3,300 km. A permanent transmitting facility was then built at Moscow AFS, a receiving facility at Columbia Falls AFS, and an operational center between them in Bangor, Maine. The coverage could be extended with additional receivers, providing for complete coverage over a 180-degree arc (each 60 degree portion known as a "sector").
GE Aerospace was awarded the development contract, expanding the existing east coast system with two additional sectors, while building another three-sector system on the west coast, a two-sector system in Alaska, and a one-sector system facing south. In 1992 the Air Force contracted to extend the coverage 15 degrees clockwise on the southern of the three east coast sectors to be able to cover the southeast U.S. border. Additionally, the range was extended to 3,000 miles (4,800 km), crossing the equator. This was operated 40 hours a week at random times. Radar data were fed to the U.S. Customs/Coast Guard C3I Center, Miami; Joint Task Force 4 Operations Center, Key West; U.S. Southern Command Operations Center, Key West; and U.S. Southern Command Operations Center, Panama.
With the end of the Cold War, the influence of the two senators from Maine was not enough to save the operation and the Alaska and southern-facing sites were canceled, the two so-far completed western sectors and the eastern ones were turned off and placed in "warm storage," allowing them to be used again if needed. By 2002, the west coast facilities were downgraded to "cold storage" status, meaning only minimal maintenance was performed by a caretaker.
Research was begun into the feasibility of removing the facilities. After a period of public input and environmental studies, in July 2005 the U.S. Air Force Air Combat Command published a "Final Environmental Assessment for Equipment Removal at Over-the-Horizon Backscatter Radar - West Coast Facilities". A final decision was made to remove all radar equipment at the west coast sector's transmitter site outside Christmas Valley, Oregon and its receiver site near Tulelake, California. This work was completed by July 2007 with the demolition and removal of the antenna arrays, leaving the buildings, fences and utility infrastructure at each site intact.
The United States Navy created their own system, the AN/TPS-71 ROTHR (Relocatable Over-the-Horizon Radar), which covers a 64 degree wedge-shaped area at ranges from 500 to 1,600 nautical miles (925 to 3,000 km). ROTHR was originally intended to keep track of ship and aircraft movement over the Pacific, and thus allow coordinated fleet movements well in advance of an engagement. A prototype ROTHR system was installed on the isolated Aleutian Island of Amchitka, Alaska, monitoring the eastern coast of Russia, in 1991 and used until 1993. The equipment was later removed into storage. The first production systems were installed in the test site in Virginia for acceptance testing, but were then transitioned to counter the illegal drug trade, covering Central America and the Caribbean. The second production ROTHR was later set up in Texas, covering many of the same areas in the Caribbean, but also providing coverage over the Pacific as far south as Colombia. It also operates in the anti-drug trafficking role. The third, and final, production system was installed in Puerto Rico, extending anti-drug surveillance past the equator, deep into South America.
The Soviets had also studied OTH systems starting as early as the 1950s. Their first experimental model appears to be the Veyer (Hand Fan) that was built in 1949. The next serious Soviet project was Duga-2, built outside Nikolayev on the Black Sea coast near Odessa. Aimed eastward, Duga-2 first ran on 7 November 1971, and was successfully used to track missile launches from the far east and Pacific Ocean to the testing ground on Novaya Zemlya.
This was followed by their first operational system Duga-3, known in the west as Steel Yard, which first broadcast in 1976. Built outside Gomel, near Chernobyl, it was aimed northward and covered the continental United States. Its loud and repetitive pulses in the middle of the shortwave radio bands led to it being known as the "Russian Woodpecker" by amateur radio (ham) operators. The Soviets eventually shifted the frequencies they used, without admitting they were even the source, largely due to its interference with certain long-range air-to-ground communications used by commercial airliners. A second system was set up in Siberia, also covering the continental United States, as well as Alaska.
In early 2014, the Russians announced a new system, called Container, that was to see over 3000 km.
A more recent addition is the Jindalee Operational Radar Network developed by the Australian Department of Defence in 1998 and completed in 2000. It is operated by No. 1 Radar Surveillance Unit of the Royal Australian Air Force. Jindalee is a multistatic radar (multiple-receiver) system using OTH-B, allowing it to have both long range as well as anti-stealth capabilities. It has an official range of 3,000 kilometres (1,900 mi), but in 1997 the prototype was able to detect missile launches by China over 5,500 kilometres (3,400 mi) away.
Jindalee uses 560 kW as compared to the United States' OTH-B's 1 MW, yet offers far better range than the U.S. 1980s system, due to the considerably improved electronics and signal processing.
The French have developed an OTH radar called NOSTRADAMUS during the 1990s (NOSTRADAMUS stands for New Transhorizon Decametric System Applying Studio Methods (French: nouveau système transhorizon décamétrique appliquant les méthodes utilisées en studio).) It entered service for the French army in 2005, but is still in development. It is based on a star shaped antenna field, used for emission and reception (monostatic), and able to detect every aircraft at a range of more than one thousand kilometers, in a 360 degree arc. The frequency range used is from 6 to 30 MHz.
Launched in 2009, the French STRADIVARIUS research project developed a new over-the-horizon radar (High Frequency Surface Wave Radar – HFSWR) capable of monitoring maritime traffic up to 200 nautical miles offshore. A demonstration site  is operational since January 2015 on the French mediterranean coast to showcase the 24/24 7/7 capabilities of the system that is now offered for sales by DIGINEXT.
A number of OTH-B and OTH-SW radars are reportedly in operation in China. Few details are known of these systems. However, transmission from these radars causes much interference to other international licensed users.
Alternate OTH approaches
Another common application of over-horizon radar uses surface waves, also known as groundwaves. Groundwaves provide the method of propagation for medium-wave AM broadcasting below 1.6 MHz and other transmissions at lower frequencies. Groundwave propagation gives a rapidly decaying signal at increasing distances over ground and many such broadcast stations have limited range. However seawater, with its high conductivity, supports groundwaves to distances of 100 kilometres (62 mi) or more. This type of radar, surface-wave OTH, is used for surveillance, and operates most commonly between 4 and 20 MHz. Lower frequencies enjoy better propagation but poorer radar reflection from small targets, so there is usually an optimum frequency that depends on the type of target being detected.
An entirely different approach to over-the-horizon radar is to use creeping waves or electromagnetic surface waves at much lower frequencies. Creeping waves are the scattering into the rear of an object due to diffraction, which is the reason both ears can hear a sound on one side of the head, for instance, and was how early communication and broadcast radio was accomplished. In the radar role, the creeping waves in question are diffracting around the Earth, although processing the returned signal is quite difficult. Development of such systems became practical in the late 1980s due to the rapidly increasing processing power available. Such systems are known as OTH-SW, for Surface Wave.
The first OTH-SW system deployed appears to be a Soviet system positioned to watch traffic in the Sea of Japan. A newer system has recently been used for coastal surveillance in Canada, and is now offered for sales by Thales as the Coast Watcher. Australia has also deployed a High Frequency Surface Wave Radar.
- Laurie states two ranges for MADRE against aircraft, 3000 and 4000 km, on the same page. The former appears to be correct, comparing against other sources. To add to the confusion, Signals describes MADRE as having an average power of 100 kW and a peak of 5 MW, much more powerful than suggested by Laurie. See Signals, Vol 31, Issue 1, p. 7.
- Laurie 1974, p. 420.
- Frissell & Hockersmith 2008, p. 3.
- Laurie 1974, p. 421.
- Fowle, E.L. Key, R.I. Millar, and R.H. Sear, "The Enigma of the AN/FPS-95 OTH Radar", MITRE Corporation, 1979
- Laurie 1974, pp. 421-422.
- AN/FPS-118 Over-The-Horizon-Backscatter (OTH-B) Radar
- [dead link]
- "Final Environmental Assessment for Equipment Removal at Over-the-Horizon Backscatter Radar - West Coast Facilities"
- Photos of the TULELAKE AFS AN/FPS-118 OTH-B RADAR FACILITY
- "Electronic Weapons". Strategy Page. StrategyWorld.com. 2004-10-31. Retrieved 2006-11-21.
In 1997, the prototype JORN system demonstrated the ability to detect and monitor missile launches by Chinese off the cost of Taiwan, and to pass that information onto U.S. Navy commanders.
- Colegrove, Samuel B.(Bren) (2000). "Project Jindalee: From Bare Bones To Operational OTHR" (PDF). IEEE International Radar Conference - Proceedings. IEEE. pp. 825–830. Retrieved 2006-11-17.
- On Onera web, the French aerospace laboratory, one can find information about Nostradamus and a movie presentation on YouTube.
- "The STRADIVARIUS OTH Radar"., DIGINEXT
- John C. Wise, "PLA Air Defence Radars", Technical Report APA-TR-2009-0103, January 2009
- "Over-the-Horizon Backscatter Radar [OTH-B]"
- "Coast Watcher 200", Thales Canada
- Senator Robert Hill, Landmark Land Use Agreement For High Frequency Surface Radar, Ministerial Press Release number 33/2004 from the Australian Department of Defence, February 25, 2004
- Peter Laurie, "An eye on the enemy over the horizon", New Scientist, 7 November 1974, pp. 420–423.
- Nathaniel Frissell and Lyndell Hockersmith, "Seeing Beyond: Over the Horizon Radar Systems and HF Propagation", Virginia Tech, 2 December 2008
- Radar Systems on Shortwave, an extensive list of OTH and similar radar systems compiled by Wolf Hadel, August 2013
- A Canadian Perspective on High-Frequency Over-the-Horizon Radar — paper by R. J. Riddolls, Defence R&D Canada, Ottawa
- The Development of Over-the-Horizon Radar in Australia — paper by D.H. Sinnott on the Australian Department of Defence Web site
- Google maps link — Russian "Steel-Yard" radar near Chernobyl.