Cobra Dane (AN/FPS-108) is a large and powerful phased-array radar located on Shemya Island at the western end of the Aleutian Island chain (52.7º N, 174.1º E). The radar’s boresite is at 319º (that is 41º west from due north) at 20º above the horizon. Its primary mission when deployed was the monitoring of Soviet ballistic missile test flights, with secondary missions of early warning and space surveillance. It became operational in 1977, and underwent a modernization in the early 1990s and a number of enhancements subsequently.
Cobra Dane was part of the Space Surveillance Network until April 1994, when this role was eliminated (and its communications link to the SSN’s Space Communications Center was closed) for budgetary reasons. Even before then, operational procedures – in particular the limitation of its surveillance fences to low elevations — limited the size of space objects it could detect. Following tests in 1999 that demonstrated its capability for tracking small space debris, it was reconnected to the SSN in October 1999. However, in order to reduce operating costs, it was only operated at one–quarter power. This was achieved by reducing its duty cycle from 6.0% to 1.5%, with the radar capable of returning to full power in less than 30 seconds if a relevant ballistic missile test took place. Cobra Dane resumed full power operations in March 2003. It is now a contributing sensor in the SSN. Cobra Dane has also been incorporated into the United States Ground-Based Midcourse Defense (GMD) national missile defense system.
Cobra Dane’s antenna is 29 m in diameter and has 15,360 active elements out of a total of 34,768 elements. The active elements are space-tapered, decreasing to 20% density at the antenna edge. In order to obtain high range resolution off-boresite the antenna elements are subdivided into 96 subarrays, each powered by a separate traveling wave tube. The radar’s peak power is 15.4 MW with an average power of 920 kW, corresponding to its maximum 6% duty cycle. Its beam width is 0.6˚.
As originally built, in wide-band mode Cobra Dane used a 1 ms pulse with a 200 MHz bandwidth (obtained using linear frequency modulation pulse compression) at frequencies between 1.175 and 1.375 GHz, and was limited to angles within 22.5˚ of its boresite. In wideband operation, it has a range resolution of about 3.75 feet (1.14 m).
In narrowband operation, it uses frequencies between 1.215 to 1.250 GHz (corresponding to a wavelength of about 24.3 cm). In search, it uses 1 MHz pulses with lengths of 1.5 or 2.0 ms, and in track it uses six different 5 MHZ bandwidth pulses with lengths between 0.15 and 1.5 ms. The radar also has two 1.0 ms, 25 MHz bandwidth pulses used for ionospheric compensation.
Cobra Danes specification’s call for it to achieve a range accuracy of 15 feet (4.6 m) and an angle accuracy of 0.05˚ at 0.6˚ elevation. Typical accuracies are 3 m in range and 0.02˚ in angle.
Because it operates at a higher frequency than the SSN’s other large phased-array radars, which all operate at about 440 MHz, a given small object will generally have a larger RCS for Cobra Dane for object sizes of about 12 cm and smaller. This RCS advantage increases to a factor of greater than 60 for spherical objects 5 cm or smaller.
When it was deployed, Cobra Dane had an azimuthal field of view of ±60º. Its field of view was subsequently extended by ±8º at elevations up to 30º. However, its northerly location and its orientation limit it to tracking space objects with inclinations between 55 and 125 degrees.
Starting in 1999, Cobra Dane was reconnected to the space surveillance network and was able to use radar time not devoted to its primary missile surveillance mission specifically for space surveillance. At this time, Cobra Dane was still operating at one-quarter power and operated primarily in a tasked mode. Operating in this mode, Cobra Dane collected about 2500 observations on about 500 cataloged objects per day. It also maintained a ten-degree-wide, high-elevation fence for detecting uncataloged objects which produced about 500 observations per day on about 100 uncorrelated objects. Together, these two tasks occupied a total duty factor of 1.14%.
In a 1999 test, a single row of pulses fence of width 30˚ at a 50˚ elevation using the then full space surveillance allotment of 1.14% duty factor produced between 700 and 800 uncorrelated tracks per day. As part of a space debris experiment later in 1999, Cobra Dane maintained a 60˚ wide (289˚- 349˚ azimuth) single row of pulses fence at an elevation of 50˚, and at ranges of 417-2501 km, with the radar running at full power and with 3.0% duty factor used for the fence.
Its return to full power operation in 2003 allowed Cobra Dane to maintain a wide space debris search fence in addition to continuing its primary missile intelligence gathering role, and resulted in the rapid addition of several thousand objects to the SSN’s analyst catalog (many of which were subsequently dropped). This fence is likely the same as or similar to the 60˚ wide, 50˚ elevation fence described in the previous paragraph.
As of about 2000, Cobra Dane was able to detect objects out to ranges of 14,000 km, but software in use at that time was designed to drop tracks on objects with orbital periods greater than 225 minutes (that is, objects not in low earth orbit, by the SSN’s definition).
Cobra Dane Performance Claims
As of 2000, Cobra Dane could achieve a S/N = 15 dB against a -20 dBsm target at a range of 1,852 km with a single 1.5 ms pulse and 13.2 db with a 1 ms pulse. However, this was at an elevation of 1.0˚, which results in atmospheric absorption and scan losses of about 2.1 dB = 1.6 relative to a beam along the boresite. Taking this absorption into account, the corresponding boresight sensitivities would be 59,500 = 47.7 dB against a 1 m2 target at 1,000 km with a 1.5 ms pulse, and 39,400 (46.0 dB) with a 1.0 ms pulse. An earlier source (1976) states that it can achieve 16.5 dB against a 0.01 m2 target at 1,000 nautical miles with a 1.0 ms pulse length (60 pulses per second) at 0.6˚ elevation, with greater sensitivity obtainable with a longer pulse. This corresponds to a boresite sensitivity of 82,500.
 Unless otherwise noted, the information in this post is from P. Chorman, “COBRA DANE Space Surveillance Capabilities,” in S. E. Andrews, ed., Proceedings of the 2000 Space Control Conference, Lincoln Laboratory, Lexington, Mass., April 11-13, 2000, pp. 159-168; E.G. Stansbery, “Growth in the Number of SSN Tracked Orbital Objects,” 55th International Aeronautical Congress of the International Aeronautical Federation, the International Academy of Astronautics, and the International Institute of Space Law, Vancouver, Canada, October 4-8, 2004 (IAC-04-IAA.5.12.1.03), Earl Filer and John Hartt, “Cobra Dane Wideband Pulse Compression System,” IEEE Electronic and Aerospace Systems Convention (EASCON) 1976, Washington, D.C., September 26-29, 1976, Paper 61., and Philip J. Klass, “USAF Tracking Radar Details Disclosed,” Aviation Week and Space Technology, October 25, 1976, pp. 41, 43, 46.
 Gene Stansbery, “Preliminary Results from the U.S. Participation in the 2000 Beam Park Experiment,” Proceedings of the 3rd European Conference on Space Debris, Darmstadt, Germany, March 19-21, 2001, pp. 49-52.
 One pulse is centered 1187.5 MHz and the other at 1363.5 MHz. Using satellite targets of opportunity, the ionospheric correction is then made based on the time delay difference between the two frequencies.
 Keith Englander, “Ground Based Midcourse Missile Defense, briefing slides, U.S. Ballistic Missile Defense Agency, August 2001.
 In the Rayleigh scattering regime, which Cobra Dane will be in for objects about 5 cm diameter and smaller, the RCS varies with the 4th power of the frequency, so compared to the FPS-85 and other SSN LPARs, Cobra Dane has an advantage of about (1.24/0.44)4 = 63.
 Chorman, ‘COBRA DANE Space Surveillance Capabilities,” p. 161.
 Stansbery, “Growth in the Number,” pp. 2-4.
 Chorman, “COBRA DANE Space Surveillance Capabilities,” p. 160.
 Nicholas L. Johnson, “U.S. Space Surveillance,” Advances in Space Research, pp. 8(5)-8(20), (p. (8)9).
 Chorman, “COBRA DANE Space Surveillance Capabilities,” p. 160.
 Atmospheric absorption loss from Fred E. Nathanson, Radar Design Principles: Signal Processing and the Environment, 2nd ed. (Mendham, NJ: Scitech, 1999), Figure 6.1 (p. 216). Scan loss from Figure 5 of Chorman, “COBRA DANE Space Surveillance Capabilities,” p.166.
 Filer and Hartt, “Cobra Dane Wideband Pulse,” p. 61-F.
 Stansbery, “Growth in the Number,” p. 3.
 Chorman, “COBRA DANE Space Surveillance Capabilities,” p. 159.