Lockheed WS-117L
(CORONA / SAMOS / MIDAS)

This article presents the development history of the United States' first generation of reconnaissance satellites. It covers the CORONA, SAMOS and MIDAS systems, which were all either direct continuations or spin-offs of the U.S. Air Force's WS-117L program. CORONA was run by the CIA and therefore not strictly a military program, but for the sake of completeness it is nevertheless included here. Also, it was by far the most important and successful of the early satellite programs.

WS-117L

The history of the USA's first military satellite system dates back to the 1951/52 time frame, when the RAND "think tank" conducted studies under an Air Force contract about the possibility of satellite-based reconnaissance. As a result of these studies, ARDC (Air Research and Development Command) established a project framework in 1953 for actual development of a military satellite system. The project was briefly designated MX-2226, but was eventually given the Weapons System designator WS-117L in December 1953. In March 1955, the Air Force issued General Operational Requirement (GOR) No. 80 for a Reconnaissance Satellite System. To fulfill the requirement, WS-117L was to lead to a multi-purpose satellite design, which could carry a variety of reconnaissance and intelligence payloads.

By the end of 1955, design study contracts for WS-117L had been awarded to RCA, Martin and Lockheed under the name "Pied Piper". In June 1956, Lockheed was selected as prime contractor for the WS-117L system. The satellite's core, developed to first flight in the following three years, became known as RM-81 Agena. Agena was a general purpose flight module, which served as upper stage of the launch vehicle and also as in-orbit control module, and which could house custom-designed payloads in its nose section.

From the start of WS-117L, the Air Force had planned to use the forthcoming SM-65 Atlas ICBM as first-stage launch vehicle. For the photo-reconnaissance mission, the USAF envisioned a film readout and transmission system, where the film would be developed and optically scanned on the satellite, and the scan converted to electrical signals, which would then be transmitted to ground stations and processed to regenerate the image. However, this approach turned out to be problematic. The film readout procedure was unproven, and there were serious doubts whether the technology of the time was good enough to produce viable results. Also, the development of the Atlas missile took longer than expected. On the other hand the intelligence community, most notably the CIA, wanted information about the Soviet ICBM programs as soon as possible by any means. Therefore, an "interim" WS-117L system was proposed in November 1957. It was to be lighter than the ultimate WS-117L, so that it could be launched by the SM-75 Thor IRBM, which was expected to be a lower risk than Atlas. And instead of the film readout method, it would return the undeveloped film to earth using a recoverable reentry vehicle.

In February 1958, the "interim" WS-117L was officially cancelled. But at the same time, a new secret program to develop such a system was established. WS-117L was an Air Force program, and the USAF was discussing it openly as a military satellite system. The new program, on the other hand, was to be primarily managed by the CIA, and its purpose as a photo-reconnaissance satellite was to be kept highly secret. The secret code name CORONA was assigned to the program, but to the outside, it was named Discoverer and described as a general-purpose satellite program for scientific, technological and biomedical research in space. It was not before 1992, that CORONA was officially declassified.

The beginning of the CORONA program didn't mean the end of the USAF's WS-117L. But the Weapons System designator was dropped in October 1958, primarily for political reasons, and replaced by a simple name - first SENTRY, and finally SAMOS. Another direct spin-off of WS-117L was an early warning system named MIDAS.

CORONA

The first camera system for CORONA was developed by Itek and built by Fairchild, and designated simply as "C Camera" (the C presumably standing for CORONA). The effective surface resolution of the photos was around 12 m (40 ft), and the first flights carried enough film for a mission completing 17 orbits in about 24 hours. CORONA satellites were launched from Vandenberg AFB in California into polar orbits, needed to be able to cover the full area of the Soviet Union. The camera was only active while the satellite was over relevant "denied territory". The exposed film was stored in a reentry capsule, built by General Electric and known as SRV (Satellite Return Vehicle or Satellite Recovery Vehicle) or informally as "bucket". At the end of the mission, a ground control signal initiated the separation of the SRV, which was equipped with small retro-rockets (replaced by cold gas thrusters very early in the program) to begin reentry. The SRV was protected by an ablative heat shield, the remains of which were jettisoned at about 18000 m (60000 ft) altitude, when the parachute was deployed. The primary method to recover the SRV was to use specially equipped transport aircraft to snatch the parachute in mid-air and reel in the capsule. In the first years, these aircraft were C-119F Flying Boxcars, later replaced by the JC-130B Hercules. SRV reentry was always timed so that the capsule descended over the ocean. Should the mid-air recovery fail, the SRV could float in the water for about 3 days, and signal its location to a recovery team via a radio beacon.

The first launch of a CORONA satellite, named Discoverer 1 for the public, occurred on 28 February 1959, but failed to reach orbit due to a failure of the Agena upper stage. The first orbitial flight was Discoverer 2 on 13 April 1959, but the recovery of the SRV failed because of a mis-timed separation. After three launches for test and evaluation, Discoverer 4 on 25 June 1959 was the first to carry a camera system, but it failed to reach orbit. None of the first 12 Discoverer/CORONA flights came close to a full success, due to lots of different failure scenarios. These included exploding Thor boosters, malfunctioning Agena modules, failing camera and film, and SRV reentry/recovery failures. Discoverer 13, a pure research/diagnostics flight without a camera system, was the first one ending with the recovery of the SRV. And Discoverer 14, launched on 18 August 1960, finally succeeded in returning the first high-resolution reconnaissance imagery taken from space back to earth.

Drawings: Guiseppe di Chiara, CC BY-SA 3.0

CORONA with C camera (KH-1) on Agena A

Not all problems were solved for good with the success of flight 14, and the next three missions were again failures. But in general, CORONA had proven that it worked in principle, and would be able to provide very valuable intelligence. In the meantime, the camera system had also been improved, and flight 16 saw the first use of the updated C' (C Prime) system, with a ground resolution of about 8 m (25 ft). Flight 17 introduced the Agena B module, which was an enlarged Agena A with an uprated engine and a significantly higher propellant load. Agena B could launch heavier payloads into orbit, and offered more options for orbital insertion.

Drawings: Guiseppe di Chiara, CC BY-SA 3.0

CORONA with C' camera (KH-2) on Agena B

In 1961, two new CORONA payloads were introduced. The first was an Army-developed mapping camera system named ARGON. Other than the C/C' cameras, which photographed narrow strips on the surface with a high resolution, ARGON would photograph wide areas with a much lower resolution. This was needed to get accurate geo-coordinates of potential ICBM targets in the whole Soviet Union. The first ARGON flight was CORONA flight 20 on 17 February 1961, but it was unsuccessful - the camera system failed, and the SRV was not recovered. Only the 5th ARGON mission was the first successful one, flown in May 1962.

Drawings: Guiseppe di Chiara, CC BY-SA 3.0

CORONA with ARGON camera (KH-5) on Agena B

For high-resolution CORONA missions, the new C''' (C Triple Prime) camera system became operational in 1961. It was both designed and built by Itek, and had several design improvements leading to a surface resolution of about 3.6 m (12 ft) and a higher overall reliability of the system. The first C''' mission was CORONA flight 29, launched on 30 August 1961. The SRV was successfully recovered, but because of a camera system failure, all photos were out of focus. The cause was fixed until the next C''' flight (CORONA 32) on 13 October 1961, which was a complete success.

Drawings: Guiseppe di Chiara, CC BY-SA 3.0

CORONA with C''' camera (KH-3) on Agena B

In 1961, the development of a camera system for stereo imagery began under the name MURAL (or simply M). It consisted of two C''' cameras, the rear one angled 15° forward, and the front one angled 15° back. Each one had its own supply of film, and the area on the ground was photographed twice - first by the forward-pointing camera, and a few frames later by the backward-pointing one. The result were two photos which, when viewed with special stereoscopic devices, gave the photo interpreters a 3-dimensional view of the ground. Even though the M camera had basically the same optical resolution as the C''', the stereo images made it significantly easier to identify structures and equipment. The MURAL system first flew on CORONA flight 38 on 27 February 1962, a mission ending in successful recovery of the SRV.

Drawings: Guiseppe di Chiara, CC BY-SA 3.0

CORONA with MURAL camera (KH-4) on Agena D

Coincidentally, the first MURAL camera mission was also the last one using the Discoverer cover name. From March 1962 on, military satellites were no longer publicly associated with any specific project, name or purpose. Space programs were to be identified by numerical program designators only, with CORONA becoming Program 162. At the same time, the different camera payloads of CORONA satellites were given numbers in a new KH series ("KH" was derived from "Keyhole"). The new nomenclature was also retroactively applied to the initial C, C' and C''' systems, which were no longer flying at that time.

In 1962, a short-term requirement arose for very high-resolution photos of suspected Soviet ABM (Anti-Ballistic Missile) sites. A camera system labeled E-5, developed in the context of the Air Force's SAMOS program, was supposed to provide a resolution of 0.6 m (2 ft). SAMOS itself was in trouble, and already on the verge of cancellation, but it was decided to adapt the E-5 camera for CORONA. As such it was named LANYARD, and designated KH-6. LANYARD was larger and heavier than the other CORONA payloads, which mandated a more powerful launch system. The Thor was equipped with three strap-on solid-fueled rocket boosters, creating the Thrust-Augmented Thor (TAT) first stage. Another novelty of LANYARD was a so called "roll-joint" between the Agena and the camera module. This meant that the camera could rotate, while the Agena maintained a stable attitude, which would improve overall camera stability and therefore image quality. The first LANYARD launch, on 18 March 1963, failed to reach orbit, and on the next two flights in May and July, the KH-6 camera failed so that no usable photos were returned. By that time, another highly secret program for a high-resolution photo-reconnaissance satellite, run by the NRO with the Air Force and named KH-7 GAMBIT, was showing promising results, and therefore LANYARD was terminated after only three flights.

Drawings: Guiseppe di Chiara, CC BY-SA 3.0

CORONA with KH-6 camera on Agena D

The next significant payload upgrade for CORONA was named JANUS (J; later renamed J-1, when the J-3 was developed), and designated KH-4A. It consisted of a slightly upgraded KH-4 camera system (resolution improved to about 2.7 m (9 ft)), with twice the amount of film supply. KH-4A satellites had two SRVs, so that the film was returned in two batches. The Agena D module could also be set dormant for up to three weeks after the first SRV was deorbited, allowing for significantly more versatile mission profiles. The first two KH-4A flights in August and September 1963 had issues which prevented the recovery of the second SRV. Nevertheless, from 1964 onwards KH-4A was used for all CORONA missions until August 1967,

Drawings: Guiseppe di Chiara, CC BY-SA 3.0

CORONA with KH-4A camera on Agena D

The final upgrade of the CORONA camera system was known as KH-4B or J-3, developed since 1965. Even though the designation suggested only a gradual upgrade from the KH-4A J-1, the system was a major redesign. The result was higher reliability and better overall image quality. The first KH-4B flew in September 1967, and was a full success. Shortly before, an upgraded Thor rocket with a longer fuel tank was introduced, known as Long-Tank Thrust-Augmented Thor (LTTAT) or Thorad. All KH-4B satellites were launched with the LTTAT.

Drawings: Guiseppe di Chiara, CC BY-SA 3.0

CORONA with KH-4B camera on Agena D

During the course of the CORONA program, its numerical designator was changed twice: From Program 162 to Program 241 in July 1964, and finally to Program 846 in November 1966.

CORONA ended after 145 missions in May 1972. By that time, the significantly larger and more capable HEXAGON satellite (KH-9 camera system) had successfully completed its first two missions, and was to become the primary US wide-area photo-reconnaissance system for the upcoming years.

CORONA Launch List

• No.: Sequential flight number for the CORONA program
• Name: Unclassified name of the satellite. The OPS number ("OPS" standing for "Operations") is a random number given to all military-related satellite launches between 1963 and 1983 (when "OPS" was replaced by a sequential "USA" number). Between the end of the Discoverer series and the beginning of the OPS numbering system, the only unclassified reference to the satellite was an FTV (Flight Test Vehicle) number. For CORONA, the FTV number was identical to the Agena serial number.
• Mission: Each CORONA mission had a unique serial number, typically assigned in a separate block for each camera system. Flights without a camera system (for testing and/or support of the "scientific" cover story) didn't get a Mission number.
• COSPAR ID: International designation of the satellite; not applicable, if orbit was not reached
• Launch Vehicle: Variant of Thor first stage and Agena upper stage
• Result:
  - Launch failure: Failed to reach orbit; includes both first stage and second stage failures
  - Camera failure: Agena reached orbit, but the camera system failed in some way
  - Deorbit failure: The camera possibly worked in orbit, but the re-entry of the SRV wasn't properly initiated
  - Recovery failure: Mission succeeded up to and including SRV reentry, but SRV couldn't be recovered
  - Partial success: SRV recovered, but only part of the film was usable; or only one out of two SRVs of a KH-4A/B was successfully recovered
  - Success: SRV(s) successfully recovered and film retrieved

CORONA launches

SAMOS

SAMOS was effectively WS-117L renamed, and as such the main focus for delivering the reconnaissance photos was still on a film readout system. The advantages seemed to be obvious, like much faster access to the photos, and potentially longer mission times not limited by the volume of a film recovery capsule.

WS-117L/SAMOS major components were designated by letters of the alphabet, followed by a sequential number, if needed:

A - Airframe
B - Propulsion
C - Auxiliary Power
D - Guidance and Control
E - Visual Reconnaissance
F - Electronic Intelligence
G - Early Warning (see MIDAS)
I - Intelligence Data Processing

The initial SAMOS film readout camera system was thus named E-1, and designed to provide pictures with a ground resolution of 30 m (100 ft) for an area of 160 x 160 km² (100 x 100 miles²). But from the beginning it became clear that the process of scanning the negatives in orbit, transforming the brightness information into electrical signals and transmitting these to a ground station in the USA was rather slow, due to the limitations of then state-of-the-art techology. Not more than around 60 photo frames per day could be received by each of the three planned receiving stations. Also, the resolution would be significantly lower than that of the original film. Nevertheless, even a second readout system was designed, named E-2, which photographed a smaller area (27 x 27 km²) at a somewhat higher resolution (6 m (20 ft)). The still-born E-3 was an electrostatic tape readout system.

Because of the anticipated problems, film recovery systems were studied in parallel. E-4 was to be a mapping system, which was abandoned because it would have duplicated what eventually became CORONA's KH-5 ARGON camera. The E-5 camera was therefore the first recovery system built for SAMOS. Initially, it was planned to bring the SAMOS E-5 SRV down on land to avoid the need for a mid-air recovery. This had several other disadvantages though, and when CORONA showed its first successes with mid-air recoveries, that method was chosen for SAMOS as well. To initiate reentry, the whole Agena stage was deorbited, and the SRV separated afterwards.

The first SAMOS launch, with an E-1 system as primary payload, occurred on 11 October 1960, but failed to reach orbit. The second launch on 31 January 1961 succeeded to place the satellite in a sun-synchronous orbit, and a few days later the first transmitted images were available for review. However, the image quality was not impressive (especially when compared to photos of the first successful CORONA flights), and at that point it was effectively decided to terminate the readout SAMOS program as soon as the first recovery system (E-5) became available. In the end, only one E-2 flight attempt followed in September 1961, which ended in a launch failure. The same happened to the first E-5 launch in November that year. After two more failed E-5 missions, that part of SAMOS was terminated as well in March 1962. But the E-5 system itself was briefly revived as CORONA's KH-6 LANYARD camera, also with very limited success.

By that time, yet another camera system for SAMOS had begun development. Labeled E-6, it was to be an advanced very high resolution film recovery system, with a ground resolution of 2.4 m (8 ft). It pioneered some technology which was to be used in a completely new photo-reconnaissance system named GAMBIT/KH-7, developed by the Air Force and the then newly established National Reconnaissance Office (NRO). GAMBIT was highly secret, and the SAMOS E-6 program was a convenient cover for GAMBIT's initial camera development.

Drawings: Guiseppe di Chiara, CC BY-SA 3.0

SAMOS configurations

In March 1962, the same directive which assigned the nomenclature Program 162 to Discoverer/CORONA, also replaced the SAMOS label by program numbers:

Between April and November 1962, five SAMOS E-6 missions were launched. While all reached orbit, not a single SRV was successfully recovered. This disappointing track record, coupled with the success of CORONA and some promising progress on the aforementioned GAMBIT, led to the cancellation of the SAMOS photo-reconnaissance program on 31 January 1963. At some time during its short existence, the unclassified numerical designation of SAMOS E-6 was changed to Program 698BJ.

The SAMOS flights with E-1 and E-2 systems also carried a secondary payload named F-1. F-1 was an experimental ELINT (Electronic Intelligence) system also known as a "Ferret". It was the precursor of later satellites, initially named SAMOS F-2, which flew similar ELINT payloads "stand-alone" on Thor-launched Agena-B/D modules. These SAMOS follow-ons are described on the SAMOS ELINT page.

SAMOS Launch List

• No.: Sequential flight number for the SAMOS photo-reconnaissance program
• Name: After SAMOS 2, contemporary statements either didn't quote any name for the satellite or mission, or only an FTV (Flight Test Vehicle) number. For SAMOS, the FTV number was identical to the Agena serial number.
• COSPAR ID: International designation of the satellite; not applicable, if orbit was not reached
• Launch Vehicle: Variant of Atlas first stage and Agena upper stage
• Result:
  - Launch failure: Failed to reach orbit; includes both first stage and second stage failures
  - Camera failure: Agena reached orbit, but the camera system failed
  - Recovery failure: Mission succeeded in orbit, but either the deorbiting failed or the SRV couldn't be recovered
  - Success: Images successfully transmitted to ground stations, or SRV recovered (which never happened)

SAMOS launches

MIDAS

Early in the WS-117L program, it was proposed to develop a payload for the detection of ballistic missile launches in the Soviet Union. It was to supplement the land-based Ballistic Missile Early Warning System (BMEWS). By the end of 1957, the early warning payload had become subsystem G of WS-117L, and in November 1958, it officially became a separate program named MIDAS (Missile Defense Alarm System). Like WS-117L/SAMOS, MIDAS was based on the Agena upper stage and on-orbit module, and launched by an Atlas first stage. The payload consisted of a low-resolution infrared camera, which was to detect the thermal signature of a missile launch. When a launch was registered, its time and approximate location were radio-transmitted to a ground control station.

The first two MIDAS missions, labeled Series I and launched in 1960, planned to use relatively low orbits of around 500 km altitude, limited by the performance of the Atlas-Agena A launcher. MIDAS-1 didn't reach orbit, and on the second flight, the radio link failed before any experiments with detecting prepared IR sources on the ground could begin.

The satellites for the next three missions, MIDAS Series II, were slightly lighter than the first two and used the more powerful Agena B upper stage, allowing an orbital altitude of around 3500 km (significantly increasing the satellite's field of view). Electrical power was supplied by two solar panels, replacing the heavy batteries used in the first two MIDAS satellites. However, the test series was not very successful. On MIDAS 3 the power supply failed after only 5 orbits, and MIDAS 4 suffered from an imprecise ascent trajectory, a tumbling Agena in orbit, and the disappointing performance of its IR sensor payload. Compared to that, the final Series II launch in April 1962 was moderately successful, yielding IR background measurements until the power failed on the 7th orbit.

Similar to other military space programs, the MIDAS name was dropped in March 1962 and replaced by a numerical designator. In this scheme, MIDAS became Program 239A, changed to Program 461 in March 1963.

In December 1962, MIDAS 6 was launched, the first of the Series III satellites. They featured upgrades supposed to increase overall system reliability, and also had a completely new infrared sensor payload. But MIDAS 6 didn't even reach orbit, because the Atlas rocket failed shortly after launch. MIDAS 7, launched on 9 May 1963, was the first truely successful mission of the program. It operated for 47 days, and successfully detected several test launches of American missiles, which were specifically timed to fall within MIDAS 7's field of view. The final two MIDAS flights resulted in another launch failure for MIDAS 8 and a final successful mission for MIDAS 9.

Photo: USAF

MIDAS 6 infrared camera system

The original MIDAS program had ended after these 9 missions, having demonstrated the feasibility of the basic concept. But Program 461 continued for a while, and in 1966, three significantly improved satellites were also flown under that program. They are usually also listed under the MIDAS label. Also known as RTS-1 (Research Test Series 1) missions, they scored one partial launch failure (leading to an undesirable orbit) and two successful flights. A planned RTS-2 series, under Program 266, was not proceeded with. By November 1966, the Air Force had replaced Program 266 by Program 949, which would eventually lead to the very successful IMEWS/DSP (Integrated Missile Early Warning System / Defense Satellite Program) satellites.

MIDAS Launch List

MIDAS launches

Main Sources

[1] Curtis Peebles: "The CORONA Project", Naval Institute Press, 1997
[2] David Baker: "US Spy Satellites", Haynes Publishing, 2016
[3] Jeffrey T. Richelson: "America's Space Sentinels: DSP Satellites and National Security", University Press of Kansas, 1999

Back to Directory of U.S. Military Rockets and Missiles, Appendix 3

---

Last Updated: 16 June 2025