by Dwayne Day
The USIB, for instance, might establish a requirement for more information on Soviet nuclear facilities. While other members of the Intelligence Community would seek information through other channels (e.g., clandestine agent reports and signals intelligence intercepts), COMOR was responsible for seeing that relevant imagery was collected. COMOR would identify specific nuclear facilities to be imaged; this list would then go to an operations center (first at the CIA and later at the National Reconnaissance Office [NRO], where it was eventually known as the “Mushroom Factory”), which was responsible for outlining the basic operational parameters of a mission.50 This small group would determine launch and recovery times, orbital paths, when to operate the cameras, and other details. This outline would then go to engineers at the Lockheed AP facility, where it was converted into specific commands for the spacecraft. CORONA’s commands were stored on 35mm Mylar tape with holes cut into it at specific points. Five wire brushes ran along one side of the tape. When a brush encountered a hole, it would touch a metal contact on the other side of the tape, closing a circuit. With several brushes, a large number of circuit combinations were possible, and thus a large number of commands could be given to the spacecraft. The system proved quite reliable, but was also limited in flexibility at first, since all commands for the spacecraft were preset before launch.
Getting the spacecraft in orbit and photographing certain targets was important, but the photographs still had to be evaluated once they were returned to Earth. During the first U-2 flights, the CIA operated the Photo-Intelligence Division (PID). By 1958, the Army and Navy joined with the CIA (the Air Force wanted to conduct its own evaluations) and formed the Photographic Intelligence Center (PIC). In fall 1960, a major imagery review group that had evaluated the Intelligence Community’s ability to conduct photo-interpretation recommended to President Eisenhower that the Air Force efforts be incorporated into the PIC. Eisenhower agreed with the recommendations. The National Photographic Interpretation Center (NPIC) was created in January 1961, with Arthur Lundahl as the first director.51
On August 18, 1960, the same day that the first film was successfully retrieved from space, James Q. Reber, chairman of COMOR, issued the “List of Highest Priority Targets: USSR.” This list included ICBM and IRBM bases and production complexes, submarine-launched ballistic missile and heavy bomber bases and manufacturing sites, and nuclear energy facilities—a total of thirty-two in all, complete with geographic coordinates. Most of the facilities were believed to be along major rail lines—which was one of the ways that U-2 targets were selected. Included in the list were the cities of Kirov, Danilov, Konosha, Murmansk, Kyshtym, Sverdlovsk, Mogilev, Leningrad, Gorkiy, and Sevastapol, to name a few.52
On this same day, a staffer to the Joint Chiefs, Colonel James E. Mahon, issued a memorandum calling for a review of scheduling of future CORONA launches with the idea of pushing the program into operational status as soon as possible. Mahon noted that the needs for reconnaissance and geodesy (the location of geographical points on the earth) information were both critical. He suggested that COMOR recommend that a CORONA mission using the newer C′ (“C Prime”) camera system be launched as soon as possible to obtain reconnaissance of the whole of the Soviet Union, particularly focusing on targets set forth in a revised May 25 “List of Highest Priority Targets.” Mahon stated that as soon as that mission was accomplished, a mapping mission using the CORONA spacecraft but a different camera, known as ARGON, should be launched to fulfill the geodesy requirements. In other words, CORONA would identify the targets and ARGON would pinpoint their location.53
The need for accurate geodetic information was very important from a strategic warfighting point of view. A bomber crew could be expected to search for its target and then attack it. But ICBMs have an inherent inaccuracy during their flight that is measured in terms of circular error probable (CEP), which, simply stated, indicates how near to their target half of the missiles fired will fall. If the exact location of the target is unknown, errors compound upon each other and there is less assurance that an ICBM will actually hit its target. ARGON was needed to remove a large part of this uncertainty by photographing vast swaths of land so that accurate geodetic measurements could be taken.
ARGON had been approved as an Army-sponsored independent mapping project a year earlier, on July 21, 1959.54 It replaced an Air Force SAMOS reconnaissance camera system known as E-4. ARGON was handled under the auspices of the CORONA program out of the need for security and a concern that it might compete for launches. ARGON employed a 3-inch focal-length still camera with 5-inch-wide film and a resolution of approximately 460 feet. Its images covered an area of 300 by 300 nautical miles.55
BUILDING BETTER SPACECRAFT
New versions of CORONA were also in development. Discoverer XVI, CORONA Mission 9011, was launched in October 1960 with the first C′ camera, but failed to reach orbit. The C′ was also designed and manufactured by Fairchild, under Itek supervision. The C′ differed from the C by having variable image motion compensation so that different orbits could be flown.56 Discoverer XVII was launched a month later but suffered a payload malfunction. However, Discoverer XVIII, CORONA Mission 9013, launched on December 7, 1960, worked perfectly.
Discoverer XVI was also the first launch using the Agena B second-stage booster. The new booster failed during this launch but was successful during the following launch. The Agena B represented a significant improvement over the A version. Not only did it have bigger fuel tanks—effectively doubling its fuel capacity (and hence increasing its payload and safety margins) over early Agena versions—but it also possessed a restart capability.57 Restarting a space vehicle in orbit represented a difficult challenge, but the only significant additional equipment required for the Agena, other than the increased fuel load, was a second starter charge and additional ullage rockets that fired to push the fuel to the rear of the tanks so that it could feed into the engines.58
On February 17, 1961, the first ARGON satellite was launched, with the designation CORONA Mission 9014A (the A stood for ARGON). But the orbital programmer and the camera failed and there was no recovery of film from this mission. The next CORONA mission after that also failed. The next two ARGONs after that failed. The following CORONA was a success. The one following that was only a partial success after the camera stopped midway through the flight. Then there was another ARGON failure (due to a rare Thor malfunction), followed by another CORONA malfunction.
Imagery of a Soviet airfield returned from CORONA Mission 9017 on June 16, 1961. This satellite used a C′ (KH-2) camera, had a maximum resolution of about 25 feet, and proved to be particularly successful; it provided the details and locations of several key Soviet strategic facilities. (Photo courtesy of the National Reconnaissance Office)
On August 30, 1961, the first C‴ (“C Triple Prime”) camera was launched. (There was no C″ version.) The C‴ swas the first Petzval f/3.5 lens camera, the first camera both designed and built by Itek, and the first camera to include significant improvements to reduce vibration. The C‴ differed from its predecessors by having the lower section of the camera containing the heavy lenses rotate a full 360 degrees while the scan arm at top rocked (or oscillated) back and forth through a 70-degree arc.
Robert Hopkins of the University of Rochester had suggested using the Petzval f/3.5 lens on CORONA. He realized that the field curvature in a Petzval lens was not as important when it was being used in a panoramic camera. The Itek-designed Tessar f/5.0 lens used on the earlier C and C′ cameras did not have sufficient light-gathering power to allow use of a slower film that could record finer detail. By switching to the Petzval lens, the designers could also switch to a slower film. Typical film speeds for household cameras are in the ASA 200–ASA 400 range. CORONA ultimately used films with speeds of ASA 2 to ASA 8, depending on the processing.59 Inserting a field flattener in the focal plane created diffraction-limited resolution, allowing the camera to achieve 180–200 lines per millimeter—mu
ch better than the C and C′ cameras at 100–120 lines per millimeter. For the earlier cameras, enlargements could be made up to 20 times the original image size. The new camera allowed enlargements up to 40 times the original image size.60
The Petzval design was large. It used three 7-inch diameter elements at the front and two 5.5-inch diameter elements at the rear. These were all encased in a 22-inch-long cylindrical cell. The lens rotated around an axis about 10 inches from the front of the cell. In addition to the five lenses in the cell, a sixth lens segment, acting as the field-flattener, was located .25 inches in front of the film. Collectively, the six lenses weighed about 20 pounds.61
The C‴ camera was also modified to reduce the effect of thermal differentials on its components. Its controls were made more reliable, and the method of metering film and achieving and maintaining camera focus were also improved. Adjusting the IMC depending upon the orbit allowed for lower orbits and higher resolution. Furthermore, timing pulses, which determined the scan velocities and IMC for each frame, were marked in the image area rather than the border area of the film. Resolution improved from 25 feet for the C′ to 12–25 feet for the C‴. Finally, 39 pounds of film could now be carried.62
This mission, number 9023, was a success. Of the sixteen CORONA and ARGON missions launched in 1961, seven CORONAs returned film (all of the ARGON missions were failures). Gradually, problems were being solved. Also in 1960 and 1961, two CORONA missions had been launched without cameras or SRVs. They were instead equipped with radiometric sensors for the MIDAS early warning satellite program. Both missions were successful.63
After a launch failure, wreckage from the launch vehicle and the classified payload would be returned to a classified facility for painstaking examination in order to determine the cause of the failure. (Photo courtesy of A. Roy Burks)
MURAL AND MAPPING
In addition to problem-solving, CORONA engineers were working on more ambitious plans. On August 9, 1961, a contract was awarded to Itek for a more capable camera system. This contract was made retroactive to March 20—a procedure that appears to have been common during the CORONA program; improvements were often begun before they had been officially approved. This new camera system was known as the “MURAL” or “M” camera, and it filled an obvious requirement from an intelligence standpoint—the need for stereoscopic imagery.
MURAL actually consisted of two C‴ cameras. One was mounted so that it was tilted slightly forward and the other so it was slightly backward, so that their fields of view overlapped. The forward-looking camera (the camera in the “front” of the spacecraft, which was actually faced backward along the flight path) photographed a swath at a 15-degree angle. About a half-dozen frames later, the rear-looking camera also photographed the same swath at a 15-degree angle, but from a different direction. When the two film swaths were properly aligned in a stereo-microscope, the image would appear to be three-dimensional and interpreters could obtain accurate height data on ground objects.64 For the Itek design team, when compared to the engineering problems encountered with the earlier cameras, MURAL proved to be relatively easy—almost fun—to design.65 In practice, the cameras worked perfectly together most of the time, but on the rare occasion when one camera failed, useful images could still be obtained from the remaining camera.66
The two cameras received their film from a dual film supply cassette at the rear of the payload. The two film strips traveled forward and were then bent at a 90-degree angle when they reached their respective cameras. Once exposed, they were then bent again and sent forward to film take-up cassettes inside the reentry vehicle buckets.
The last C‴ camera was launched aboard Discoverer XXXVII, CORONA Mission 9030, in January 1962, but the Agena failed to orbit. A month and a half later, the first MURAL camera was launched on Mission 9031.67 This launch, on February 27, was publicly known as Discoverer XXXVIII. It was the last Discoverer launch. Those in charge of CORONA had realized that the cover story had worn thin; there were now simply too many launches for this to be considered an experimental, scientific program. Thereafter, all launches would simply be listed as classified Air Force launches under the designation “Program 162.”68
Mission 9031, in addition to the MURAL camera, also included a small Index camera with a focal length of 1.5 inches that took a small-scale photograph to be used for matching the panoramic swaths to the terrain. A few missions later a nearly identical Stellar camera was added which, in combination with the horizon cameras, provided a much more precise indication of vehicle attitude.69
The panoramic camera sacrificed geometric relationships in order to maximize image detail. The true geographic location of an object on the panoramic film frame could be positioned to only a few miles, using knowledge of the orbital track and altitude, the time that the photograph was taken, and the attitude of the spacecraft. But this was not precise enough for targeting purposes and hence the government asked Itek to develop a camera that would photograph a large area on each frame to permit the panoramic photographs to be precisely positioned within a geographic framework.
The Index camera, as it was known, was actually a commercial Hasselblad camera modified for use in space. The German-manufactured Hasselblad is a very high quality camera frequently used by professional photographers, but it was not designed for use in a weightless vacuum, where lubricants and sealants could evaporate and condense on a cold lens surface, thereby fogging the image. It was also not designed to carry the large amount of film that would be required for heavy on-orbit use. The camera required some redesign. Early versions of the Index camera had a slight distortion problem. The camera was redesigned with a new shutter manufactured by a commercial firm lacking any security clearance; thus the firm could not be told the purpose for which the shutters were to be used. The manufacturer of the shutter used talcum powder as a shutter blade lubricant, but this was not acceptable for space use. Itek engineers suggested the use of whale oil instead. Because whale oil cost $11,000 per gallon, the manufacturer balked. This meant that Itek engineers had to accept the new shutters with talcum, disassemble them, clean and lubricate them with whale oil, and reassemble them. Only about a drop of oil was needed for each shutter, meaning that Itek was left with a lot of unused whale oil.70
A KH-4 CORONA payload during vibration testing at Lockheed’s Advanced Projects facility. The KH-4 had only a single film-return bucket. This version of the spacecraft was bigger than the early version (shown in the introduction) due primarily to the addition of a second camera within the cylindrical section. But it was not as big as the KH-4A shown in chapter 2, which included a second film-return bucket below the first.
Panoramic images could be used for mapping purposes in combination with the dedicated Index camera. Because they were curved, one inch at the center of the photograph represented a different ground distance than at the ends of the photograph, which meant that the images were not suitable for mapping purposes. Itek manufactured optical rectifiers—essentially a photographic enlarger that reversed the original process (in which the “flat” earth was imaged on the cylindrical film surface). In the optical rectifier, the film was held in cylindrical form (just like in the camera) and a light behind it projected an image via a lens onto a flat sheet of photographic paper which represented the flat earth (the rectifier also compensated for the earth’s curvature). The print that was produced removed the distortion inherent in the film process; one inch on the print represented the same ground distance, no matter where it was located on the image.71
Before CORONA, geographic positions of many locations within denied territory were known to no better than an accuracy of 30 miles. Due to the development of cartographic cameras, by the end of CORONA they were known to within 400 feet in the horizontal plane and 300 feet vertically. Many points were known to within 150 feet in the horizontal plane.
Sometime before the first MURAL camera was flown, officials at the NRO, which was created in 1961 to assume overall charge of American
reconnaissance efforts, designated all reconnaissance satellites with the code name KEYHOLE and a numerical designator. The MURAL system was labeled KH-4. The designation was retroactively extended to earlier cameras: the C-camera missions were labeled KH-1, the C′ missions KH-2, and the C‴ missions KH-3. ARGON was labeled KH-5.
Sixteen KH-4 missions were launched in 1962, with only two failures. Three ARGON missions were successfully launched during this same period. Satellite reconnaissance had become an operational intelligence method, well integrated with the intelligence process overall.
CORONA PLAGUES CORONA
Film from early CORONA flights occasionally returned faded or marked with spectacular branch-like patterns. This phenomenon, common to many different reconnaissance cameras, was known as “corona.” Although the problem had been somewhat sporadic and inconsistent in earlier CORONA missions, CORONA Mission 9040, launched in July 1962, returned with extensive film damage due to corona and radiation fogging.
The problem was not new to the optical engineers who had designed the CORONA cameras, although the extent of the problem was unexpected. Corona occurred when the film dried out in the vacuum of space after unwinding from the film-supply cassette on its way to the camera. Static electricity built up on the moving parts of the camera, particularly the rubber film rollers, and the resulting sparks affected the film. Although the cause was well understood, finding a solution was far more problematic. A group of Itek environmental test-lab people, after careful trial-and-error testing, replaced the material in the film rollers with a new material that did not build up any static electricity.72 The rollers were ground to size and steamed with water in a pressure cooker, and then mounted in the engineering model of the camera and tested in a vacuum chamber. If they produced no corona marking, they were tested in a flight camera two more times. If they failed any of these tests, they were rejected and refurbished. The acceptance rate was about 10 percent. Frank Madden, who was then in charge of the camera at Itek, had to procure precisely machined film rollers from subcontractors who could not be told what they were for, a factor that limited their ability to respond to the problem.73