Eye in the Sky: The Story of the CORONA Spy Satellites

by Dwayne Day

  The first mission we received material from was the KH-1 CORONA. When the materials arrived, we were absolutely amazed by their content and detail. We said, Boy! This is going to be great material for compiling and revising 1:250,000 maps and 1:200,000 ATCs [Air Target Charts].” Preliminary experiments revealed that the KH-1’s pan geometry was manageable, but, as expected, was totally incompatible with the exisiting state-of-the-art stereo analog mapping equipment. Nevertheless, its imagery content and stable geometry suggested its great potential for mapping and targeting using analytical photogrammetry technology.

  The CORONA managers and contractors’ attitudes and actions were also an important factor in bringing about major changes in mapmaking during the period. One day, for instance, we were talking with some Itek representatives, and we mentioned to Ron Ondrejka, “You know, Ron, if we could just get a framed camera in this thing to calibrate the panoramic imagery, we’d have a first-class mapping system.” So we talked and talked, and about a month later, Ron came back and said, “You know, I think we can get that camera you want because we’ve found a hole in the CORONA vehicle where we can pack it.” Believe it or not, by the time the KH-4 system came along, we had our little 1½-inch camera system sitting inside that vehicle. It didn’t have all the bells and whistles that we would’ve liked to have on it, but the KH-4 “bird,” as we called it, was the first bird that was adapted to support MC&G mapping and target location requirements. Further improvements included adding a more precise timing system for frame exposure, and horizon and stellar cameras for attitude determination.4 With these additions, CORONA started to become a true geodetic camera system.

  By the time the KH-4A imagery began rolling in, we were able to make accurate panoramic photogrammetric measurements. Pan photo rectification had been established. Orbit positions were more accurate and the time of frame exposures was becoming more precise, which facilitated the beginning of analytical triangulation techniques using both orbital and attitude constraints. MC&G was finally in a position to meet the ATC requirement of 300 feet (linear error—90 percent) for relative vertical accuracy anywhere in the Soviet bloc. We were also able to use the new material to do stereo-compilation and point positioning. The KH-4A and B added other improvements. In addition to improving the on-board cameras, Doppler was added for improved orbit determination. MC&G multiple orbit strip and block triangulation adjustment techniques were developed to better calculate the “drag” variable affecting change in the movement of the space vehicle. These adjustments were also designed to improve the consistency of joins between frame models over large areas of the Soviet bloc. The WGS 66 [World Geodetic System, 1966] ellipsoid was constantly improved in order to achieve the WGS 72, which better defined the earth’s geoid and gravity field.5 These accuracies determined the way the Minuteman system evolved by allowing the use of a single missile to carry multiple lower-yield warheads.

  CORONA wasn’t just an intelligence and target location tool. It was also a major instrument for compiling the cartographic and digital products that were necessary to support ground forces. By 1969, we could reliably predict any position on the earth’s surface to within 450 feet with a 90 percent accuracy rate. That rate, besides meeting the SAC’s ICBM target location requirements, also met the DMA’s [Defense Mapping Agency] mapping requirements for medium-scale maps. CORONA also got into the business of supporting air tactical weapon systems. During the Vietnam War, for instance, we took CORONA stereo pan image models, geodetically calibrated them, and passed them out in the field as point-position databases (PPDB). New reconnaissance aircraft photographs were then correlated to these PPDBs. As a result, field forces could get a quick response target position readout over densely vegetated areas within two to four hours. Photoreconnaissance materials from the SR-71 Blackbird—which was being flown over Vietnam at that time—would come in. Targets would be identified and located by correlating them to pan PPDBs, to give target coordinates for future weapon strikes. This approach to ground force targeting addressed the major difficulty in Vietnam of finding targets under heavy tree cover. The only way you could determine a precise target location was by doing a stereo correlation to the pan PPDBs. Another application was the use of CORONA pan materials to compile TERCOM (terrain contour matrices) digital terrain matrices, which were used for the in-flight guidance control of cruise missiles.

  A comparison of the Dual Image Stellar Index Camera (DISIC) mapping camera coverage and the stereoscopic panoramic camera coverage. The Index and later DISIC cameras were used both for mapping and for locating the position of target objects on the ground.

  By WGS 72, MC&G had developed an accurate mathematical geoid and gravity model of the earth. Various geodetic datums from around the world were tied together using this model to create a world geodetic system. This information, in turn, was fed into CORONA’s guidance and control to improve its orbit. Improved orbit data was then fed synergistically back to MC&G to improve its target location capability.

  To complete the story of the impact of CORONA, I again call your attention to the fact that before CORONA, DoD map compilation was based on the use of standardized 6-inch focal-length frame, distortion-free, low-resolution stereophotographs processed in Multiplex/Kelsh and other commercially available analog instruments. To make the transition to CORONA, a major technological revolution also had to be accomplished in the management and growth of DoD MC&G mapping organizations, including the in-house training of the DoD MC&G’s workforce to use analytical techniques; the redesign of the DoD MC&G’s production process; and the coordination of DoD MC&G’s research and development through the Army, Navy, and Air Force labs.6

  The techniques offices were associated with production elements so that every element essentially had its own R&D team that could solve its production problems. Originally, there was a conflict over whether there should be one techniques office solving everybody’s problems, or whether several techniques offices should head up smaller production teams. The site-centered management method brought immediate improvement.

  In summary, the impact of the CORONA system on MC&G was profound. It got us into the satellite business and made us develop new techniques and new instruments that enabled us to handle panoramic material as well as any type of future image-collection system. Eventually, as a result of the entire MC&G community’s expansion, an OMB Federal Mapping Task Force was formed in 1972, and a study was conducted which resulted in all DoD MC&G resources being consolidated under the Defense Mapping Agency.7

  Kenneth I. Daugherty was one of Mahoney’s colleagues at the ACIC. Unlike Mahoney, Daugherty was a geodesist, not a photogrammetrist.8 While geodesists and photogrammetrists often worked together very closely during the CORONA program, there was still an important distinction between them. For example, geodesists attempted to map the world based on ground surveys, whereas photogrammetrists mapped the earth based on aerial photographs. Speaking from a geodesist’s perspective, Daugherty details the training process that such specialists went through during the CORONA era and how they developed a global mapping system based on a series of old ground surveys plus CORONA images. He also explains how mapmakers would then take suspected enemy targets and locate them on the new map system and gain a more complete intelligence picture. Recounting the complexities of the CORONA mapping process from the very beginning, Daugherty notes:

  There were only a few courses in geodesy offered in this country at that time. Syracuse University was one of the few schools that offered geodesy at the undergraduate level. Geodesy was not taught at the graduate level in this country until Ohio State formed a graduate program in the 1950s. Geodesists were traditionally trained on the job or in Europe, which was the center of geodetic thought and practice. With financial support and encouragement from the Department of Defense and the CIA, however, Ohio State imported, almost lock, stock, and barrel, a European faculty that established a program at the M.S./Ph.D. level in geodesy, photogrammetry, and cartography in the early 19
50s. Out of that program came folks like Bill Kola, Bill Mahoney, me, and many others.

  During the 1950s, surveys were based on local datums. Two such datums were the North American Datum of 1927 and the European Datum of 1950, which was the result of an adjustment made by the Army Map Service after World War II. We also had datums of the Far East which had been tenuously calculated by using long lines of triangulation and some measurements across the ocean. Asia was tied to Europe by a survey along the trans-Siberian railway and then an arc that went down through India.

  The Roman Ruins at Lejjun, Jordan, September 29, 1971. Note the heavy defensive walls. CORONA revealed many terrain features not visible by other means. (Photo courtesy National Photographic Interpretation Center)

  During this period, we had some rather forward-looking folks in the Air Force who went out and started to build a program by hiring folks who were already in the business, like Bill Mahoney. Then those people would hire folks who were just coming out of college, like me. I had a brand-new degree in mathematics, geography, and geology and was recruited by the ACIC. The recruiter said to me, “We’re really interested in your cartography and your geography skills, but your math probably won’t hurt either.” So, within a year I was doing projections and grids on a desk calculator at the ACIC. Then they came around to me and said, “How would you like to go to Ohio State and study geodesy?” And I said, “All right!” Then I looked in the dictionary to see what the hell “geodesy” was.9

  A large group of us, 45 civilians from ACIC and five officers from SAC, went to Ohio State, and it was one of the most intensive educational experiences I had ever had. The folks at OSU followed the ACIC’s advice and made sure that since we were getting paid eight hours a day, that we remained in the classroom eight hours each day.10

  After training at OSU, we came back to the ACIC and started to work on putting together a geodetic system in which you could plot points from different continents on a coordinate system. This would allow us to deploy ballistic missiles accurately on target from one continent to another, and to track satellites. So we came up with a World Geodetics System in 1959, 1960, 1966, and 1972. Every day, we were doing new and revolutionary work because there were no reference books on the subject, and there were no programs or manuals. We had to invent everything from the outset.

  Our source materials on the Soviet Union were pre-Revolution 1917 surveys and early surveys of the communist era that had been captured by the Germans during World War II. We used methods like looking along the trans-Siberian railway for the old wooden triangulation towers that the Russians had used when they did some of their previous surveys. When CORONA flew over, we would go and find these towers—which were still sitting there rotting along the trans-Siberian railway—and correlate their geodetic coordinates, which we already knew from the old surveys, with the new satellite photos we had obtained.

  We also soon realized that what we needed to do was survey the entire earth from space. We could not correlate the mishmash of information about an area without such a survey. We needed a common denominator to hold it together. So we came up with a process that required a satellite with a calibrated camera, a precise orbital position, and an exact camera attitude. As we fed each of these factors into the system, they were incorporated into the CORONA program and into later programs to the point where we could survey from space and create accurate maps and targeting materials. After all, east of the Urals, we only knew where some places were within 15 to 30 miles, and some places we didn’t know about at all.

  In 1966 or 1967, mostly out of intellectual curiosity, I also decided that since we were doing what we called G&G calculations (i.e., the geodetic and gravity effect on ballistic missiles), I would sit down and do some reverse calculations by making assumptions about what the Soviets knew based on the information that was available to them. In fact, they could go out and buy the USGS (U.S. Geologic Survey) quads which unintentionally revealed our missile fields by showing where our missile fields’ fences were located.11 My reverse calculations uncovered that they did not have the same problem we had. The targeting problem was much simpler for them. They could go out and, for less than one dollar, buy a 1:24,000 map and then actually go out and walk around a missile field’s fence. So, from the outset, their ability to have a more precisely targeted system was there. Whether they did or not, I can’t say for sure.

  During this period, we were also working on a program to improve the one-nautical-mile accuracy of the Atlas ICBM system to something that would be closer to a tenth of a nautical mile in the Minuteman series. For those of you who visited St. Louis during that era, you will remember that they were building the St. Louis Arch. Interestingly, the St. Louis Arch is almost exactly a tenth of a nautical mile in height. What we were planning to do in the ballistic missile and geodetic world was to stand 5,400 nautical miles away from a target and lob every other missile into an area the size of that arch. This analogy gives you a visual picture of what we were trying to accomplish.

  Because gravity variations affected the way satellites and ballistic missiles moved, we also set out to build a gravity library so that we could correlate gravity information from all over the world and improve the “Earth Gravity Model,” which in turn helped us define the geoid. We managed to accomplish these goals in a rather rudimentary fashion for WGS 66, in a much more refined fashion in 1972, and then again in another 12 years with WGS 84. Today’s accurate worldwide navigation and surveying capabilities came about as a result of WGS 84 and the Global Positioning System.12 Back in 1967, on behalf of the community of geodesists and folk of that ilk, I briefed Dr. Harold Brown, who was then DDR&E [Director of Defense, Research, and Engineering], and convinced him that the department ought to spend about $25,000,000 to update the geodetic and gravity models so that we could improve the Minuteman I system—and he bought it. From that point on, we never looked back, and as a result, the DoD now has the world’s finest geodetic and gravity source materials and the ability to exploit those materials in support of national programs.

  The Kodak-developed Dalton Spray Processor, which was used to provide precision processing of high-resolution CORONA film. As CORONA materials proliferated and were used by customers such as the U.S. Geological Survey, more copies of the film were needed. (Photo courtesy of the Eastman Kodak Company)

  Lowell E. Starr, the fourth and final cartographer, was a civilian mapmaker with the U.S. Geological Survey (USGS) during the CORONA era. Starr was a key innovator in the CORONA mapmaking process and helped significantly modernize cartographic capabilities during the period.13 In his discussion, he reviews the USGS’s participation in the CORONA program and the way that the various branches of the Intelligence Community cooperated in order to map the earth’s surface. He also explains the important impact that the Intelligence Community’s cooperation had on improving civilian maps. Starr notes:

  In the mid-1960s, there was a study that the BoB [Bureau of the Budget—now the Office of Management and Budget] and the Department of Defense commissioned to analyze the best methods for exploiting CORONA materials to the fullest. The study essentially served as the springboard for getting the civilian community involved with the program. It provided civilians with the opportunity to grasp technological innovations and use them, even though we didn’t get involved until about 1965.

  The USGS was a civilian organization that was very conservative and scientific. We were steeped in the culture of near-academic type research—the public dissemination of maps, and data and scientific publications—and many of us were therefore reluctant to participate in a covert operation. After all, covert data did not support the “publish or perish” world’s fundamental objective because the associated information could not be referenced in a publication. Nevertheless, Dr. Thomas Nolan, who was the director of the USGS at the time, and William Radlinski of the National Mapping Division saw the value of USGS participation in the CORONA program.

  In early 1965, I was reassigned to hea
dquarters from a USGS Mapping Center located in Rolla, Missouri, to serve the final year of a three-year training program. This one-year assignment actually lasted approximately 30 years. Winston Sibert, my supervisor and chief of the Office of International Activities, came to me and asked if I might be interested in participating in a special classified program which could be of substantial benefit to the USGS National Mapping Program. Well, I didn’t have the foggiest idea what was going on, but I understood the general drift of what was possible, so I said, “Yes, I’m interested.” At that time, the USGS was heavily involved with NASA and the Gemini and Apollo programs. So I started working on NASA/USGS cooperative activities so that I could gain experience with space imagery and programs. This occurred during a period when I was waiting to receive my clearance, which usually required only about three months in those days. I later learned the real reason I had been assigned to those activities was so that I could learn about orbital science and a few other things that were applicable to mapping using CORONA materials.

  In 1966, we formed a small team of people and proceeded to pull off a near miracle—at least by U.S. government standards. Winston Sibert formed a small team with the goal of appyling CORONA materials to USGS mapping programs. He hired a security officer, program officer, some USGS production and photolab experts, and me. We were able to establish a liaison with the CIA through Dr. Joe Baclowski, the man who deserves much of the credit for the success of the civilian community’s program. We then received a series of billets—I think 40 to begin with—and a token budget. We selected a site in Reston, Virginia, and built a classified facility (Reston would later become the home of the USGS headquarters). We had a great deal of difficulty building that facility in Reston. Nevertheless, what was so phenomenal about the endeavor was that we built the facility, established a staff, and occupied the building in 13 months for only about $300,000. That would be impossible today.