Eye in the Sky: The Story of the CORONA Spy Satellites
Page 29
As satellite techniques progressed, we could easily prove that our strategic forces were clearly superior to those of the Soviets, and that U.S. superiority would continue to grow. Sherman Kent, the Director of the Office of National Estimates, after seeing our tabulation of Soviet missile, bomber and submarine forces, said, “Hell, this is no longer an estimate, it’s a fact book.”
Overall, each day and each frame of film was a most rewarding adventure and satellite imagery became the most valuable intelligence source of physical information. We knew that our interpretation analysis had to be precise because we knew we were only a few steps away from the president, his policymakers, and Congress.
Like Brugioni, David Doyle was an imagery analyst and manager at NPIC during the CORONA era. He specialized in the interpretation of Soviet ballistic missile systems and was one of the key imagery analysts involved in the assessment of the Soviet intermediate-range ballistic missiles in Cuba during the Cuban missile crisis of 1962. Doyle was present when the first CORONA images came in for analysis in August 1960. In the following passages, he recounts the processes, techniques, and strategies analysts developed to tease information out of the imagery. He also notes the teamwork that made the CORONA program so successful.
The Steuart Building in Washington, D.C., located at the corner of Fifth and K Streets NW, served as the site of the U.S. photo-interpretation effort for the U-2 and early CORONA missions. A car dealership, petroleum company, and real estate company occupied the first three floors of the building; the Photographic Intelligence Center (later the National Photographic Interpretation Center) occupied the top four floors. Intelligence specialist Dino Brugioni is in a trench coat at right. (Photo courtesy of Dino Brugioni)
When the film from Mission 9009 [the first successful CORONA mission] arrived at the Steuart Building at 5th and K Streets in northwest Washington, it was the start of a new era. That mission gave us more coverage than all the U-2 missions. And that was to be repeated on every day that a CORONA satellite was in orbit. So, over a year’s time, it was a tremendous amount of information at our disposal.
Photo-interpreters using micro-stereoscopes in the Steuart Building. A rear-projection viewer used for CORONA imagery is on the right. The ceiling consisted of pressed seaweed, which constantly flaked and fell on the photo-interpreters and their materials. The facility was less than ideal for the rapidly expanding photographic interpretation effort. The building was used for photo-interpretation and mission planning from 1956 until the move to Building 213 at the Washington Navy Yard in January 1963. (Photo courtesy of Dino Brugioni)
That first mission, even with its 30- to 40-foot ground resolution, was good enough to actually see SAM [surface-to-air missile] sites on it, and many of them we identified. The same thing held true for the second successful mission in December of 1960. We hit paydirt on Mission 9017 in June 1961. It was on this mission that we found the first deployed ICBM sites still under construction and a number of MRBM [middle-range ballistic missile] sites west of the Urals. And, by the fall of 1961, we had established imagery signatures for five different deployed ballistic missile systems—the SS-4, -5, -6, -7, and -8. The -6, -7, and -8 were the ICBM range missiles.
By that time, we had reorganized how we approached things, and had broken up into different groups to increase efficiency at NPIC. We had what we called “search” teams, which were groups of two or three image analysts, two of which were looking at a rear-projection viewer that the film slowly went across. You had another image analyst that sat at a 40-inch light table with a stereoscope who could take photographic coordinates in x and y. We had another person, a collateral support person, who got maps, other data, and generally helped the image analysts in their chores. And we had specialist teams who knew offensive and defensive missiles, nuclear energy, and naval installations. And these same teams kept following the same targets. So, as we got multiple coverages of installations, we rapidly became as familiar with those installations as we were with our own neighborhoods.
In the spring of 1962, we had our first stereo coverage. That’s when we added two cameras to the first KH-4 and that was as good as increasing resolution by a substantial amount. We were able to see that the “scratching” on the earth’s surface was actually an excavation for a hardened group of silos that was being put in at a missile site. It wasn’t just scarring on the ground getting ready for something else.
It was a very arduous search process to follow everything once you found out that the Soviets were deploying missiles, or doing something unexplained. It was all very time-consuming. The intensity of your analysis was the big variable. You took a difficult situation and you could spend literally weeks going over a couple frames of imagery if you were really looking for very, very subtle signatures. Other times, you could do many frames of imagery in a day if you were just looking for growths, such as new road constructions or something like that.
How did we approach the deception problem where they would have rubber submarines or rubber airplanes or models? We did come across some of those. Luckily, human beings are not totally disciplined and over time make mistakes. The rubber deflates and people are too lazy to go out and blow it back up again. The tail on the wooden airplane falls off, or a piece of equipment doesn’t move for two and a half years. Or they build a missile site with no road to it.
Also, the maps that we had in the early days were really quite rudimentary and we had to update them for a large part of the Soviet Union. We found railway lines, power grids, and microwaves, and all of these things had to be cataloged and literally tens of thousands of new targets were entered into a database, which in those days wasn’t simply sitting down at a word processor and putting it in and having the computer do all the work for you. It was writing out the information in longhand, and then having a data-entry clerk put it onto IBM cards and running those cards through the computer. It was a very manpower-intensive process that went on in those first years. Those readouts then were published in hard copy over the next few days and sent out electrically on the old low-speed printers as basic reports, which tied up many communications lines for hours and hours on end all over the defense establishment.
We had a lot of “possibles” and “probables” thrown around in those reports. In general you can interpret a “possible” to mean a 50/50 chance of what it might be and that depends on how the individual P.I. feels on the day he’s writing it up. The “probable” may mean 90 percent, 95 percent, or if the P.I. is real conservative, 99 percent. We still use that to this day in imagery analysis reporting. It gives a feeling to the person reading the report of just how sure we are of the situation. Over the years we established not only the fact that you can find ICBM sites; we also learned more about their organization by tracing cable lines between silos. We could see that they were being deployed in groups of six or ten and then we could separate them by when they were being constructed at launch sites in launch groups. Then, after the sites went operational, we could determine that the Soviets had, depending on the system, a two- or three-year maintenance cycle where they would remove the missiles, or went in and worked on them, so that the sites were off-line for a while.
The Severodvinsk shipyard, USSR, on February 10, 1969. A large missile-firing submarine assembly and outfitting facility is located at the center of the photograph. Note the ice-breaker tracks in the river indicating that naval vessels are about to leave. (Photo courtesy National Photographic Interpretation Center)
We could do the same thing for a ground forces installation. We could tell what kind of unit was there, whether it was a regiment or a division, and to some extent what their readiness was in terms of how and what they had assigned to them.
Overall, this system was a success, especially when you consider that over a 12-year period it went from 30 or 40 feet [of resolution] down to 5 or 6 feet, and increased the amount of imagery per mission ninefold. You can see that a number of improvements were made.
Despite the p
rogram’s overall success, imagery analysts were still sometimes unable to answer some intelligence questions. For example, sometimes they had trouble accurately evaluating several sites and installations in a timely manner. As Richard Kerr explains:
We were not perfect in this process. We had a lot of anomalies. We had an anomalies list at one time and it was significant. There were tens of tens of tens of facilities that were big and complex and surrounded by triple barbed wire. Sometimes an airplane would be sitting inside with its engine turned on and we’d say, “What the hell is this?!” You know, some of these things we never figured out. We called them “sensitive operation complexes.”
A Mann stereo-comparator used for precise measurements of strategic installations and objects. (Photo courtesy of Dino Brugioni)
David Doyle adds:
Sometimes analysts became overloaded and missed targets and made misidentifications. Sometimes we found just about everything, but some of it was later than it should have been. I think our batting average was pretty good, but we certainly weren’t perfect. We misidentified some things and had to recant and tell them, “That’s not a missile site, that’s a haystack.” These things happened in the process, but, over time and through repetitive coverage, we ended up getting a pretty accurate readout and feel for installations and overall capabilities.
As this last statement suggests, regardless of such miscalculations, analysts were still able to make a number of important discoveries and assessments by using CORONA imagery. As Doyle, Huffstutler, and Brugioni recount, there were many specific sites, installations, pieces of hardware, and world events that the U.S. Intelligence Community learned about through CORONA. For example, as Doyle notes:
Every satellite pass in those days was a new event. We were seeing areas of the Soviet Union that had not been seen since World War II. We found cities that people had only heard of. We followed a lot of different things. We saw, at the Sary Shagan building, that initial ABM [antiballistic missile] prototype which was later deployed at Moscow. The patterns were the same, so we knew what we were dealing with. The Soviets were pretty well deployed out in the Far Eastern maritime provinces and in the West.
CORONA also was a critical source of information along the Sino-Soviet border. It was the best source for telling U.S. policymakers just how many divisions the Soviets had along the border and what the Chinese had on the other side. And we followed that in detail over the years. We followed their missile systems. The Shuang Cheng-Tzu missile test center was found by CORONA and we followed it extensively during China’s subsequent ballistic missile deployment.
Huffstutler remembers that when the Russians started constructing the Berlin Wall in 1961, CORONA film gave
President Kennedy a very good idea of what was around the Berlin area and what kinds of reinforcements from the Warsaw Pact were available in the event they decided to either seal off Berlin, or decided somehow to take over the entire Allied sector of Berlin. That helped senior foreign policymakers to keep from making mistakes.
In a similar vein, Brugioni recalls that during the Cuban missile crisis of 1962, CORONA allowed
Dave [Doyle] to follow the intermediate-range ballistic missile sites in Russia, and make this call: “This is an intermediate-range ballistic missile site [in Cuba].” He had to convince [U.S. Attorney General] Bobby [Kennedy] and [Secretary of Defense Robert] McNamara that the patterns that we were seeing in Cuba were exactly those we’d seen in Russia.
And, on a more inclusive note, Brugioni reveals:
Among the world events captured on this film were the Russian and Chinese nuclear programs, and the Six-Day War between Israel and the Arab nations. We could look and see how in a matter of about three hours, the Egyptian, Jordanian, and Syrian air forces were taken out by the Israelis. You could see the Soviet invasion of Czechoslovakia; the Soviet race to the moon; the Chinese/Soviet border conflict; the India/Pakistan war; the Vietnam conflict; the Kystym nuclear incident; the construction of the Berlin Wall; the Chinese takeover of Tibet; and you can also see the remains of the Gulag Archipelago.
An illustration of a Hen House phased-array radar system and a CORONA photograph of a similar facility. Such line drawings were frequently produced from satellite photographs in order to conceal the technical capabilities of the satellite cameras. Hen House radars were used primarily for tracking space objects and served well into the 1990s. (Illustration courtesy Teledyne Brown Engineering)
In sum, CORONA provided analysts with the imagery necessary to make some vital assessments about other countries’ military, political, social, economic, and cultural status during a difficult period in foreign relations. As Kerr, Huffstutler, and Brugioni note, CORONA has continued to have an impact on some of the problems and challenges that the United States faces, including military strategic planning and even environmental issues. Kerr, for example, thinks that
the CORONA project was quite honestly an extraordinary period. It was the business of imagery, and the business of taking that information and integrating it into a product. The problem was then to try to figure out how to control it, what to do with it, and how to put the information into context, and in some ways, that’s the magic, at least from my perspective, that we did.
Are there lessons to be learned from CORONA that would apply to current threats from missiles, whether ICBMs or others? I think the simple answer is absolutely yes. I don’t think there’s any question that there’s a transferability here of information and knowledge. This breadth and depth of knowledge and understanding of facilities and movements and activities and equipment is fundamental to being able to understand new developments in the former Soviet Union and other countries.
Rae Huffstutler adds the following assessment of CORONA’s overall influence:
It was imagery that set the stage for the arms limitation talks. We began drafting the verification capabilities of the Intelligence Community in some interagency papers written in 1968 and 1969, several years before the dialogue began. With imagery, we could go to a numbers-based strategic arms limitation negotiation with the high confidence that we didn’t need any help from the other side to verify it. The first phase of the arms limitation negotiations which dealt with either pulling forces back from a border, taking them out of a zone, or forbidding the presence of specific kinds of weapons systems, was only possible because of CORONA imagery. It gave us a politically acceptable way of demonstrating if forbidden forces were present, or if the treaty had been violated. That simply would not have been possible with other sources of evidence.
And Dino Brugioni notes that
the information gleaned from the CORONA program saved millions of defense dollars. Defense planners came to NPIC and, with information gleaned from satellite imagery, could plan counterweapon programs against our adversaries more effectively, more efficiently, and much cheaper.
New methods and equipment had to be designed. The microstereoscope and advanced comparators were developed to speed the interpretation effort. Some of the equipment designed to enhance satellite imagery by the National Photographic Interpretation Center has been found by researchers to aid in the detection of breast cancer.
But perhaps the best overall assessment of CORONA’s impact on intelligence is Rae Huffstutler’s:
The great contribution of intelligence during this 1960s period—and I think particularly a contribution of the CORONA system and all those that worked on it—was to help the United States plan a safe passage through a very dangerous period where a lot of mistakes could have been made had you simply assumed you understood the enemy, his intentions, and his capabilities, but had not based those judgments on hard data and the facts that came out of imagery.
Appendix A
PROGRAM OVERVIEW AND CAMERA DATA
PROGRAM OVERVIEW
CAMERA DATA
ADDITIONAL CAMERAS ON THE KH-4. KH-4A. AND KH-4B
Camera model Index and Stellar/Index DISIC
Type Frame Frame
Resolution
Ground (ft) 400-500 (est) 100-400 (est)
Film (lines/mm) — —
Coverage 166 × 166 miles (est) 140 × 140 miles (est)
Appendix B
LAUNCH LISTINGS
PREPARED BY JONATHAN MCDOWELL
Notes to Appendix B
aLaunch sequence number in CORONA program.
bType and serial number of Agena space vehicle. The satellites were referred to by this number at Lockheed. SS-01A and SS-01B are Standard Stage variants of the Agena D.
cProgram within CORONA project.
dMass of satellite after orbit insertion.
eDates may differ by one day from sources which give local time.
fThe CORONA flights were referred to by this classified mission number by the CIA.
gDate of SRV recovery, Greenwich Mean Time. Many SRVs were recovered late on the previous date according to local time.
hType and serial number of booster. The DM-18A was similar to the Thor IRBM.
The DM-21 was modified for space use with a shorter guidance section, and the DSV-2A (USAF designation SLV-2) was a standardized version of the DM-21. The DSV-2C (USAF designation SLV-2A) had three Castor strap-on boosters and was also called TAT (Thrust Augmented Thor). Finally, the DSV-2L (USAF designation SLV-2G) had three Castor II solid strapons and a stretched tank; it was also known as LTTAT (Long Tank Thrust Augmented Thor) and Thorad.