From Sputnik II to Transit

Damn, I’m going to wrangle a place on that project.

—Harold Black, Transit team member, on hearing of Guier and Weiffen – bach’s work on orbital determination and prediction at the end of 1957

O

n November 8, 1957, Ralph Gibson dipped into the director’s dis­cretionary fund for $20,000 to fund Project D-54—to determine a satellite orbit from Doppler data, and he assigned technical and engineer­ing support to Guier and Weiffenbach.

Shortly afterwards, Guier and Weiffenbach briefed their colleagues in the research center, some of whom would later play an important part in the development of Transit. One, Harold Black, was bowled over. Guier had explained methods and coordinate systems that were on the edge of his understanding, and Black knew that he wanted to be a part of the work. He would be, but not for a few more months. Not until the Transit program was underway.

In the closing months of 1957, Guier and Weiffenbach were tackling the problem of the ionosphere—the region of the earth’s upper atmo­sphere where solar energy separates electrons from atoms and molecules, creating layers of free electrons that reflect or refract radio waves. Reflected waves pass round the curve of the Earth and carry transmissions from, say, Voice of America, or provide medium – and high-frequency channels for voice communication. Higher frequencies pass through the ionosphere, enabling communication between the ground and a satellite.

On passage through the ionosphere, radio waves interact with the free electrons, and the frequency of a signal received on Earth from a satel­lite appears different from the frequency transmitted by the satellite. This muddies the water if you are trying to relate the Doppler shift to the satel­lite’s motion. There is a qualitative explanation that gives a rough idea of what is happening.

Radio waves are, of course, examples of electromagnetic radiation. Sputnik’s oscillator was creating an electric field (with its associated mag­
netic field), and the influence of the field extended through space. The field’s influence set the free electrons of the ionosphere oscillating with the same frequency as the field. The oscillating electrons set up their own elec­tromagnetic fields, which, in turn, extended their influence out through space. The fields from Sputnik’s oscillator and from the free electrons were of the same frequency but were not exactly in step. They overlapped. If one returns to the wave analogy of electromagnetic radiation, it as though the crests of the waves from the field generated by the free electrons occurred at a slightly different time than the crests of the waves from the satellite oscillator’s field. So a receiver on Earth detected more wave crests over a given time than it would if the ionosphere were to conveniently disappear (convenient at least for this application).

The Doppler shift (received minus transmitted frequency) is related to the range rate, that is, to the way the satellite’s position with respect to the lab was changing. So when the received frequency was altered by the presence of the ionosphere, the satellite’s path appeared to be different from what it actually was.

How could the receiver distinguish between the number of crests received in a given time as a result of ionospheric refraction as distinct from the number of crests detected because of the Doppler shift?

It couldn’t. And because Sputnik I generated only one frequency, there wasn’t much that Guier and Weiffenbach could then do by judi­ciously juggling theory and observation. Fortunately, the Soviets launched Sputnik II, and Weiffenbach started recording both frequencies. By com­paring the two signals received, and knowing, from theory, that the iono­spheric effect was roughly inversely proportional to the square of the transmitter frequency, they were able to calculate the amount by which the ionosphere altered the received frequency. They were able, therefore, to remove the effects of the ionosphere from their experimental data. Once Guier and Weiffenbach corrected for ionospheric refraction, the frequen­cies computed were those due to the satellite’s motion and not to passage through the ionosphere. Needless to say, this again sounds far simpler than it was, and considerable effort subsequently went into theoretical and experimental explorations of the ionosphere’s nature.

Fortunately, the IGY was generating new observations of the iono­sphere. Of particular importance for greater theoretical understanding were the Naval Research Laboratory’s data showing that the then existing simplified description of the ionosphere was less complete than had been thought previously. The nonuniform electron density through a cross sec­tion of the ionosphere raised concerns at APL that more than two fre­quencies would be needed to correct for ionospheric refraction with suffi­cient accuracy Guier’s and Weiffenbach’s work with Sputnik II did not settle this question.

Another potential complication was that electron density, which influences the amount of refraction, varies with the amount of energy the sun is pouring onto the earth. There are, for example, more free electrons at noon than at night, and the density varies according to the latitude, the time of year, and the stage of the solar cycle.

Thus in November 1957, the ionosphere was the first quagmire that Guier and Weiffenbach encountered as they analyzed their Doppler data. Their experience led them to conclude in the first Transit proposal, erro­neously as it turned out, that the ionosphere would be the biggest obstacle to developing a navigation system.

Once the Transit program was underway, Weiffenbach and then Guier and then others tackled the ionospheric physics in greater detail and with more sophistication than in Guier and Weiffenbach’s initial paper in April 1958, entitled Theoretical Analysis of the Doppler Radio Signals from Earth Satellites.

Weiffenbach went back to basics, studying Maxwell’s equations describing the behavior of electromagnetic waves as well as the equations that provide a theoretical and simplified description of the ionosphere. He collected much of the published data and satisfied himself that theory and experiment seemed to be in synch. Weiffenbach’s aim was to find the opti­mum frequency for Transit given the constraints that physics and current or foreseeable technology imposed. Ideally, the highest possible value should be chosen, but electronic components of the day did not work at the higher frequencies. Such juggling of theory and practicality was an essential aspect of determining the early specifications for Transit, and it focused the minds of many other satellite designers.

By the end of 1958, Weiffenbach had concluded that two frequencies should do the job. The analysis he submitted to Richard Kershner, the team leader, warned that only experiment would clarify the situation, and the first experimental Transit satellite was aimed at doing just that.

Guier’s later basic theoretical analysis, bolstered by results from early experimental satellites and discussions with Weiffenbach, showed that for Transit’s purposes, physical phenomena such as the irregularity of electron density could be ignored, and that two frequencies provided an accurate enough correction for ionospheric refraction.

Shortly after the first Transit proposal was written in April 1958, they realized that the earth’s magnetic field would also affect passage of radio waves through the ionosphere. Therefore, Weiffenbach, who became responsible for space physics and instrumentation at APL, recommended circularly polarized transmitters. Such transmitters were not unusual, but Henry Riblet, who went on to head space physics and instrumentation at the lab, had to modify existing designs so that they would be suitable for the spherical surface of Transit. Having chosen circular polarization, Tran­sit’s designers were constrained in their approach to stabilizing the satellite in orbit. Thus the physics affects the technology and each technological decision affects others.

As 1957 drew to a close, Guier, later responsible for space analysis and computation, was refining the mathematical model. Weiffenbach was gath­ering data and improving the experimental setup, and together they con­tinued to explore what the model could tell them about orbits. In the meantime they were reading madly, educating themselves in ionospheric physics and orbital mechanics.

Around this time they met someone who was to become their “good angel.” His name was John O’Keefe, and he was working with the Van­guard tracking team, was in fact at the radio tracking station near Washing­ton D. C. on the morning after the launch of Sputnik I.

O’Keefe, who was to become a pioneer of satellite geodesy, had stud­ied the moon’s motion in search of clues to the nature of the earth’s gravi­tational field and had eagerly anticipated the launch of satellites. For the previous year, in preparation for the space age, he had studied orbital mechanics every morning before going to work. He was devastated when the Soviets launched first.

O’Keefe, whose son heard Sputnik’s Doppler shift on a ham radio set, was intrigued when he learned of the quality of Guier’s and Weiffenbach’s Doppler data. He went to APL to take a look and was impressed. When Guier and Weiffenbach determined Sputnik’s orbit, O’Keefe was astounded; he didn’t think they would be able to get that much informa­tion from the data.

Yet there was actually even more in the data. Later, the same curve­fitting technique enabled them to vary values for ionospheric refraction and thus to find indices of refraction in addition to transmitter frequency and the orbital parameters.

O’Keefe tweaked Guier and Weiffenbach about their lack of knowl­edge of the ionosphere and pointed out, too, that the Earth’s gravitational field would turn out to be far more complicated than people yet realized. He brought much humor and knowledge of physics to their relationship.

As a regular attendee at IGY meetings, O’Keefe was up-to-date on emerging results. In the following year, he would analyze observations of Vanguard i, known as “the grapefruit,’’ and he recognized that for the satel­lite to be in the observed orbit, the Earth must be pear-shaped in addition to being oblate. The southern hemisphere, in effect, is larger than the northern hemisphere.

The pear-shaped term for the Earth eventually became the second brush stroke on the canvas of that very simple mathematical model that Guier had initially constructed for his and Weiffenbach’s research during the autumn of 1957. And when the gravitational consequences of the pear-shaped earth were folded into the model, the fit between theoretical and experimental curves improved yet again, reducing the error of the orbital determination and prediction.

Immediately after the launch of Sputnik I, O’Keefe had very little data to work with. Hence his fascination with Guier’s and Weiffenbach’s Doppler curves. He invited them “downtown” to Washington D. C. to meet the Vanguard’s Minitrack team. At the time the Minitrack group was inundated. Calls were coming in from many people whom they really didn’t want to talk to. Guier and Weiffenbach fell into the category of nui­sance, and to their chagrin, the Minitrack people dismissed them and their work. The rebuff rankled. Weiffenbach recalls thinking, “OK, we’ll show you; you will take notice.”

Throughout the early part of 1958, Guier and Weiffenbach focused on improving orbital determination. By early spring, they thought they had the satellite’s position to within two or three miles, and they began writing their key paper on the theoretical analysis of Doppler data. But what next?

It was their boss, Frank “Mack” McClure, head of the lab’s research center, who took the next step. McClure seems to have been a compli­cated character. Various descriptions crop up from those who knew him: smart as hell; prone to terrific rows; blunt; impatient; enormous ego (I doubt this was a unique characteristic); someone who “didn’t fool himself about himself” recalls Weiffenbach. Certainly he seems to have impressed the research center’s scientists with the breadth and depth of his knowl­edge of physics.

At this time McClure’s main interest was in solving the problem of insta­bilities in the burning of fuel in solid-fuel rockets, one of the critical technolo­gies for the Polaris missiles. Hence he was spending a lot of time at Special Pro­jects, as was the man who became the Transit team leader—Richard Kershner.

On Monday, March 17,1958, Frank McClure called Guier and Weif­fenbach to his office. He told them to close the door; a clue, they knew, that something interesting or classified would be said. McClure asked them, Can you really do what you say you can? In the manner of the best courtroom attorneys, McClure was asking a question to which he was sure he already knew the answer. Guier and Weiffenbach did not know this, but they answered yes.

In that case, McClure told them, if you can determine a satellite’s orbit by analyzing Doppler data received at a known position on the earth, it should be possible to determine position at sea by receiving Doppler data from a satellite in a known orbit. The navigation problem, in fact, would be easier, argued McClure, because only two values—latitude and longitude—would have to be determined.

McClure sketched out the concept. The satellite would transmit two continuous waves, which would allow a receiving station on a submarine to record two Doppler curves and to correct for ionospheric effects. The satellite would also broadcast its position at the time of the transmission.

The submarine’s inertial guidance system could then provide an esti­mated value of latitude and longitude. A computer program would gener­ate the Doppler curves that a submarine at that estimated location would receive from a satellite at a known position. It would then continue to generate Doppler curves that would be received at nearby values of lati­tude and longitude. When a best fit was found between the estimated and received Doppler curves, those values of latitude and longitude would be the latitude and longitude of the submarine.

Thus the method that Guier and Weiffenbach had developed would be used in two ways. First, to find a satellite’s orbit by observing it from ground stations at well-determined locations, allowing a prediction of its position. The positions for the next twelve hours would then be uploaded

From left to right:William H. Guier, FrankT McClure and George C. Weiffenbach. Drs. Guier and Weiffenbach first employed Doppler tracking to determine the orbits of satellites when tracking Sputnik. Dr McClure used this principle as a basis for inventing a worldwide navigation system. Courtesy of Applied Physics Laboratory.

twice a day. Second, the same computational and statistical techniques would provide a means of fixing latitude and longitude.

McClure asked them what they thought. They agreed that it sounded plausible. Go away, McClure told them, and do an error analysis.

They did. After the computer simulations they had been working with, simulating the errors in the navigation scheme proposed by McClure was, they say, next to trivial. They looked at the answers and didn’t believe them. They seemed too good. They wondered whether they had left something out, but they knew that they had not. So they increased the errors in the assumed depth and speeds of the submarines as well as several other parameters. Then they went back and told McClure that it would work, and work beautifully.

McClure smiled and said, “I know.”

McClure told Guier that he had had the idea the weekend before and had called Richard Kershner. The two of them, who frequently spent time together away from the lab, had fleshed out the idea over the week­end. They had concluded that the navigation method would work if Guier and Weiffenbach could do what they said they could.

On the day after his meeting with Guier and Weiffenbach, McClure sent a memo to Gibson, APL’s director. The topic was “satellite Doppler as a means of ship navigation.” McClure suggested that APL’s patent group should consider seeking protection for the idea. He pointed out that “While the possible importance of this system to the Polaris weapons sys­tem is clear… an extension of thinking in regard to peacetime use is to me quite exciting.” Other advantages did not escape him, and he added that the establishment of such a system “ … might provide a wonderful cold war opportunity…. This would put the U. S. in the position of being the first nation to offer worldwide service through its venture into outer space.” Work on a technical proposal began immediately.

Developing specifications for a complex engineering system is like plan­ning an elaborate meal without reference to recipes, knowing only what combinations are most likely to work, what fresh produce is likely to be available, and making last-minute adjustments for what is not. The occa­sion for the meal guides the process. First there is the overall plan of suit­able and compatible courses. Then comes the detail, working out all of the recipes, testing quantities, temperatures, cooking times.

McClure defined the occasion and made clear what the “diners” would swallow. He knew the technical needs of Polaris in detail, but of equal importance was his sense of the technological likes and dislikes of Special Projects. If he thought Special Projects wouldn’t like a particular technical approach, he would point that out.

Several factors, therefore, converged to ensure that the Transit pro­posal would win the backing that it did in the face of opposing schemes that were emerging for satellite-based navigation. First, APL was present­ing a good idea and had some smart people able to see it through from theoretical analysis to engineering application. Second, the lab had an acknowledged history of completing complex and difficult applied physics and engineering projects, starting with the proximity fuse. (The Naval Research Laboratory’s experience with Vanguard after the success of Vi­king, however, showed that previous success is not always a harbinger of future triumphs.) Neither factor would have been enough alone, but Tran­sit would clearly have been dead in the water without the lab’s expertise and reputation. Additionally, of the emerging ideas for satellite navigation, Transit met the needs of Special Projects and, because of the Brickbat-01 status of Polaris, was compatible with national priorities.

Of the competing schemes, one, taking angles to emitting radio waves was impossible because the antennas needed would have been far too large for submarines. Even had the antennas been usable by the surface fleet, the scheme would not have been serious competition for Transit because the surface fleet was not expressing a strong desire for improved position fixing, nor had its missions the same high priority as those of Spe­cial Projects.

Another scheme, radar ranging, might have met the needs of the U. S. Air Force, but although Air Force missions did have a high priority and the Air Force had proven in its ten years of existence to be highly success­ful at lobbying the government, the preferred scheme within the Depart­ment of Defense at that time was to develop inertial guidance for aircraft. Since aircraft are not aloft for months at a time, the errors do not have as long to accumulate as they do during a submarine’s tour of duty. (Radar ranging resurfaced as the technical basis for the GPS satellites, which have replaced Transit.)

Also operating in APL’s favor was the lab’s relationship, as the consul­tant assessing the Polaris weapons system overall, with Special Projects. The organizational relationship was bolstered by professional relationships at all levels and by McClure’s political astuteness in making his inside knowledge of Special Projects work to further the technical acceptability of the Transit proposal. In short, the people who proposed Transit were the right people, in the right place, at the right time, and with the political and scientific smarts to see it through.

There was another influential factor, and that was Richard Kershner’s reputation. Kershner’s approach to technical problems was similar to that taken by Captain Levering Smith, then technical director of Special Proj­ects. Commander “Chuck” Pollow, who assisted in the day-to-day man­agement of Transit, is convinced that the perception that Special Projects and Smith had of Kershner was critical to Transit’s initial acceptance. And the Transit team members are convinced that Kershner was critical to the success of Transit.