In the late forties, William Guier worked at Los Alamos using the computational modeling techniques developed by Richard Feynman, and with the IBM CPC hardware that had followed the CPC precursors used during the war. In the early fifties he joined John Wheeler’s thermonuclear simulation effort Project Matterhorn and worked at the National Bureau of Standards programming extensions of the Los Alamos methods on the newly completed Standards Electronic Automatic Computer, working seven nights a week on the machine. Meanwhile he had begun working at Johns Hopkins, and when Sputnik was launched in 1957 he used its newly acquired Univac mainframe to model the orbit determination accuracy achievable using the frequency shift tracking data his colleagues were gathering from Sputnik passes. From early in the effort, John O’Keefe collaborated on the effects of irregularities in the Earth’s gravity field, in particular the north-south hemispheric asymmetry O’Keefe had predicted and termed the pear-shaped effect. Guier’s work demonstrated that accurate orbit determination from a single satellite pass was possible, given processing of an entire horizon-to-horizon frequency shift curve and a nine-dimensional estimation space comprising the six classical orbital elements alongside three instrumental parameters. [1]
In early 1958 Guier’s colleagues proposed inverting the problem. Rather than solving for an unknown orbital trajectory given a tracking station location, an observer at an unknown location could determine terrestrial coordinates from a known satellite orbital trajectory. Fixing the satellite orbital elements collapsed the dimensionality of the parameter space from nine unknowns down to five, with two geographic coordinates and three instrumental parameters, yielding very high observer location accuracy. The work rapidly evolved towards cooperative satellites broadcasting at two frequencies to cancel ionospheric delay and reduce the observer’s estimation problem to just three parameters: latitude, longitude, and local frequency offset. Computer simulations confirmed extraordinary positioning accuracy that directly addressed a pressing strategic requirement for Polaris submarines to determine their launch coordinates at sea within a few hundred feet without exposing themselves. The complete operational architecture of the Transit Navigation System was drafted over a single weekend, pairing a constellation of polar-orbiting satellites broadcasting stable dual frequencies encoded with predicted orbital elements, a ground tracking network solving the direct orbit-determination problem, and an injection station uploading refreshed orbit information so submerged submarines could compute global position fixes. Accepted immediately by the Navy, Transit joined early meteorological spacecraft as the first two operational satellite systems in the world.
For geodesy, Guier and his colleagues recognized that the Transit system provided the continuous, time-dense observations necessary to capture longitude dependent gravitational signatures. Pre-Transit satellites had already provided latitude dependent information concerning Earth’s equatorial bulge and north-south pear-shaped asymmetry, based on the slow, secular precession over months of a satellite’s orbital plane. Longitude dependent geodetic information remained elusive however, because the Earth rotates beneath the orbital plane and longitudinal mass anomalies exert rapid, periodic forces that reverse sign before the orbit can accumulate large positional deviations, producing sub-daily orbit perturbations of only a few tenths of a mile. Working with an IBM 7094, researchers introduced a rigorous data-compression and perturbation framework to extract the subtle longitudinal perturbations without overwhelming the computer’s limited core-memory. Rather than fitting gravity field parameters directly to raw data, they aggregated each satellite pass into two orthogonal geometric observables: the along-track residual, which reflected timing and velocity errors near the time of closest approach, and the slant-range residual, which captured radial and cross-track errors. With this data-flow and iterative least-squares modeling of the residuals using the 7094, by 1963 the Transit system was reliably mapping longitudinal Earth gravity features. [2]
![]() |
| The Transit system was very much the parent of the GPS system, a generation older and retired in 1996. |
![]() |
| Transit and the 7094 grew up with the Space Age, and both the 7094 and PDP-6/10 were 36-bit classics of the era. |
[2] Guier, W. H. “Determination of the Non-Zonal Harmonics of the Geopotential from Satellite Doppler Data.” Nature 200, no. 4902 (October 1963): 124–25.
[3] National Aeronautics and Space Administration. Satellite Geodesy, 1958–1964. NASA Special Publication, NASA SP-94. Washington, D.C.: National Aeronautics and Space Administration, 1966.
Utopia Time Capsule for satellite geodesy at UT Center for Space Research is evolving here.
Previous posts connected with Transit
https://decwarorg.blogspot.com/2026/05/tranet-mcmurdo-and-austin.html
https://decwarorg.blogspot.com/2026/09/transit-satellites-and-1961-trw-130.html





No comments:
Post a Comment