Summary

  • West’s 1982 result was specific and reproducible in scope. Assuming the altimetric geoid was the best available reference, she adjusted the geoid derived from potential coefficients to minimize differences, applied that correction to the ellipsoid fitted to the coefficient-derived geoid, and reported a mean ellipsoid radius of 6,378,134.9 metres, x and y origin shifts below one metre, and an approximately −2.5-metre z shift.
  • The 1982 paper sits inside a longer Dahlgren sequence rather than a timeless algorithm: West’s 1979 GEOS-3 work compared Wiener and Kalman smoothing using third-order Markov statistics; her 1981 SEASAT report described the reduction of radar-altimeter measurements, precise Doppler orbits and environmental corrections into along-track geodetic products; and her 1986 GEOSAT specification documented another processing system.
  • GPS uses a later WGS 84 Earth-centred, Earth-fixed reference system and broadcasts satellite state in that frame. Receiver positioning is a separate estimation problem involving satellite geometry, clocks, ephemerides, signal propagation and receiver processing; high-precision applications can add external corrections and surveying infrastructure. An ellipsoid is necessary reference geometry, not a complete positioning system.

The 1982 paper is narrow—and strong

The strength of West’s 1982 paper is precisely that it does not claim to solve “GPS.” It asks a smaller geodetic question: what radius and origin shift best describe a mean Earth ellipsoid when SEASAT-1 altimetric geoid heights are compared with geoid heights generated from gravitational potential coefficients?

The abstract states the assumption and adjustment explicitly. The altimetric geoid is treated as the best available reference. Differences between that altimetric geoid and a geoid derived from potential coefficients are minimized; the resulting adjustment is then applied to the ellipsoid that best fits the coefficient-derived geoid. West reports solutions from two altimetric-geoid data sets and coefficient-based geoid heights using DoD WGS 72 and GEM10B. The resulting mean ellipsoid has radius a = 6,378,134.9 metres, with x and y origin shifts under one metre and a z shift of roughly −2.5 metres.

Those numbers matter because they bind the claim to a method and evidence set. They do not travel forward automatically into every later geodetic standard. A fitted parameter is not a permanent property of all future coordinate frames; it is the output of a specified model, data set, preprocessing history and objective function.

The 1982 paper is therefore unusually useful for separating contribution from mythology. It gives a concrete result while also exposing how much else had to exist before that result could be computed: radar measurements, orbital information, geoid estimates, gravity coefficients, reference systems and numerical procedures.

From a radar pulse to a geoid height

A radar altimeter does not observe an ellipsoid or a GPS coordinate directly. It measures range from a spacecraft toward the reflecting ocean surface. The SEASAT altimeter’s stated objectives included measuring mean sea level and supporting refinement of the geoid, while NASA’s mission history describes the radar altimeter as one instrument within a broader ocean-observing spacecraft.

To turn a radar range into a useful geodetic height, the analyst must also know where the spacecraft was. That is why orbit determination is a separate dependency.

West’s 1981 SEASAT processing report describes measurements from the radar altimeter being combined with a precise Doppler orbit, corrected for atmospheric and environmental effects, and reduced to along-track geoid heights and vertical deflections with an adaptive Kalman smoother based on a third-order Markov process.

The distinction matters. Raw radar-altimeter observations are instrument measurements. An orbit solution places the satellite relative to a reference system. Atmospheric propagation, tides and other environmental effects alter the relationship between the observed sea surface and the geodetic quantity being estimated. Corrected observations can then be reduced to along-track geoid-related heights and smoothed products. Those products can subsequently be compared with a geoid calculated from gravity-field coefficients and used in an ellipsoid-fitting problem.

That chain is the evidentiary boundary that public retellings often erase. West’s contribution included mathematical and computational systems that performed important parts of this reduction and estimation. The spacecraft, radar hardware, tracking infrastructure, gravity coefficients and upstream mission observations were not the work of one individual.

Geoid, ellipsoid and coordinate frame are different layers

Three concepts are often compressed into “the shape of the Earth,” although they perform different jobs.

The geoid is an equipotential surface of Earth’s gravity field that approximates mean sea level. It is irregular because gravity reflects Earth’s mass distribution and rotation. The reference ellipsoid is a smooth mathematical surface, generated from a simple geometric form and chosen as a tractable approximation to Earth. The Army Geospatial Center explicitly distinguishes the two and notes that GPS-derived positions are relative to an ellipsoid while a geoid model connects that geometric reference to gravity-related elevations.

A terrestrial coordinate frame adds another layer. It supplies an origin and axes, along with the practical realization that makes coordinates operational. IS-GPS-200M specifies a WGS 84 Earth-centred, Earth-fixed coordinate system for satellite antenna-phase-centre positions: the origin is Earth’s centre of mass, while the x, y and z axes have defined orientations.

That is not synonymous with the geoid, and it is more than an ellipsoid radius.

Satellite orbit determination is another problem again: estimating where a transmitting spacecraft is. Receiver positioning then estimates the user’s state from radio measurements whose interpretation depends on satellite ephemerides, clocks, constellation geometry, propagation conditions and receiver processing. Precision surveying can further depend on reference stations, correction products and other infrastructure.

Collapsing these layers makes one historical ellipsoid parameter appear to determine end-user accuracy. It does not.

West’s Dahlgren work was a sequence of systems

The better way to understand West’s record is chronologically.

In 1979, she published on smoothing GEOS-3 radar-altimeter data at the Naval Surface Weapons Center. The paper says a Wiener smoother based on a third-order Markov process had been used, and that a Kalman smoother using the same third-order Markov statistics replaced it because it more closely satisfied the conditions of the altimetry data.

That is evidence of method comparison and system evolution, not one frozen algorithm.

The 1981 SEASAT report moves from filter comparison to a broader reduction system. Its abstract describes altimeter measurements combined with precise Doppler orbit information, environmental and atmospheric corrections, and reduction to along-track geoid heights and vertical deflections.

The 1982 paper then uses altimetric-geoid products for an ellipsoid solution.

Four years later, West’s 1986 GEOSAT specification documented another system at Dahlgren for reducing radar-altimeter measurements to best-estimated along-track geoid heights and vertical deflections. The specification identifies inputs including precise orbit information, environmental data, tidal constants and sensor records; it also lays out organizational interfaces rather than presenting the computation as a self-contained formula. NASA’s GEOSAT record places that later satellite in its own mission history, beginning with its 1985 launch and subsequent altimetric operations.

The continuity is expertise: geodetic data reduction, smoothing, gravity-related modelling, software design and documentation. The continuity is not one immutable routine handed unchanged from GEOS-3 to SEASAT to GEOSAT and then into contemporary GPS.

The contribution map is institutional, not solitary

West’s authorship deserves to remain specific. The Library of Congress finding aid says she planned, implemented, documented and supervised complex computer programs involving gravity coefficients, vertical deflection, geoid coordinates and satellite orbital products. It records work using GEOS-3, SEASAT and GEOSAT data and describes increasing supervisory responsibilities across her Dahlgren career.

The Air Force’s 2018 Space and Missile Pioneers notice separately recognizes her decades of geodetic modelling work. Dahlgren’s own historical material places her employment there from 1956 and records her transition from manual calculation to computer programming.

But attribution becomes less accurate when recognition is converted into sole invention.

SEASAT was a NASA ocean-observing mission. Its radar-altimeter documentation identifies Johns Hopkins Applied Physics Laboratory as the contractor for the instrument. Precise orbit determination and gravity modelling came from wider technical programs. DoD geodetic systems supplied reference frameworks and coefficient sets. West’s own later GEOSAT documentation shows responsibilities distributed among Dahlgren, APL, mapping organizations, naval entities and downstream users.

That map does not dilute West’s contribution. It locates it.

In infrastructure, authorship is often clearest when the interfaces remain visible: who designed the mission, who built the sensor, who estimated the orbit, who generated the gravity model, who developed and supervised reduction software, who defined the reference system, and who later operationalized positioning. Turning that network into a single-inventor story removes much of the mechanism that made the work valuable.

WGS 84 and GPS are later receipts

The Virginia General Assembly’s 2026 memorial resolution says West’s mathematical models “still reside in every GPS-enabled device.” As commemoration, that sentence expresses technological legacy. As engineering evidence, it is too broad to demonstrate literal inheritance of the 1982 WGS 72/GEM10B numerical solution.

Modern GPS documentation supplies a separate receipt. IS-GPS-200M specifies satellite positions in WGS 84 Earth-centred, Earth-fixed coordinates. The GPS geodetic-reference material describes WGS 84 in relation to the International Terrestrial Reference Frame and shows the reference system as an actively realized geodetic framework rather than a synonym for West’s earlier WGS 72/GEM10B calculation.

The distinction is important. A current receiver does not need to contain West’s 1982 result unchanged for her work to belong to GPS’s technical lineage. Legacy can propagate through methods, institutional capability, geodetic modelling, software practice and accumulated improvements.

The historically supportable claim is that West’s documented satellite-geodesy work contributed to the technical foundation from which precise global positioning developed. The stronger literal claim—that one 1982 ellipsoid solution is the permanent numerical core of every receiver—requires evidence the memorial resolution does not provide and the GPS interface specification does not state.

Why this evidentiary boundary matters

Technology procurement and resilience depend on decomposing systems into dependencies. A positioning service can degrade because of satellite availability, clock or ephemeris problems, radio interference, propagation effects, receiver implementation, correction-service outages or reference-station failures. None of those risks can be understood by pointing to a single ellipsoid parameter.

Standards governance has the same problem. Reference systems are realized and maintained; missions and processing systems change; data products acquire different correction models and quality controls. Treating a historical calculation as an eternal technical artifact obscures the institutions responsible for keeping the system coherent.

Authorship claims also affect how infrastructure is understood and funded. If public history turns networked engineering into a one-person invention, the laboratories, mission teams, standards custodians, software maintainers and measurement networks required to sustain capability disappear from view.

A more exact account can credit West more securely: not as the sole inventor of GPS, but as an author, programmer, system designer, documenter and project leader whose satellite-geodesy work addressed difficult Earth-modelling problems on which later positioning systems depended. The Library of Congress record is especially useful here because it preserves both her individual responsibility and the wider program context.

Sources