Skip to content

WGS-84: Modern Geodetic Datum

The detailed constant comparison table lives on the WGS-72 page, which is the canonical location for this data in the SGP4 theory archive. In brief: the two systems differ by 2 meters in semi-major axis, approximately 10710^{-7} in J2J_2, and use gravitational models of dramatically different resolution (degree 12 vs. degree 360).

The SGP4/SDP4 propagators are mathematically coupled to WGS-72. This coupling is not an oversight — it is structural.

  1. TLE generation uses WGS-72. The orbit determination process at CSpOC (Combined Space Operations Center) fits observed positions using SGP4 with WGS-72 constants. The mean elements in a TLE are parameters of that specific model.

  2. J2J_2 perturbation terms are calibrated to WGS-72. SGP4 models the secular and periodic effects of Earth’s oblateness using J2=1.0826158×103J_2 = 1.0826158 \times 10^{-3}. The higher-precision WGS-84 value would not correctly cancel the fitting residuals.

  3. Consistency, not accuracy, is the goal. TLEs are not truth data. They are a compact representation valid only when propagated with the matching algorithm and constants.

  4. Vallado’s standardization. The 2006 “Revisiting Spacetrack Report Number 3” paper explicitly states that WGS-72 constants must be used with standard TLEs. The WGS-84 constant mode exists solely for research comparison.

Once SGP4 produces an Earth-Centered Inertial (ECI) position vector using WGS-72, Craft must convert that position to a ground-observer-relative frame. This is where WGS-84 enters the pipeline.

  • Observer location. GPS coordinates are WGS-84 geodetic coordinates. The observer’s latitude, longitude, and altitude must be converted to ECEF (and then ECI) using the WGS-84 ellipsoid.
  • Topocentric frame. Computing azimuth, elevation, and range from observer to satellite requires both the satellite’s ECI position (from SGP4 with WGS-72) and the observer’s ECI position (from WGS-84 geodetic-to-ECEF conversion).
  • Sub-satellite point. Converting a satellite’s ECEF position to geodetic latitude/longitude for ground track plotting uses WGS-84 ellipsoid parameters.
  • CesiumJS visualization. CesiumJS uses WGS-84 natively. All positions rendered on the 3D globe are WGS-84 Cartesian or cartographic coordinates.
TLE --> SGP4 (WGS-72) --> ECI position --> coordinate rotation --> ECEF (WGS-84) --> geodetic/topocentric

The satellite.js library (used by Craft’s web frontend for client-side propagation) uses WGS-72 constants internally for SGP4, matching the CSpOC orbit determination process. The Skyfield library (used by Craft’s API backend) handles the appropriate reference frame transformations automatically.

The position difference between WGS-72 and WGS-84 is on the order of 1—2 meters. For satellite tracking applications where TLE accuracy itself is 100 m to 1+ km (see Kelso 2007), this difference is negligible. But using the correct constants in each stage prevents systematic bias from accumulating.

The following chapters were transcribed from the 175-page source document:

FileDescription
Defining parametersAll defining parameters, derived geometric and physical constants, coordinate conversion formulas, and WGS-72-to-84 transformation
Machine-readable constantsJSON of all constants with values, units, uncertainties, and formulas

The following chapters are available in the source PDF but were not transcribed, as they are not directly needed for Craft’s coordinate transformation pipeline:

ChapterTitleContent
4WGS-84 Ellipsoidal Gravity FormulaDetailed gravity field, free-air corrections, Bouguer anomalies
5WGS-84 GeoidEGM96 geoid undulations, geoid-ellipsoid separation
6WGS-84 and Local Geodetic SystemsTransformation parameters for 100+ local datums
7Reference Frame TiesITRF relationships, tectonic plate motion
App. A—DVariousTest cases, datum shift tables, supplementary formulas