1 The question this series will answer

Kaplan and Hegarty introduce the ideal two-body equation of motion in the form

(1)

where

(2)

Equation (1) looks compact. The GPS broadcast ephemeris algorithm does not. A receiver encounters quantities such as

(3)

and the six harmonic coefficients

(4)

It then performs a sequence involving mean anomaly, eccentric anomaly, true anomaly, argument of latitude, harmonic corrections, orbital-plane coordinates, and a final transformation to Earth-fixed coordinates.

The central question of this series is therefore

(5)

The complete conceptual path is shown in Figure 1.

2 Start one step before Kaplan’s equation

The two-body equation is itself a consequence of Newton’s law of universal gravitation. If an Earth of mass M attracts a satellite of mass m, the gravitational force on the satellite is

(6)

where

(7)

Using Newton’s second law,

(8)

gives

(9)

The satellite mass cancels:

(10)

Defining

(11)

produces Eq. (1).

This cancellation is physically important. In the ideal Newtonian two-body model, the orbital acceleration of a satellite is independent of its own mass. A one-kilogram test mass and a GPS spacecraft placed at the same position with the same velocity follow the same trajectory.

3 Why does a three-dimensional orbit require six constants?

A common sentence in orbital mechanics is that the two-body solution contains “six constants of integration” or “six integrals of motion.” Before deriving them, it is useful to understand why the number six must appear.

Write Eq. (1) in Cartesian components:

ẍ	= −,	(12)
ÿ	= −,	(13)
z	= −.	(14)

These are three coupled second-order differential equations. Equivalently, define the six-dimensional state

(15)

Then the dynamics become six first-order differential equations,

(16)

A unique trajectory is selected by six independent scalar initial conditions, for example

(17)

This is the simplest reason that a general three-dimensional orbit needs six independent parameters. The classical orbital elements provide a more geometrically meaningful set of six numbers describing the same solution.

3.1 A terminology warning: constants of integration versus conserved quantities

The phrase “six integrals of motion” can be confusing. It does not mean that one simply writes down six unrelated scalar conservation laws. The two-body problem contains vector invariants such as angular momentum and the eccentricity vector, but their components are constrained and are not all independent. The clean statement for this series is:

The general three-dimensional solution of the two-body equations contains six independent constants. Those six degrees of freedom can be represented by a Cartesian state at an epoch or by six orbital elements.

Later lessons will derive the useful conserved quantities and show exactly how they lead to the conic orbit and the classical elements.

4 The first physical integral: angular momentum

The acceleration in Eq. (1) is always parallel to r. Therefore the torque per unit mass about Earth’s center is zero:

(18)

Using

(19)

we obtain

(20)

Hence the specific angular momentum vector

(21)

is constant in the ideal two-body problem.

This immediately establishes a major geometric fact: because both r and v remain perpendicular to the fixed vector h, the trajectory remains in one fixed plane through the attracting center. The three-dimensional problem has therefore acquired a two-dimensional orbital plane.

GPSORB03 will develop this result in detail.

5 Energy gives the size of the orbit

Taking the scalar product of Eq. (1) with v leads to conservation of specific mechanical energy,

(22)

For an elliptic orbit this constant is related to the semimajor axis by

(23)

Equation (23) is the origin of the first classical orbital element, a. It is also the basis of the vis-viva relation,

(24)

GPSORB04 will derive these equations rather than assume them.

6 The eccentricity vector gives shape and periapsis direction

Another invariant of the ideal two-body problem is the eccentricity vector,

(25)

Its magnitude

(26)

is the orbital eccentricity, and its direction points toward periapsis.

Combining h and e leads to the polar equation of a conic,

(27)

For an ellipse,

(28)

Thus Newton’s vector differential equation produces an ellipse without inserting an ellipse by assumption.

This is one of the main derivations of the series: Newtonian dynamics → conserved quantities → conic geometry.

7 The six classical orbital elements

For an elliptic orbit, a convenient set of six independent parameters is

(29)

Their roles separate naturally:

For geometry, true anomaly ν₀ can temporarily replace M₀. For time propagation, mean anomaly is more convenient, so GPS ultimately uses M₀.

The conversion

(30)

will be derived in both directions later in the series.

8 Why three different anomalies appear

The satellite’s location along an elliptic orbit is described using three related angles.

8.1 True anomaly

The true anomaly ν is the geometrically obvious angle measured in the orbital plane from periapsis to the satellite position vector. It appears directly in the conic equation (27).

8.2 Eccentric anomaly

The eccentric anomaly E is an auxiliary angle associated with a circle surrounding the ellipse. It is not merely a geometric curiosity; it provides a convenient parameterization

xp	= a(cos E − e),	(31)
yp	= a sin E.	(32)

From this parameterization one obtains

(33)

8.3 Mean anomaly

Mean anomaly is defined so that it increases uniformly in the ideal two-body problem:

(34)

The constant rate n is the mean motion,

(35)

Kepler’s equation connects mean anomaly to eccentric anomaly:

(36)

The relation between eccentric and true anomaly can be written as

(37)

These equations form the time-propagation heart of the GPS ephemeris algorithm. GPSORB06 will derive them from area swept by the radius vector and the geometry of the ellipse.

9 Why Kepler’s equation needs an iterative solution

Equation (36) contains E both algebraically and inside a sine function. There is no elementary algebraic rearrangement that isolates E. A receiver therefore solves

(38)

Newton’s method uses

(39)

and gives

(40)

This is not an arbitrary numerical trick added to GPS. It follows directly from applying Newton-Raphson iteration to Kepler’s equation. IS-GPS-200 gives this iteration explicitly in its legacy LNAV user equations.

10 The ideal six elements are not constant for a real GPS satellite

Equation (1) treats Earth as a point mass and assumes that no other forces act. A real satellite instead experiences a more complete acceleration model schematically represented as

(41)

Under these perturbations, the six Keplerian parameters can still be used to describe the instantaneous orbit, but they no longer remain exact constants.

A useful conceptual distinction is therefore

(42)

The legacy GPS broadcast quantities are Keplerian in appearance, but the control/space segment obtains operational values by fitting the broadcast model to a propagated reference trajectory over a finite interval. This is why it is dangerous to interpret every broadcast correction coefficient as if it were a fundamental constant derived directly from one perturbing force.

11 From six Keplerian elements to the broadcast ephemeris

The legacy LNAV broadcast model can be organized into three groups, as shown in Figure 2.

11.1 Kepler-like core

The core contains quantities closely associated with the six classical elements:

(43)

The broadcast quantity is rather than A itself for legacy LNAV.

11.2 Reference and secular terms

The receiver also uses

(44)

Here the ephemeris reference-time parameter identifies the epoch of applicability, the mean-motion correction modifies the nominal Keplerian mean motion, the node-rate parameter describes fitted node evolution, and IDOT describes inclination-rate behavior.

11.3 Second-harmonic correction coefficients

Six coefficients apply periodic corrections. The pair C_uc,C_us corrects argument of latitude, C_rc,C_rs corrects orbital radius, and C_ic,C_is corrects inclination. The corresponding correction structure is

(45)

(46)

(47)

The appearance of 2Φ_k is connected to the second-harmonic structure that naturally arises in important nonspherical-gravity effects. Later lessons will develop that physics while keeping separate the distinction between physical perturbation theory and the fitted coefficients actually transmitted in the navigation message.

12 The receiver algorithm that we will derive

The legacy LNAV position algorithm can be viewed as a pipeline rather than as a collection of unrelated formulas.

First recover the semimajor axis:

(48)

Then form the nominal mean motion

(49)

and the elapsed time from the ephemeris epoch

(50)

After the GPS-week crossover convention is applied, compute

(51)

and

(52)

Solve Kepler’s equation

(53)

for E_k, then obtain ν_k. Form the uncorrected argument of latitude

(54)

Apply Eq. (47), followed by

uk	= Φk + δuk,	(55)
rk	= A(1 − e cos Ek) + δrk,	(56)
ik	= i0 + IDOT tk + δik.	(57)

The in-plane coordinates become

x′k	= rk cos uk,	(58)
y′k	= rk sin uk.	(59)

The corrected longitude of the ascending node is

(60)

where the Earth-rate quantity is the Earth rotation rate used by the GPS user algorithm.

Finally,

xk	= x′k cos Ωk − y′k cos ik sin Ωk,	(61)
yk	= x′k sin Ωk + y′k cos ik cos Ωk,	(62)
zk	= y′k sin ik.	(63)

The output is the satellite antenna phase-center position in Earth-fixed coordinates.

Every equation in this pipeline will be derived in subsequent articles.

13 Why the final coordinates are Earth-fixed

Classical orbital mechanics is most naturally developed in an inertial reference frame. GPS users, however, need satellite positions relative to the rotating Earth. This introduces a second layer of physics.

If an inertial frame and an Earth-fixed frame rotate relative to one another with angular velocity Ω_e, derivatives obey

(64)

Applying this relation twice generates the familiar Coriolis and centrifugal terms. Those terms reappear explicitly when IS-GPS-200 gives Earth-fixed satellite acceleration equations.

Thus the series begins with inertial Newtonian mechanics but ends in a rotating Earth-fixed navigation frame. That connection is not incidental; it is the same rotating-frame physics that appears in inertial navigation mechanization.

14 What lies beyond Kaplan’s position table

Kaplan’s Table 2.3 gives the core route to the satellite ECEF position. The current IS-GPS-200N legacy LNAV user equations extend the chain further by presenting equations for velocity and acceleration.

Differentiating Kepler’s equation gives

(65)

Differentiating the anomaly relationship produces the true-anomaly rate, and successive chain-rule applications give

(66)

and eventually the Earth-fixed velocity components.

The acceleration equations bring the discussion back to force models. IS-GPS-200 includes a J₂ oblateness contribution together with rotating-Earth terms. The conceptual loop of the course is therefore

(67)

15 A GPS-scale sanity check

Before working through the full derivation, it is useful to estimate the magnitude of the quantities involved. Take a representative GPS semimajor axis

(68)

and the GPS user value

(69)

Then Eq. (35) gives approximately

(70)

The corresponding ideal period is

(71)

and a circular-orbit speed at this radius is approximately

(72)

These estimates will serve as useful order-of-magnitude checks in later numerical work.

16 Scope: legacy LNAV first, then modern extensions

IS-GPS-200 contains more than one navigation-message family. This series first follows the legacy LNAV structure because it corresponds closely to Kaplan’s Table 2.3 and to the familiar parameters , Δn, and the six harmonic correction coefficients. In IS-GPS-200N this is the Table 20-IV family of broadcast navigation user equations.

Modernized message structures, including the CNAV equations in the Table 30-II family, use related mechanics but introduce different parameterizations and additional time-dependent terms. Once the legacy derivation is complete, the modern equations can be treated as an extension rather than mixing two parameter sets during the foundational development.

Clock correction parameters such as a_f0, a_f1, and a_f2 are also part of the navigation data, but they are not orbital elements. They belong to the satellite-clock model and will be treated separately from this orbit-mechanics sequence.

17 Series roadmap

Separate companion articles labeled E1, E2, and so on can contain hand calculations, MATLAB/Python exercises, and numerical verification without interrupting the main derivation sequence.

18 Notation to keep fixed throughout the series

Angles will be treated in radians inside the mathematics unless a broadcast-message scaling explicitly uses semicircles. Frame superscripts will be added whenever the frame is not obvious from context.

19 What should be understood before moving to GPSORB01

The important result of this introductory article is not a single equation. It is the architecture of the problem.

A satellite trajectory is fundamentally a solution of a differential equation. The three-dimensional second-order equation requires six independent scalar conditions. Classical orbital elements re-parameterize those six degrees of freedom in a way that exposes the geometry of the conic. Kepler’s equation supplies the time evolution along that conic. Real GPS motion is not exactly Keplerian, so the operational navigation message uses a fitted Kepler-like model with secular and periodic correction parameters. The receiver evaluates that model at the desired transmit time and finally expresses the satellite position in Earth-fixed coordinates.

In compact form,

(73)

GPSORB01 now returns to the beginning and derives Eq. (1) carefully from Newtonian mechanics, including the assumptions hidden inside the phrase “two-body problem.”

References

References