Orbital perturbations — J2 effect (oblateness), derivation of nodal precession

The Earth is not a perfect sphere. It bulges at the equator. That tiny bulge silently drags satellite orbits around, rotating the whole orbital plane. This note derives exactly how fast.

The Big Picture (80/20)

The 20% you must own: the equatorial bulge adds a small extra potential term $J_2$. Its time-averaged torque leaves orbit size ($a$), shape ($e$), and tilt ($i$) alone, but slowly rotates two orientation angles:

• the right ascension of ascending node $\Omega$ (the orbit plane swivels) → nodal precession
• the argument of perigee $\omega$ (the ellipse rotates in its own plane)

The headline result: $\dot{\Omega} = -\frac{3}{2}\, n\, J_2 \left(\frac{R_E}{p}\right)^2 \cos i$

[!intuition] WHY does a bulge cause precession?

A perfect sphere pulls exactly toward its center → gravity is central → the orbit plane is fixed (angular momentum $\vec L$ is conserved).

The equatorial bulge adds extra mass around the equator. When a satellite is above or below the equatorial plane, the bulge pulls it slightly back toward the equator — a component of force not pointing at the center. This off-axis pull exerts a torque on the orbit. Because the satellite constantly sweeps around, the torque averages into a steady drift: the orbital plane cannot tumble randomly, so instead it precesses like a spinning top under gravity.

[!definition] The J2 coefficient

$J_2$ is the largest coefficient in the multipole (spherical-harmonic) expansion of Earth's gravitational potential. It measures oblateness (equatorial bulge). For Earth $J_2 \approx 1.0826\times10^{-3}$ — small, but it dominates all other perturbations for low orbits.

$U(r,\phi) = -\frac{\mu}{r}\left[1 - J_2\left(\frac{R_E}{r}\right)^2 \frac{3\sin^2\phi - 1}{2} - \dots \right]$

• $\mu = GM_E$, $R_E$ = equatorial radius, $\phi$ = geocentric latitude.
• $P_2(\sin\phi)=\tfrac{1}{2}(3\sin^2\phi-1)$ is the Legendre polynomial of degree 2.
• The leading $-\mu/r$ is the point-mass (Keplerian) term.

[!formula] Derivation of nodal precession $\dot\Omega$

Step 1 — The disturbing function $R$

Split potential into Kepler + perturbation. The disturbing function is the extra piece: $R = \frac{\mu J_2 R_E^2}{2 r^3}\left(3\sin^2\phi - 1\right)$ Why this step? Lagrange's planetary equations feed on $R$ = potential energy per mass above the point-mass baseline.

Step 2 — Express latitude via orbital geometry

For a satellite at argument of latitude $u=\omega+\nu$ in an orbit of inclination $i$: $\sin\phi = \sin i \,\sin u$ Why this step? The satellite's latitude depends only on how far it is around the orbit ($u$) and how tilted the orbit is ($i$). This is straight spherical trigonometry of the orbital triangle.

So: $R = \frac{\mu J_2 R_E^2}{2 r^3}\left(3\sin^2 i\,\sin^2 u - 1\right)$

Step 3 — Average over one orbit

Perturbations per revolution are tiny; only the secular (non-repeating) drift matters. Average over one period. Two facts: $\langle \sin^2 u\rangle = \tfrac12,\qquad \left\langle \frac{1}{r^3}\right\rangle = \frac{1}{a^3(1-e^2)^{3/2}}=\frac{1}{p^3/... } \;\Rightarrow\; \frac{1}{a^3(1-e^2)^{3/2}}$ Why this step? The short-period wiggles cancel out over a full loop; the leftover constant part is what accumulates orbit after orbit.

Averaging (with careful $\nu$-weighting, using $r^2\,d\nu = \sqrt{\mu p}\,dt$) gives the mean disturbing function: $\bar R = \frac{n^2 J_2 R_E^2}{4(1-e^2)^{3/2}}\left(3\sin^2 i - 2\right)\cdot(\text{const})$ where $n=\sqrt{\mu/a^3}$ is the mean motion and $p=a(1-e^2)$.

Step 4 — Lagrange's planetary equation for $\Omega$

$\frac{d\Omega}{dt} = \frac{1}{n a^2\sqrt{1-e^2}\,\sin i}\,\frac{\partial \bar R}{\partial i}$ Why this step? Node motion is driven by how the averaged energy changes when you tilt the orbit — that's the meaning of $\partial \bar R/\partial i$.

Since $\bar R \propto (3\sin^2 i - 2)$, we have $\partial\bar R/\partial i \propto 6\sin i\cos i$. The $\sin i$ cancels, leaving $\cos i$:

$\boxed{\;\dot\Omega = -\frac{3}{2}\,n\,J_2\left(\frac{R_E}{p}\right)^2 \cos i\;}$

with $p=a(1-e^2)$, $n=\sqrt{\mu/a^3}$.

Companion result (same machinery, $\partial\bar R/\partial e$... actually $\partial/\partial i$ for $\omega$):

$\dot\omega = \frac{3}{4}\,n\,J_2\left(\frac{R_E}{p}\right)^2\left(5\cos^2 i - 1\right)$

[!intuition] Reading the two results

Nodal precession $\dot\Omega \propto -\cos i$:

• $i=0°$ (equatorial): $\cos i=1$, max westward ($\dot\Omega<0$).
• $i=90°$ (polar): $\cos i=0$, no precession.
• $i>90°$ (retrograde): $\cos i<0$, so $\dot\Omega>0$ — the plane drifts eastward. This is the trick behind Sun-synchronous orbits: pick $i\approx 98°$ so $\dot\Omega = +0.9856°/\text{day}$, exactly Earth's orbit around the Sun → the orbit plane keeps a fixed angle to the Sun.

Apsidal rotation $\dot\omega \propto (5\cos^2 i - 1)$:

• Zero when $\cos^2 i = 1/5 \Rightarrow i = 63.4°$ (or $116.6°$). This critical inclination freezes perigee — used by Molniya orbits so the apogee stays over the northern hemisphere.

[!example] Worked: precession of the ISS

Given: $a\approx 6.79\times10^6$ m, $e\approx 0$, $i=51.6°$, $\mu=3.986\times10^{14}$, $R_E=6.378\times10^6$, $J_2=1.0826\times10^{-3}$.

1. Mean motion: $n=\sqrt{\mu/a^3}=\sqrt{3.986\times10^{14}/(6.79\times10^6)^3}=1.13\times10^{-3}$ rad/s. Why? Sets the base timescale; precession scales with $n$.
2. $p=a(1-e^2)=a$ (nearly circular). So $(R_E/p)^2=(6.378/6.79)^2=0.882$. Why? This is the "how deep in the bulge field" factor — larger orbits feel less J2.
3. $\dot\Omega=-\tfrac32(1.13\times10^{-3})(1.0826\times10^{-3})(0.882)\cos 51.6°$ $=-\tfrac32(1.13\times10^{-3})(1.0826\times10^{-3})(0.882)(0.622)$ $\approx -1.01\times10^{-6}$ rad/s.
4. Convert: $\times 86400\times \tfrac{180}{\pi} \approx -5.0°/\text{day}$. ✔ (ISS node regresses ~5°/day, matching reality.)

[!example] Worked: designing a Sun-synchronous orbit

Want $\dot\Omega=+1.991\times10^{-7}$ rad/s ($=0.9856°/$day). For $a=7.2\times10^6$ m, $e=0$:

1. $n=\sqrt{\mu/a^3}=1.033\times10^{-3}$ rad/s. Why? Needed in the formula.
2. Solve $\cos i = \dfrac{\dot\Omega}{-\tfrac32 n J_2 (R_E/p)^2}$.
3. $(R_E/p)^2=(6.378/7.2)^2=0.785$. Denominator $=-\tfrac32(1.033\times10^{-3})(1.0826\times10^{-3})(0.785)=-1.317\times10^{-6}$.
4. $\cos i = 1.991\times10^{-7}/(-1.317\times10^{-6})=-0.151 \Rightarrow i=98.7°$. ✔ Why negative? A prograde-plus effect needs retrograde inclination.

[!mistake] Steel-manning the classic errors

"J2 changes the orbit's size and energy." Why it feels right: it's an extra force, and forces usually change energy/speed. The fix: the secular (averaged) parts of $\dot a=\dot e=0$; J2 conserves orbital energy on average and only rotates orientation angles $\Omega,\omega$ (plus a small $\dot M$). Size and shape only wobble with short periods that cancel each orbit.

"Polar orbits precess fastest." Why it feels right: polar orbits cross the equatorial bulge most steeply. The fix: the torque about the polar axis needs the orbit to have angular momentum along that axis — that's the $\cos i$. At $i=90°$, $\cos i=0$: no nodal precession at all.

"The sign doesn't matter." Why it feels right: precession is precession. The fix: the sign of $\dot\Omega$ (from $-\cos i$) is exactly what makes retrograde Sun-synchronous orbits possible. Get the sign wrong and your satellite design fails.

"Use $a$ everywhere instead of $p$." Why it feels right: they're equal for circular orbits. The fix: for eccentric orbits use $p=a(1-e^2)$; the $(1-e^2)^2$ factor hidden inside significantly speeds precession for eccentric orbits.

[!recall]- Feynman: explain to a 12-year-old

Imagine the Earth is a slightly squished ball — fatter around its belly. A satellite is like a marble looping around it. Because the belly bulge pulls the marble a bit sideways whenever it's above or below the equator, the whole loop can't stay pointing the same way — it slowly turns, like the way a spinning top slowly circles instead of standing straight. How fast it turns depends on how tilted the loop is: a straight up-and-down (over the poles) loop doesn't turn at all, while a flat loop near the equator turns the most. Space engineers use this on purpose — they pick the exact tilt so the loop turns just enough to always face the Sun.

[!mnemonic]

"NOde precession is a COsine; PErigee is Five-COs-squared."

• $\dot\Omega \sim \cos i$ → "Node—Cosine"
• $\dot\omega \sim 5\cos^2 i -1$ → "Perigee—Five"
• Both carry the minus/plus and $\tfrac32 n J_2 (R_E/p)^2$ front factor. Node is $-\tfrac32$, perigee is $+\tfrac34$.

Flashcards

Connections

• Two-body problem & Keplerian orbits — the unperturbed baseline this modifies.
• Lagrange planetary equations — the machinery converting $\bar R$ into element rates.
• Spherical harmonics of gravity fields — where $J_2, J_3, C_{nm}, S_{nm}$ come from.
• Sun-synchronous orbits — direct engineering use of $\dot\Omega$.
• Molniya & critical inclination orbits — application of $\dot\omega=0$.
• Angular momentum & torque — the deep reason precession happens.
• Legendre polynomials — the $P_2$ in the potential.

Concept Map

Hinglish (regional understanding)