Equatorial bulge and the figure of the Earth

In 1672 the astronomer Jean Richer sailed from Paris to Cayenne, five degrees north of the equator, carrying a pendulum clock that kept perfect time at home. In Cayenne it lost about two and a half minutes a day. Nothing was wrong with the clock — gravity itself was weaker there. Huygens and Newton both saw what that meant: a spinning Earth cannot be a perfect sphere. The centrifugal effect of the rotation weakens effective gravity most at the equator and lets matter pile up there, so the planet must bulge — and the size and shape of that bulge became the eighteenth century's most famous measurement war, pitting Newton's oblate prediction against the Cassini family's prolate one, and sending expeditions to Lapland and Peru to settle it with surveying chains. The physics behind the fight is a single, beautiful equilibrium condition. A rotating, self-gravitating fluid settles into the shape whose free surface is an equipotential of the combined gravitational and centrifugal potential — any tilt of the surface away from a level set would drive flow until the tilt is gone. From that one statement follow the flattening of the Earth ( $f \approx 1/298$ , a 21.4 km difference between equatorial and polar radii), Clairaut's elegant theorem linking shape to the pole-to-equator variation of $g$ , the visibly squashed profiles of Jupiter and Saturn, and even the maximum spin a fluid body can survive before it sheds mass at the equator. Mastering it turns "the Earth is slightly flat" from a factoid into a calculation you can do on one page.

Key takeaways

A rotating self-gravitating fluid cannot be a perfect sphere: the centrifugal effect adds an outward push that grows with distance from the rotation axis, so the equilibrium surface bulges at the equator and flattens at the poles.
Equilibrium requires the total effective potential (gravitational plus centrifugal) to be constant on the free surface — an equipotential — because any tangential gradient would drive fluid flow.
For the Earth the dimensionless smallness parameter is $m=\omega^2 a/g\approx 1/289$ ; the flattening $f=(a-c)/a$ is observed to be $\approx 1/298.3$ , close to the simple Maclaurin/Clairaut estimate $f\approx \tfrac54 m$ minus a correction for mass redistribution.
Clairaut's theorem links the geometric flattening $f$ to the equator-to-pole variation of surface gravity: $f+\beta=\tfrac52 m$ , independent of the internal density profile to first order.
The same balance governs Jupiter ( $f\approx 0.065$ ), Saturn ( $f\approx 0.098$ ), and the existence of a maximal rotation rate (Maclaurin/Jacobi sequences) beyond which an axisymmetric fluid figure becomes unstable.
Key idea: The free surface of a rotating self-gravitating fluid is an equipotential of the effective potential $\Phi_{\mathrm{eff}}=\Phi_{\mathrm{grav}}-\tfrac12\omega^2\rho^2$ ; solving this self-consistently with the gravitational field of the deformed body gives the oblate equilibrium figure.

The core insight

The free surface of a rotating self-gravitating fluid is an equipotential of the effective potential $\Phi_{\mathrm{eff}}=\Phi_{\mathrm{grav}}-\tfrac12\omega^2\rho^2$ ; minimizing this with the gravitational field of the deformed body gives the oblate equilibrium figure.

Where this lives today

The reference figure that came out of this physics is now baked into every phone. GPS positions are computed on the WGS-84 ellipsoid ( $a=6378.137$ km, $1/f=298.257$ ) — a direct descendant of Clairaut's equilibrium theory fitted with satellite data — and "altitude" in your maps app means height above that mathematical spheroid, corrected toward the geoid, the true $\Phi_{\mathrm{eff}}$ equipotential at mean sea level, which undulates by up to $\pm100$ m over mantle density anomalies. The bulge also steers spacecraft: Earth's $J_2$ makes satellite orbit planes precess, and mission designers exploit it — Sun-synchronous imaging satellites choose an inclination near $98°$ so that $J_2$ precession turns their orbit plane exactly $360°$ per year, keeping the lighting constant beneath them. The sharpest modern twist is that the figure of the Earth is now a climate observable. The twin GRACE-FO satellites, tracking each other to micrometres, watch $J_2$ and the geoid change month by month as ice sheets melt and groundwater shifts: mass moving from the poles toward the equator has measurably increased $J_2$ since the 1990s, reversing the slow post-ice-age decline as the mantle rebounds. The eighteenth century needed frozen surveyors in Lapland to get one digit of $f$ ; today two satellites read the planet's changing waistline every thirty days — and what they measure is still, line for line, the equipotential condition of equation (2).

The physical picture

Imagine spinning a water balloon. Water near the equator is being whirled around in big circles, so it "wants" to fly outward; water near the poles barely moves. The surface adjusts until its slope everywhere exactly cancels the leftover outward tendency — the surface becomes a level set of the effective potential, like the tilted surface of water in a rotating bucket, but wrapped around a sphere.

Figure

The bulge as an equilibrium — effective gravity must be perpendicular to the surface everywhere

Why an ellipsoid and not some other shape? To leading order the centrifugal potential $-\tfrac12\omega^2 r^2\sin^2\theta$ is a pure quadrupole ( $\propto P_2(\cos\theta)$ plus a constant). A quadrupole forcing produces a quadrupole deformation, and a sphere deformed by a $P_2$ harmonic is, to first order, an oblate spheroid. Higher harmonics ( $P_4$ , …) appear only at order $m^2$ . The bulge is also self-reinforcing. Once mass moves to the equator, that extra mass gravitationally attracts even more fluid toward the equator. This feedback makes the actual flattening larger than the naive "rigid sphere" estimate — captured by the factor $5/4$ in the homogeneous Maclaurin result versus the bare $1/2$ from centrifugal lifting alone. And the relevant smallness parameter $m=\omega^2 a/g$ is tiny for Earth ( $\approx 0.0035$ ) but large for gas giants and for any rapidly spinning star, where the figure departs strongly from a sphere and the linear theory breaks down.

The equilibrium condition and the shape

Everything begins with the potential a co-rotating fluid element actually feels. The centrifugal force in the rotating frame is conservative — it points away from the rotation axis with magnitude $\omega^2\rho$ at cylindrical radius $\rho$ — so it can be folded into the potential energy bookkeeping alongside gravity. (The centrifugal force itself, and why it takes this form, is the business of the centrifugal force article; here we only need its potential.)

\Phi_{\mathrm{eff}}(\mathbf r)=\Phi_{\mathrm{grav}}(\mathbf r)-\tfrac12\omega^2\rho^2,\qquad \rho=r\sin\theta \qquad (1)

where: $\Phi_{\mathrm{grav}}$ — gravitational potential of the (deformed) body; $\omega$ — angular velocity of rigid rotation; $\rho$ — cylindrical distance from the rotation axis; $r,\ \theta$ — spherical radius and colatitude (angle from the pole). The centrifugal term deepens the effective potential most where $\rho$ is largest — at the equator — which is precisely why fluid piles up there. For a static fluid in the rotating frame, pressure and effective gravity are the only players, and hydrostatic balance forces a striking geometric conclusion:

\nabla p=-\rho_m\,\nabla\Phi_{\mathrm{eff}}\ \Rightarrow\ \Phi_{\mathrm{eff}}=\text{const on the free surface} \qquad (2)

where: $p$ — pressure; $\rho_m$ — mass density. The surface of the ocean — and, over geological time, of the deformable solid Earth too — is a level surface of $\Phi_{\mathrm{eff}}$ . This single condition determines the planet's shape.

Derivation AWhy the free surface must be an equipotential

We are deriving the central boundary condition of the whole subject: that the outer surface of a rotating fluid body in equilibrium must be a surface of constant effective potential, equation (2). The strategy is pure hydrostatics — write the force balance for a fluid element at rest in the rotating frame and read off what it implies about the geometry of isobars. There is no approximation here; the result holds for any rotation rate. The point to watch is the logic direction: we do not assume the surface is level, we prove that any non-level surface could not be static.

Step 1. In the rotating frame the fluid is at rest, so the net force density on every element vanishes:

\mathbf f_{\text{net}}=-\nabla p-\rho_m\nabla\Phi_{\mathrm{eff}}=0

Why: with the centrifugal force absorbed into $\Phi_{\mathrm{eff}}$ by equation (1), the only remaining forces on a parcel are pressure and effective gravity; "the planet is not flowing" means they cancel exactly.

Step 2. Rearranging, $\nabla p=-\rho_m\nabla\Phi_{\mathrm{eff}}$ : the pressure gradient is everywhere parallel to $\nabla\Phi_{\mathrm{eff}}$ . Surfaces of constant pressure therefore coincide with surfaces of constant effective potential. Why: a gradient is perpendicular to its own level surfaces, so if the two gradients are parallel, the two families of level surfaces are the same family.

Step 3. The outer free surface is an isobar ( $p=0$ , or atmospheric). Hence it must be a surface $\Phi_{\mathrm{eff}}=\text{const}$ . If the surface had any tangential component of effective gravity, that force could not be balanced by pressure — pressure acts normal to an isobar — and fluid would flow sideways until the tilt was erased. Verified: this is the standard equilibrium-figure boundary condition (Landau–Lifshitz Fluid Mechanics §10; Stacey & Davis); it is the same physics that makes the water surface in a rotating bucket a paraboloid.

Two dimensionless numbers now organize everything. The first compares centrifugal lift to gravity; the second describes the shape itself:

m\equiv\frac{\omega^2 a}{g}=\frac{\omega^2 a^3}{GM},\qquad m_{\oplus}\approx\frac{1}{289}\approx3.46\times10^{-3} \qquad (3)

where: $a$ — equatorial radius; $g$ — surface gravitational acceleration; $M$ — total mass; $G$ — Newton's constant. For Earth the centrifugal acceleration at the equator is $\omega^2 a = 0.034\ \mathrm{m/s^2}$ — a third of a percent of gravity — so every shape effect will be of this order.

f\equiv\frac{a-c}{a},\qquad f_{\oplus}=\frac{1}{298.257}\approx3.353\times10^{-3} \qquad (4)

where: $a$ — equatorial (semi-major) radius, $6378.137$ km; $c$ — polar (semi-minor) radius, $6356.752$ km. The equator stands $21.4$ km farther from the center than the poles — two and a half Everests of extra radius, invisible to the eye on a globe but decisive for satellite orbits and for where "sea level" actually sits. The deformed body's own gravity field is captured, outside the body, by a multipole expansion whose leading correction is the quadrupole:

\Phi_{\mathrm{grav}}(r,\theta)=-\frac{GM}{r}\Big[1-J_2\Big(\frac{a}{r}\Big)^2 P_2(\cos\theta)+\dots\Big] \qquad (5)

where: $J_2$ — dimensionless quadrupole coefficient, $J_2=1.0826\times10^{-3}$ for Earth; $P_2(\cos\theta)=\tfrac12(3\cos^2\theta-1)$ — the Legendre polynomial that encodes "more mass around the waist". Equivalently $J_2$ measures the moment-of-inertia asymmetry:

J_2=\frac{C-A}{Ma^2} \qquad (6)

where: $C$ — moment of inertia about the polar (rotation) axis; $A$ — about an equatorial axis. A perfect sphere has $C=A$ and $J_2=0$ ; everything the outside world can learn about the bulge at quadrupole order is contained in this one number, which is why satellite geodesy reports it to nine digits.

Go deeper · Derivation B · 2 minNewton's canal argument, from the equipotential condition to the flattening

Derivation BNewton's canal argument, from the equipotential condition to the flattening

The goal is the first-order relation between shape, gravity field, and spin: $f=\tfrac32 J_2+\tfrac12 m$ , equation (7). The physical idea is Newton's famous thought experiment: bore two canals to the center of the Earth, one from the pole and one from the equator, and fill them with water. They meet at the center, where the pressure at the bottom of each column must be the same — but the equatorial column is lighter per unit length, because centrifugal lift subtracts from gravity there. Lighter fluid must stand taller to press equally hard, so the equatorial leg must be longer: the bulge. Our modern version simply demands that the effective potential (1), with the quadrupole field (5), take equal values at the pole and the equator of the surface $r(\theta)=a\,[1-f\cos^2\theta]$ . Watch the bookkeeping: $f$ , $J_2$ and $m$ are all first-order small, and we discard their products.

Step 1. Write the effective potential on the surface, keeping terms to first order in $J_2$ , $m$ , and $f$ :

\Phi_{\mathrm{eff}}=-\frac{GM}{r}\Big[1-J_2\frac{a^2}{r^2}P_2(\cos\theta)\Big]-\tfrac12\omega^2 r^2\sin^2\theta

Why we may stop at the quadrupole: the centrifugal forcing is itself a pure $P_2$ shape, so it excites $P_4$ and higher harmonics only at order $m^2$ .

Step 2. Evaluate at the two extremes. At the pole, $\theta=0$ : $P_2=1$ , $\sin\theta=0$ , $r=c=a(1-f)$ . At the equator, $\theta=\pi/2$ : $P_2=-\tfrac12$ , $\sin\theta=1$ , $r=a$ . Why these two points suffice: at first order the surface is exactly the $P_2$ shape $r(\theta)=a[1-f\cos^2\theta]$ , so matching the potential at pole and equator fixes the single coefficient $f$ ; every other point then matches automatically.

Step 3. Set $\Phi_{\mathrm{eff}}(\text{pole})=\Phi_{\mathrm{eff}}(\text{equator})$ and expand $1/r$ at the pole as $(1/a)(1+f)$ . The radius difference contributes $GMf/a$ , the quadrupole terms contribute $\tfrac32 J_2\,GM/a$ , and the centrifugal term contributes $\tfrac12\omega^2a^2$ :

\frac{GM}{a}f=\frac{GM}{a}\cdot\tfrac32 J_2+\tfrac12\omega^2 a^2

Step 4. Divide by $GM/a$ and use $\omega^2a^3/GM=m$ :

f=\tfrac32 J_2+\tfrac12 m \qquad (7)

where: each term now tells a story — the shape $f$ is held up partly by the bulge's own gravity ( $\tfrac32 J_2$ ) and partly by direct centrifugal lift ( $\tfrac12 m$ ).

Step 5. Read off the rigid-sphere limit. If the gravity field were frozen spherical ( $J_2=0$ — imagine the bulge material had no gravity of its own), the flattening would be $f=\tfrac12 m\approx1/578$ : this is the bare answer of the two-canal balance with unmodified gravity. The real Earth does better than this because the bulge pulls on itself. Verified: equation (7) with the measured $J_2=1.0826\times10^{-3}$ and $m=3.46\times10^{-3}$ gives $f=\tfrac32(1.0826\times10^{-3})+\tfrac12(3.46\times10^{-3})=3.36\times10^{-3}=1/298$ — the observed flattening, confirming the first-order theory to three digits.

Figure

Newton's two canals — equal pressure at the center forces the equatorial column to stand taller

How much does it flatten? Two classic results

The relation (7) contains two unknowns, $f$ and $J_2$ — the shape and the field it generates. Closing the system requires a model of the interior. The oldest and cleanest choice is Newton's: uniform density.

Go deeper · Derivation C · 1 minThe Maclaurin flattening of a homogeneous Earth

Derivation CThe Maclaurin flattening of a homogeneous Earth

We are deriving the classical slow-rotation flattening of a uniform-density fluid body, $f=\tfrac54 m$ — Newton's number, later embedded by Maclaurin in his exact spheroid solution. The strategy is a self-consistency loop: assume a small flattening $f$ , compute the quadrupole moment $J_2$ that a homogeneous spheroid of that shape produces, feed it back into the geodetic relation (7), and solve for $f$ . This closes the feedback the canal picture anticipated — the bulge's own gravity amplifies the bulge. Watch how a modest-looking factor ( $\tfrac35 f$ ) turns $\tfrac12 m$ into $\tfrac54 m$ : self-gravity more than doubles the deformation.

Step 1. Compute the quadrupole of a uniform spheroid. For a homogeneous body of mean radius $R$ flattened by $f\ll1$ , the principal moments of inertia differ by $C-A=\tfrac25 MR^2 f$ , so by equation (6):

J_2=\frac{C-A}{MR^2}=\frac{2}{5}f

Why: for uniform density the moments are pure geometry — squashing the sphere by $f$ moves mass from the poles to the equator, and the standard integrals give exactly the factor $\tfrac25$ .

Step 2. Insert into the geodetic relation (7):

f=\tfrac32 J_2+\tfrac12 m=\tfrac32\cdot\tfrac25 f+\tfrac12 m=\tfrac35 f+\tfrac12 m

Why this is the feedback loop: the term $\tfrac35 f$ is the bulge's own gravity helping to hold the bulge up — deformation creating the field that sustains more deformation.

Step 3. Solve for $f$ :

f-\tfrac35 f=\tfrac12 m\ \Rightarrow\ \tfrac25 f=\tfrac12 m\ \Rightarrow\ f=\tfrac54 m

The bare centrifugal lift alone would give $f=\tfrac12 m$ ; self-gravity raises it by the factor $\tfrac52$ to $\tfrac54 m$ .

Step 4. Numerical check for Earth:

f_{\text{Maclaurin}}=\tfrac54\times3.46\times10^{-3}\approx4.3\times10^{-3}\approx\tfrac{1}{231}

Why this overshoots the observed $1/298$ : the real Earth is centrally condensed — its dense iron core sits near the axis and contributes almost nothing to the bulge-driving quadrupole, so the true feedback is weaker than the homogeneous model assumes. *Verified: reproduces the standard Maclaurin slow-rotation limit; the chain $J_2=\tfrac25 f\Rightarrow f=\tfrac54 m$ matches Goldstein §5 and Stacey & Davis, and $1/231$ vs the observed $1/298.3$ correctly brackets the effect of central condensation.*

Remarkably, one combination of observables does not care about the interior at all. Clairaut found it in 1743, and it is the reason gravity surveys could map the Earth's shape a century before anyone knew what the core was made of. Define the gravity flattening $\beta=(g_p-g_e)/g_e$ , the fractional increase of surface gravity from equator to pole:

g(\theta)=g_e\big[1+\beta\sin^2\theta\big],\qquad \beta=\tfrac52 m-f \qquad (8)

where: $g(\theta)$ — surface gravity at colatitude $\theta$ ; $g_e=9.7803\ \mathrm{m/s^2}$ — equatorial surface gravity; $\beta\approx5.3\times10^{-3}$ — measured by timing pendulums (Richer's lost minutes) or, today, by absolute gravimeters. Rearranged, this is the celebrated identity:

f+\beta=\tfrac52 m \qquad (9)

where: $f$ — geometric flattening; $\beta$ — gravity flattening; $m$ — rotation parameter. Shape plus gravity variation equals $\tfrac52$ times spin — for any spherically layered density profile.

Go deeper · Derivation D · 2 minClairaut's theorem from pole-to-equator gravity

Derivation DClairaut's theorem from pole-to-equator gravity

We are deriving Clairaut's 1743 theorem, equation (9): the sum of the geometric flattening and the gravity flattening equals $\tfrac52 m$ , independent of the internal density distribution to first order. The strategy is direct: surface gravity is the magnitude of $-\nabla\Phi_{\mathrm{eff}}$ evaluated on the equipotential surface, so we differentiate the potential of equation (5) plus the centrifugal term, evaluate at the pole and at the equator — remembering that the pole sits at radius $c=a(1-f)$ , closer to the center — and form the fractional difference. The subtle point is that three separate effects contribute to $\beta$ : the centrifugal subtraction at the equator, the smaller polar radius, and the pull of the bulge mass itself; Clairaut's magic is that $J_2$ drops out of the final combination.

Step 1. Differentiate the potential radially to get surface gravity to first order:

g=-\partial_r\Phi_{\mathrm{eff}}=\frac{GM}{r^2}\Big[1-3J_2\frac{a^2}{r^2}P_2(\cos\theta)\Big]-\omega^2 r\sin^2\theta

Why the radial component suffices: on an equipotential the tangential component of effective gravity vanishes by construction — that was Derivation A — so to first order the magnitude is the radial part.

Step 2. Equatorial gravity ( $\theta=\pi/2$ , $r=a$ , $P_2=-\tfrac12$ ):

g_e=\frac{GM}{a^2}\big(1+\tfrac32 J_2\big)-\omega^2 a

Why both corrections lower the count differently: the bulge mass under your feet at the equator actually raises the field there (the $+\tfrac32 J_2$ ), while the centrifugal term $\omega^2a$ subtracts directly.

Step 3. Polar gravity ( $\theta=0$ , $r=c=a(1-f)$ , $P_2=1$ ):

g_p=\frac{GM}{a^2}\big(1+2f-3J_2\big)

Why: standing at the pole you are $fa$ closer to the center — the inverse square law gives $+2f$ — but the equatorial ring of bulge mass pulls sideways-and-up on you, reducing polar gravity by $3J_2$ .

Step 4. Form the fractional difference, writing $g_0\equiv GM/a^2$ and $\omega^2a=m\,g_0$ :

\frac{g_p-g_e}{g_0}=2f-\tfrac92 J_2+m

Step 5. Eliminate $J_2$ using the shape relation (7), $J_2=\tfrac23\big(f-\tfrac12 m\big)$ :

\beta=2f-3\big(f-\tfrac12 m\big)+m=\tfrac52 m-f

Hence $f+\beta=\tfrac52 m$ — the density profile has vanished from the relation.

Step 6. Numerical check with modern values: $f=3.353\times10^{-3}$ and $\beta=(9.8322-9.7803)/9.7803=5.31\times10^{-3}$ give $f+\beta=8.66\times10^{-3}$ , while $\tfrac52 m=\tfrac52\times3.47\times10^{-3}=8.67\times10^{-3}$ . Agreement to a part in a thousand. Verified: identity confirmed numerically with WGS-84/GRS-80 values ( $g_e=9.7803$ , $g_p=9.8322\ \mathrm{m/s^2}$ ); matches Stacey & Davis and the first-order theory in Chandrasekhar's monograph.

Limiting cases

The character of the equilibrium figure depends on how fast the body spins relative to its self-gravity, measured by $m=\omega^2 a/g$ (or by $\Omega^2=\omega^2/(2\pi G\bar\rho)$ for the homogeneous problem):

Case	Result	Meaning
Slow rotation (Earth, $m\ll1$ )	$f\approx \tfrac{5}{4}m$ (homogeneous); $f=\tfrac32 J_2+\tfrac12 m$ (real Earth)	Linear-in- $m$ theory. Clairaut's theorem holds; the figure is an oblate spheroid with flattening of order $m$ . Earth's $f\approx1/298$ .
Rapid rotation (gas giants)	$f_{\text{Jup}}\approx0.065$ , $f_{\text{Sat}}\approx0.098$	$m$ is no longer tiny ( $m_{\text{Sat}}\approx0.155$ ); higher-order terms and internal density structure matter. Visibly oblate through a telescope.
Critical rotation (Maclaurin limit)	$\Omega^2_{\max}\equiv \dfrac{\omega^2}{2\pi G\bar\rho}\approx0.2247$	Maximum dimensionless rotation for an axisymmetric homogeneous fluid; beyond this no oblate spheroid exists and the body sheds mass at the equator.
Bifurcation to triaxial (Jacobi)	$\Omega^2\approx0.1871$ , $e\approx0.8127$	At eccentricity $e\approx0.813$ the axisymmetric Maclaurin sequence becomes unstable to triaxial Jacobi ellipsoids — the figure loses its axial symmetry.
No rotation ( $\omega\to0$ )	$f\to0$ , $J_2\to0$	The equipotentials of an isolated self-gravitating fluid are spheres; the sphere is recovered as the unique equilibrium.

Worked examples

1 · Newton's prediction from two numbers (confidence builder)

Using only Earth's spin rate and size, predict the flattening of a homogeneous Earth and the height of the equatorial bulge, and compare with observation.

Step 1. The rotation parameter. With $\omega=7.292\times10^{-5}\ \mathrm{rad/s}$ (sidereal day, $86164$ s), $a=6.378\times10^{6}$ m and $g=9.78\ \mathrm{m/s^2}$ : $m=\omega^2a/g=(7.292\times10^{-5})^2\times6.378\times10^{6}/9.78=3.47\times10^{-3}\approx1/289$ .

Step 2. Homogeneous flattening: $f=\tfrac54 m=4.33\times10^{-3}\approx1/231$ .

Step 3. Bulge height: $a-c=f\,a=6378\ \mathrm{km}/231\approx\boxed{27.6\ \mathrm{km}}$ predicted for uniform density.

Step 4. Compare: the observed bulge is $a-c=21.4$ km, i.e. $f=1/298.3$ . Newton's 1687 estimate of $1/230$ was high by 29% — impressively close for a one-page calculation, and the direction of the error is itself physics: the real Earth hides a dense core near the axis, which weakens the self-gravity feedback and flattens the planet less. A homogeneous Earth would bulge 27.6 km; the real, core-heavy Earth bulges 21.4 km.

2 · Saturn, the planet that fails the homogeneous test (exam level)

Saturn rotates in $10.66$ h with equatorial radius $a=6.027\times10^{7}$ m and mass $M=5.68\times10^{26}$ kg. Compute its rotation parameter $m$ , the homogeneous flattening prediction, and compare with the observed $f_{\text{Sat}}=0.098$ . What do you conclude about Saturn's interior?

Step 1. Spin rate: $\omega=2\pi/(10.66\times3600\ \mathrm{s})=1.637\times10^{-4}\ \mathrm{rad/s}$ — only $2.2\times$ Earth's, but on a body ten times larger.

Step 2. Rotation parameter, using the $m=\omega^2a^3/GM$ form: $a^3=2.19\times10^{23}\ \mathrm{m^3}$ , $GM=6.674\times10^{-11}\times5.68\times10^{26}=3.79\times10^{16}\ \mathrm{m^3/s^2}$ , so $m=(1.637\times10^{-4})^2\times2.19\times10^{23}/3.79\times10^{16}=\boxed{0.155}$ — forty-five times Earth's value. Centrifugal lift at Saturn's equator is a sixth of its gravity.

Step 3. Homogeneous prediction: $f=\tfrac54 m=0.193$ — nearly one part in five.

Step 4. Observation gives $f_{\text{Sat}}=0.098$ , about half the homogeneous value. Saturn is strongly centrally condensed — most of its mass sits in a deep, dense interior (heavy-element core plus metallic-hydrogen envelope) that barely participates in the bulge. The measured $J_2$ and $J_4$ of exactly this kind of comparison are how the Cassini mission constrained Saturn's core mass. (At $m=0.155$ the first-order theory is also visibly straining — higher-order terms shift $f$ at the ten-percent level, which is why planetary geodesists expand to $m^2$ and beyond.)

Origins · 1 minPierre Louis Maupertuis and how this was discovered

Pierre Louis Maupertuis

1698–1759 · Paris Academy → Lapland expedition

Maupertuis made his name not with a theorem but with a journey. In 1736 he led the Academy's expedition to Tornio, in Lapland, to measure the length of one degree of latitude near the pole while a sister expedition under La Condamine measured a degree near the equator in Peru. If Newton was right and the Earth was oblate, polar degrees had to come out longer. Through a subarctic winter of frozen instruments, sledges, and sightings from church towers and bare hilltops, his team triangulated an arc across the Tornio valley — and found the polar degree longer by roughly a kilometre. The Earth was flattened, the Cassinis' prolate Earth was dead, and Voltaire teased Maupertuis as "the man who flattened the Earth and the Cassinis." He later became president of Frederick the Great's academy in Berlin and formulated an early version of the principle of least action — but it is the frozen survey, gloriously confirming a pen-and-paper prediction about the shape of the whole planet, that fixed his place in the history of mechanics.

History

1672

— Jean Richer finds his pendulum clock loses about 2.5 minutes per day in Cayenne: first evidence that gravity weakens toward the equator.

1687

— Newton's Principia predicts an oblate Earth with

f\approx1/230

from the two-canal equilibrium argument for a homogeneous fluid.

1720

— Jacques Cassini publishes French arc measurements suggesting a prolate Earth, igniting the oblate-versus-prolate controversy.

1737

— Maupertuis' Lapland expedition finds a polar degree of latitude longer than a French one: the Earth is oblate, Newton vindicated.

1743

— Clairaut's Théorie de la figure de la Terre proves

f+\beta=\tfrac52 m

, linking shape to gravity surveys independently of the interior.

Going further

References. Morin, Introduction to Classical Mechanics, ch. 9-10 (rotating frames, effective gravity) · Goldstein, Poole & Safko, Classical Mechanics 3rd ed., ch. 4-5 · Landau & Lifshitz, Mechanics (Vol. 1) and Fluid Mechanics (Vol. 6, §10 ellipsoidal figures) · Taylor, Classical Mechanics, ch. 9 · Stacey & Davis, Physics of the Earth (geodesy, J2, Clairaut) · Chandrasekhar, Ellipsoidal Figures of Equilibrium (1969)

Provenance. Numerical constants are standard textbook/reference values: $a=6378.137$ km, $c=6356.752$ km, $1/f=298.257$ (WGS-84/GRS-80); $J_2=1.0826\times10^{-3}$ ; $g_e=9.7803$ , $g_p=9.8322$ m/s $^2$ ; $m=\omega^2a/g\approx1/289$ ; giant-planet flattenings $f_{\text{Jup}}\approx0.065$ , $f_{\text{Sat}}\approx0.098$ ; Maclaurin/Jacobi critical values $\Omega^2_{\max}\approx0.2247$ , bifurcation $e\approx0.8127$ (Chandrasekhar). A verifier should recompute: the equipotential condition giving $f=\tfrac32 J_2+\tfrac12 m$ , the Maclaurin chain $J_2=\tfrac25 f\Rightarrow f=\tfrac54 m$ , and Clairaut's identity $f+\beta=\tfrac52 m\approx8.65\times10^{-3}$ . Sources: Morin ch. 9-10; Goldstein ch. 4-5; Landau-Lifshitz Vol 1 & Vol 6 §10; Stacey & Davis, Physics of the Earth; Chandrasekhar, Ellipsoidal Figures of Equilibrium.

Builds on: Centrifugal force · Effective gravity on the rotating Earth · Non-inertial reference frames · Fictitious (inertial) forces Leads to: Coriolis force · Effective gravity on the rotating Earth

If the centrifugal force is fictitious, how can it physically reshape a whole planet?

"Fictitious" describes the bookkeeping, not the reality of the effect. In the inertial frame there is no centrifugal force — instead, every parcel of a spinning planet is accelerating centripetally, and the only forces available to supply that acceleration are gravity and pressure. A sphere cannot arrange its gravity and pressure to give each parcel exactly the required centripetal pull; an oblate spheroid can. The rotating-frame language ("centrifugal lift makes fluid pile up at the equator") and the inertial language ("the shape adjusts until gravity minus pressure provides the centripetal acceleration") describe the identical equilibrium and yield the identical equation. Fictitious forces are a convenience of description; the 21.4 km bulge is measured with satellites and is as real as the planet itself.

Why doesn't the ocean flow downhill from the 21-km-high equatorial bulge toward the poles?

Because "downhill" means "in the direction of the tangential component of effective gravity," and on an equipotential surface there is no such component — that is exactly what Derivation A proves. The bulge is not a hill sitting on a level Earth; it is the level. A drop of water at the equator, 21 km farther from the center than a drop at the pole, has precisely the same value of $\Phi_{\mathrm{eff}}$ , because the extra centrifugal potential compensates the gravitational difference. If you somehow removed the bulge while the Earth kept spinning, the ocean would flow toward the equator to rebuild it.

Is Mount Everest really the point farthest from the center of the Earth?

No — that record belongs to Chimborazo in Ecuador, and the bulge is the reason. Everest's summit is 8.8 km above sea level at latitude 28°N; Chimborazo is only 6.3 km high but sits one degree from the equator, where sea level itself stands about 21 km farther from the center than at the poles. Adding it up, Chimborazo's summit is roughly 2 km farther from the center of the Earth than Everest's. "Highest above sea level" and "farthest from the center" are different questions on an oblate planet — a neat everyday consequence of the figure of the Earth.

Newton predicted f = 1/230 but the truth is 1/298. Was his argument wrong?

The argument was right; only the assumed interior was wrong. The canal balance and the equipotential condition are exact first-order physics, and with the measured quadrupole $J_2$ they reproduce $f=1/298$ on the nose (equation 7). Newton had no way to know the density profile, so he took it uniform, which maximizes the self-gravity feedback and overshoots by 29%. The gap between $1/230$ and $1/298$ is not an error term — it is a measurement of the Earth's central condensation, and eighteenth-century physicists (Clairaut among them) understood it exactly that way: the figure of the Earth was the first probe of the deep interior.

Why did the Cassinis think the Earth was prolate, stretched along the axis?

They had real data — arc measurements along the Paris meridian — but over too short a baseline. Comparing degree lengths between northern and southern France means distinguishing a change of a few hundred metres in about 111 km, and the surveying errors of the day were comparable to the signal. Their arcs came out slightly shorter to the north, which reads as prolate. The fix was leverage: put one arc near the pole and one near the equator, where the oblate-versus-prolate difference is maximal. That was precisely the design of the Lapland and Peru expeditions, and Maupertuis' polar degree came out about a kilometre longer than a French one — an unambiguous oblate verdict.

Does the equatorial bulge do anything to satellites, or is it just geography?

It dominates their long-term behaviour. The bulge makes Earth's external field depart from $1/r^2$ by the $J_2$ term of equation (5), and that quadrupole exerts a torque on every orbit: the orbital plane precesses (the node regresses for prograde orbits) at a rate proportional to $J_2\cos i/a^{7/2}$ . GPS operators must model it or the constellation drifts; Sun-synchronous satellites are deliberately placed at $i\approx98°$ so the precession tracks the Sun. Historically the logic ran backwards: watching how early satellites' orbits precessed is how $J_2$ — and thus the figure of the Earth — was first measured to modern precision, within months of Sputnik.

If the Earth stopped spinning, would it relax back to a perfect sphere?

Eventually, yes — but not instantly, and the lag is itself informative. The mantle behaves as a very viscous fluid on timescales of tens of thousands of years (that is why the theory treats the Earth as a fluid in the first place), so the bulge would decay over roughly the same $10^4$ – $10^5$ -year timescale seen in post-glacial rebound, where Scandinavia is still rising a centimetre a year in "memory" of the last ice sheet. On shorter timescales the solid Earth has elastic rigidity and would hold a fossil bulge. Something similar happens as Earth's rotation slowly brakes under lunar tides: the equilibrium figure tracks the decreasing $m$ , and the slow readjustment feeds into the long-term evolution of the length of the day.