Why is Earth's gravity stronger at the poles?
Lots of places state that the Earth's gravity is stronger at the poles than the equator for two reasons:
- The centrifugal force cancels out the gravity minimally, more so at the equator than at the poles.
- The poles are closer to the center due to the equatorial bulge, and thus have a stronger gravitational field.
TL;DR version: There are three reasons. In order of magnitude,
The poles are closer to the center of the Earth due to the equatorial bulge. This strengthens gravitation at the poles and weakens it at the equator.
The equatorial bulge modifies how the Earth the gravitates. This weakens gravitation at the poles and strengthens it at the equator.
The Earth is rotating, so an Earth-bound observer sees a centrifugal force. This has no effect at the poles and weakens gravitation at the equator.
Let's see how the two explanations in the question compare to observation. The following table compares what a spherical gravity model less centrifugal acceleration predicts for gravitational acceleration at sea level at the equator ($g_{\text{eq}}$) and the north pole ($g_{\text{p}}$) versus the values computed using the well-established Somigliana gravity formula $g = g_{\text{eq}}(1+\kappa \sin^2\lambda)/\sqrt{1-e^2\sin^2 \lambda}$.
$\begin{matrix} \text{Quantity} & GM/r^2 & r\omega^2 & \text{Total} & \text{Somigliana} & \text{Error} \\ g_\text{eq} & 9.79828 & -0.03392 & 9.76436 & 9.78033 & -0.01596 \\ g_\text{p} & 9.86431 & 0 & 9.86431 & 9.83219 & \phantom{-}0.03213 \\ g_\text{p} - g_\text{eq} & 0.06604 & \phantom{-}0.03392 & 0.09995 & 0.05186 & \phantom{-}0.04809 \end{matrix}$
This simple model works in a qualitative sense. It shows that gravitation at the north pole is higher than at the equator. Quantitatively, this simple model is not very good. It considerably overstates the difference between gravitation at the north pole versus the equator, almost by a factor of two.
The problem is that this simple model does not account for the gravitational influence of the equatorial bulge. A simple way to think of that bulge is that it adds positive mass at the equator but adds negative mass at the poles, for a zero net change in mass. The negative mass at the pole will reduce gravitation in the vicinity of the pole, while the positive mass at the equator will increase equatorial gravitation. That's exactly what the doctor ordered.
Mathematically, what that moving around of masses does is to create a quadrupole moment in the Earth's gravity field. Without going into the details of spherical harmonics, this adds a term equal to $3 J_2 \frac {GMa^2}{r^4}\left(\frac 3 2 \cos^2 \lambda - 1\right)$ to the gravitational force, where $\lambda$ is the geocentric latitude and $J_2$ is the Earth's second dynamic form. Adding this quadrupole term to the above table yields the following:
$\begin{matrix} \text{Quantity} & GM/r^2 & r\omega^2 & J_2\,\text{term} & \text{Total} & \text{Somigliana} & \text{Error} \\ g_\text{eq} & 9.79828 & -0.03392 & \phantom{-}0.01591 & 9.78027 & 9.78033 & -0.00005 \\ g_\text{p} & 9.86431 & 0 & -0.03225 & 9.83206 & 9.83219 & -0.00013 \\ g_\text{p} - g_\text{eq} & 0.06604 & \phantom{-}0.03392 & -0.04817 & 0.05179 & 0.05186 & -0.00007 \end{matrix}$
This simple addition of the quadrupole now makes for a very nice match.
The numbers I used in the above:
$\mu_E = 398600.0982\,\text{km}^3/\text{s}^2$, the Earth's gravitational parameter less the atmospheric contribution.
$R_\text{eq} = 6378.13672\,\text{km}$, the Earth's equatorial radius (mean tide value).
$1/f = 298.25231$, the Earth's flattening (mean tide value).
$\omega = 7.292115855 \times 10^{-5}\,\text{rad}/\text{s}$, the Earth's rotation rate.
$J_2 = 0.0010826359$, the Earth's second dynamic form factor.
$g_{\text{eq}} = 9.7803267714\,\text{m}/\text{s}^2$, gravitation at sea level at the equator.
$\kappa = 0.00193185138639$, which reflects the observed difference between gravitation at the equator versus the poles.
$e^2 = 0.00669437999013$, the square of the eccentricity of the figure of the Earth.
These values are mostly from Groten, "Fundamental parameters and current (2004) best estimates of the parameters of common relevance to astronomy, geodesy, and geodynamics." Journal of Geodesy, 77:10-11 724-797 (2004), with the standard gravitational parameter modified to exclude the mass of the atmosphere. The Earth's atmosphere has a gravitational effect on the Moon and on satellites, but not so much on people standing on the surface of the Earth.
The point is that if we approximate Earth with an oblate ellipsoid, then the surface of Earth is an equipotential surface,$^1$ see e.g. this Phys.SE post.
Now, because the polar radius is smaller than the equatorial radius, the density of equipotential surfaces at the poles must be bigger than at the equator.
Or equivalently, the field strength$^2$ $g$ at the poles must be bigger than at the equator.
--
$^1$ Note that the potential here refers to the combined effect of gravitational and centrifugal forces. If we pour a bit of water on an equipotential surface, there would not be a preferred flow direction.
$^2$ Similarly, the field strength, known as little $g$, refers to the combined effect of gravitational and centrifugal forces, even if $g$ is often (casually and somewhat misleading) referred to as the gravitational constant on the surface of Earth.
Here's a simple argument that doesn't require any knowledge of fancy stuff like equipotentials or rotating frames of reference. Imagine that we could gradually spin the earth faster and faster. Eventually it would fly apart. At the moment when it started to fly apart, what would be happening would be that the portions of the earth at the equator would at orbital velocity. When you're in orbit, you experience apparent weightlessness, just like the astronauts on the space station.
So at a point on the equator, the apparent acceleration of gravity $g$ (i.e., what you measure in a laboratory fixed to the earth's surface) goes down to zero when the earth spins fast enough. By interpolation, we expect that the effect of the actual spin should be to decrease $g$ at the equator, relative to the value it would have if the earth didn't spin.
Note that this argument automatically takes into account the distortion of the earth away from sphericity. The oblate shape is just part of the interpolation between sphericity and break-up.
It's different at the poles. No matter how fast you spin the earth, a portion of the earth at the north pole will never be in orbit. The value of $g$ will change because of the change in the earth's shape, but that effect must be relatively weak, because it can never lead to break-up.