Earth bulges at the equator, and that uneven pull slowly swings a satellite's orbit plane around and rotates its ellipse — a drift called precession.
A satellite circling a perfect, uniform ball would trace the same path in space forever. Earth is not perfect. It is wider at the equator than pole to pole, and the Sun and Moon pull on satellites too.
These extra tugs make orbits slowly change. NASA calls the result a walking, or precessing, orbit. For satellites at geostationary height and below, Earth's equatorial bulge is the strongest cause. Above that, the Sun and Moon dominate.
The bulge does two things.
It swings the orbit plane around Earth's axis, like a tilted spinning top wobbling. The point where the satellite crosses the equator going north drifts around the globe.
It also turns the ellipse within its own plane, so the perigee slowly moves around the orbit. At two special tilts, 63.4° and 116.6°, this second turning stops. Molniya orbits use 63.4° so their high, slow apogee stays over high northern latitudes.
Sun-synchronous orbits use the first effect on purpose. Cartosat-3's height (509 km) and tilt (97.5°) are chosen so its orbit plane swings round once a year, keeping pace with the Sun. Its equator crossings then always happen at about the same local time. Even this needs care: NASA notes that the Sun's and Moon's pull slowly shifts the inclination of sun-synchronous orbits, so it has to be trimmed every year or two.
For geostationary satellites the drift is a nuisance. The Sun's and Moon's pull slowly tilts their orbits, so operators must fire thrusters to keep them over the equator.
The ISS orbits at 51.6°, about 414 to 424 km up. At that height and tilt, Earth's bulge swings its orbit plane westward by about 5° a day, a full circle in about 72 days. This drift is one reason the times of visible ISS passes over an Indian city change from week to week.
Note the local time when Cartosat-3 crosses India; precession keeps that time nearly the same from week to week. Track Cartosat-3
To first order the bulge (J2) produces two secular drifts:
Node: dΩ/dt = −1.5 · n · J2 · (R/a)² · cos i / (1 − e²)² Perigee: dω/dt = 0.75 · n · J2 · (R/a)² · (4 − 5 sin² i) / (1 − e²)²
with n the mean motion, R = 6,378 km and J2 = 1.08263 × 10⁻³.
Worked numbers.
Cartosat-3 (a = 6,887 km, n = 5,468°/day, i = 97.5°): 1.5 × 5,468 × 0.00108263 × 0.8577 = 7.616°/day, times −cos 97.5° = 0.1305, gives dΩ/dt ≈ +0.99°/day. That is the sun-synchronous rate of 360° / 365.256 days = 0.9856°/day.
The ISS (51.6°, about 414 × 424 km, so a ≈ 6,797 km): the same formula gives about −5.0°/day. The node runs westward and laps Earth about every 72 days. That is why the times of ISS passes over your city shift from week to week.
The perigee term vanishes when 4 − 5 sin² i = 0, so sin² i = 0.8 and i = 63.43° or 116.57°. This is the critical inclination used by Molniya orbits.
At i = 90° the node term is zero too: a truly polar orbit's plane does not swing at all under J2. And because both rates scale as (R/a)², the bulge matters far less for high orbits — which is why the Sun and Moon take over above GEO.
Its equatorial bulge (oblateness). The extra mass at the equator pulls unevenly on the orbit and slowly swings its plane.
At that inclination the bulge does not rotate the perigee, so the high apogee stays over the same northern latitudes.
No. The node drift is proportional to cos i, which is zero at 90°.