Trying to tie any unit to earth geography is fraught with difficulties. The major problem is the earth's equatorial bulge. It gave the French difficulties when they tried to tie the meter to the equator-to-pole difference. (But remember, the exact value of the base of any measurement system is never important - good measurement systems are dependent on other characteristics.)

The equatorial bulge affects satellites as well. The RAAN (Right Ascension of Ascending Node - the longitude at which the south to north path of the satellite crosses the equator - more here...

https://en.wikipedia.org/wiki/Longit...ascending_node

drifts westward due to the bulge. There is an equation that predicts the drift rate...

https://en.wikipedia.org/wiki/Nodal_precession

as a function of the earth's oblateness.

I was once tasked to provide an improved gravity model that could be used to integrate the satellite's motion without using the drift rate equation. As it turns out, the oblate earth can be accurately modeled as a combination of a perfectly spherical mass with a torus around the equator to mimic the bulge. Then the net gravity acting on the satellite can be computed as a vector addition of the components arising from the sphere and the torus.

Need I mention that computing the gravity field of a torus is a daunting mathematical exercise ? Using the equation is vastly easier.