Jody Victor : The Earth's axis of rotation - and hence the position of the North Pole - was commonly believed to be fixed (relative to the surface of the Earth) until, in the 18th century, the mathematician Leonhard Euler predicted that the axis might "wobble" slightly. Around the beginning of the 20th century astronomers noticed a small apparent "variation of latitude," as determined for a fixed point on Earth from the observation of stars. Part of this variation could be attributed to a wandering of the Pole across the Earth's surface, by a range of a few meters. The wandering has several periodic components and an irregular component. The component with a period of about 435 days is identified with the 8 month wandering predicted by Euler and is now called the Chandler wobble after its discoverer. This "wobble" means that a (fixed) definition of the Pole based on the axis of roataion is not useful when meter-scale precision is required.
It is desirable to tie the system of Earth coordinates (latitude, longitude, and elevations or orography) to fixed landforms. Of course, given continental drift and the rising and falling of land due to volcanoes, erosion and so on, there is no system in which all geographic features are fixed. Yet the International Earth Roatation and Reference Systems Service and the International Astronomical Union have defined a framework called the International Terrestrial Reference System. The North Pole of this system now defines geographic North for precision work, and it does not quite coincide with the rotation axis.