If you look at the star canvas from the north poll .. it will rotate around a point, called the celestial (north) pole, which lies fairly close to the zenith - the highest point of the sky.
If you move away from the north pole, the celestial north pole moves further and further away from the zenith. At the equator, you don't see the point the star sky rotates around anymore.
When moving towards the south pole, another point comes into view - the celestial south pole. Around which the sky rotates. If you get close to Antarctica, this point comes closer and closer to the zenith again .. till the star sky rotates around the same point again.
How is this possible with a flat earth? I understand the explenation of the star canvas in FE is that it rotates around the north pole.
Then WHY does it appear like the sky rotates around a single point at the south pole? Why does it look like one is looking at the centre of a rotating disc, and NOT the rim, as it would have to be if the FE explenation was correct?
While on that point, how come Astronomical system conversions, based on the roundness of the earth, *work*, 100%? Seems like a wee bit of a coincidence.
If I'm in an observatory in Puerto Rico and map the location of a star on the sky at a certain time, I don't put its location relative to my location on the records .. I note that down and from that calculate its position relative to the centre of the earth.
And I can, from an observatory in, say, the Canary Islands, calculate this data back into data applicable to my local coordinates .. that will tell me exactly where this star is on my sky at the given time.
Why oh why does this method .. WORK? Quite well, too, I've used centre-of-the-earth based databases myself to find the location of a star on my evening sky.