Originally Posted by
JDNSW
I think this is the appropriate spot to point out that the mapped coordinates of a point on the earth depends on the spheroid and datum used for the mapping.
If the ones used for the map data are not the same as the one used by the GPS system, the position shown by GPS will be incorrect. Local mapping uses a spheroid and datum such that this mathematical surface corresponds to the equipotential surface as closely as possible for a mathematical model, e.g. Australia typically uses the Australia datum and Clarke 1858 spheroid. GPS, on the other hand, uses a spheroid and datum that tries to match the equipotential surface over the whole earth.
Unless there is a conversion between the two, usually in the GPS receiver or the map supplied to the GPS receiver, there will be a mismatch, which can run into hundreds of metres. And then there are straight mapping errors.