jimirwin a écrit dans le message ...
Rod Burns wrote in sci.geo.geology:

I think the reference is not a point but a surface, the geoid surface,
which is an oblate spheroid that is determined I think by measuring
the gravity (I don't know how, thpough). This gives 0 meter altitude.


Very close. The ellipsoid is a mathematical reference surface used by
geodesists and cartographers. There are several, but GRS80 or WGS84 are
most commonly used today.

The geoid is a gravitational equipotential surface undulating above and
below the ellipsoid. The WGS84 value for the gravitational potential was
chosen for best fit to the global sea surface, but it doesn't follow the
sea surface. [In fact "the geoid" is a misnomer, since there are actually
several different potentials that have merit.]

Mean sea level is location dependent and may vary from the geoid due to
ocean dynamics. Different datums specify mean sea level by averaging
designated tide guages.

Elevation is measured with reference to a datum.

--
Jim Irwin
http://www.holoscenes.com


mean sea level -A tidal datum: the arithmetic mean of hourly water
elevations observed over a specific 19-year cycle. Points on land can be
referenced to a mean sea level, in which case the datum assumes zero
elevation.

" sea level is not a simple surface. Since the sea surface conforms to the
earth's gravitational field, MSL also has slight hills and valleys that are
similar to the land surface but much smoother. However, zero elevation as
defined by Spain is not the same zero elevation defined by Canada, which is
why locally defined vertical datums differ from each other."

The geoid corresponds to mean sea level.
In land covered areas, the geoid is the surface that would be determined by
the level to which water would rise in narrow canals cut through the
continents.

JOL