# Meaning of “Elevation above surface of ellipsoid”?

I'm using implementation of Vincenty's algorithms, and I have their class called "position" that requires lat, lon, and elevation defined by: "Elevation above the surface of the ellipsoid".

What is the meaning in such applications? Is it just simple "Elevation above MSL"?

The elevation above the ellipsoid (ellipsoidal height) is the elevation above a mathematical model that approximates the shape of the earth. The current most common one is WGS84. These are the elevations that you'd get from a GPS.

Orthometric heights are measured above the geoid or equipotential surface, that is, the surface of equal gravity. MSL is "mean sea level," which is supposed to roughly approximate the equipotential surface, but obviously can't be directly measured inland.

Unlike the ellipsoid, the geoid can't be represented by a function is complicated (see 2NinerRomeo's comment), so conversions have to use a grid shift raster to find the ellipsoidal separation at any given location. NRCAN has a decent page describing this stuff.

If you have orthometric (e.g. MSL) heights, you'll need to transform them using the appropriate grid shift file.

• The geoid is represented by a function. It just happens to be a harmonic function with many, many coefficients. The National Geospacial Intelligence agency provides both the coefficients, a rasterized height map to a resolution of 15", and a piece of code to interpolate the function. – 2NinerRomeo Aug 4 '15 at 21:17
• Wikipedia has a great graphic which shows the deviation of the EGM96 geoid with respect to the WGS48 Ellipsoid. – 2NinerRomeo Aug 4 '15 at 21:17
• Added a note highlighting your comment. – Rob Skelly Aug 4 '15 at 21:30
• +1 Concerning the geoid and orthometric heights, see the diagrams (in the question itself) and analysis at gis.stackexchange.com/questions/112195. For references to earth models (including spheroids and ellipsoids), see gis.stackexchange.com/a/53737. – whuber Aug 4 '15 at 21:58
• Kind of, @2NinerRomeo. There are functions which are published as approximations to the geoid, but the geoid isn't defined by them. Though of course such models may be referred to like you did as "the EGM96 geoid", and people will know what is meant. – Doug McClean Aug 5 '15 at 4:14

An ellispoid is a mathematical model of the earth that approximates its three dimensional shape. See this definition. Elevation on top of the ellipsoid is 0, but since it's just an approximation one can be above or below the ellipsoid at any given point. "Elevation above the surface of the ellipsoid" is the distance between the measurement and the 0 value of the ellipsoid.

The Z value in a given coordinate system has to be based on something--a height above a generalized shape of the earth. MSL is one way to do it, but in my experience the majority of cases use ellipsoids as approximate figures. GPS, for example, uses WGS84 as the global coordinate system, and with it is the WGS84 ellipsoid.

An "ellipsoid" is a mathematical approximation of the shape of the Earth. Many different ellipsoids exist, but the two most widely used today are the GRS80 and the WGS84, which attempt to provide a best-fit across the globe. Heights were traditionally referenced to MSL, but with satellite and other technologies, we can often do better in terms of accuracy. HAE (Height above ellipsoid) is the term often used. (info summarized from Bolstad, 2012)