The Mismatch Before Any Projection
Go back to the two points from the last chapter, a hundred meters apart, sitting at a modest inland site, five hundred meters above the ellipsoid. No projection has entered this story yet. State Plane, UTM, none of it has come up. And the two distances still do not match.
Elevation alone, with no projection involved at all, causes a GNSS-derived distance and a ground-measured distance to differ, because the ground sits above the ellipsoid, and the ellipsoid is what a purely geodetic GNSS distance is referenced to.
Two Spokes on a Wheel
Picture two spokes of a wheel, close together near the hub, spreading slightly farther apart the further out they extend. Two points a fixed angular distance apart behave the same way: measured out at the Earth's actual surface, sitting above the ellipsoid by the site's elevation, they sit slightly farther apart than the equivalent points would if projected straight down onto the ellipsoid itself. A GNSS-derived distance, computed from each point's position relative to that ellipsoid, describes the shorter, inner spacing. The ground, sitting higher up, is the longer one.
The ellipsoid is not a stand-in for the ground, or for sea level. A site's height above the ellipsoid and its ordinary elevation above sea level are two different numbers, close in many places but not identical, and in some low-lying locations the ground can even sit below the ellipsoid rather than above it. A purely geodetic distance is always referenced to the ellipsoid surface itself, not to the ground a crew is actually standing on.
Putting a Number on It
At five hundred meters of elevation, the reduction between a ground distance and its ellipsoidal equivalent works out to roughly eight parts in one hundred thousand. Over the hundred meter baseline from the last chapter, that comes out to about eight millimeters, roughly three-tenths of an inch. Small, on a single baseline. It does not stay small forever, and this companion will return to that fact directly once a longer distance enters the picture.
This elevation-driven gap is not a projection artifact, and it does not go away by working in raw latitude and longitude instead of State Plane or UTM. It is present the moment any elevation separates a site from the ellipsoid, projected or not.
Why This One Is Easy to Miss
Grid scale factor, covered in the next chapter, gets talked about often enough in survey circles that at least the name is familiar to some GNSS users. This one rarely gets its own name at all, because it hides inside every geodetic distance calculation a receiver ever performs, quietly, whether or not a projection is involved. A user working entirely in latitude and longitude, never touching a projected coordinate system, still cannot assume a GNSS-derived distance matches a taped one, purely because of the elevation the site sits at.
A GNSS user never uses State Plane or UTM, working entirely in raw latitude, longitude, and ellipsoid height. Does that mean their GNSS-derived distances will always match a total station's ground measurement? No. Elevation alone still separates the two, with no projection involved at all.
A GNSS-derived distance is referenced to the ellipsoid, not to the ground. Because the ground sits above the ellipsoid by the site's elevation, a ground-measured distance is very slightly longer than its ellipsoidal equivalent, an effect that exists with or without any map projection involved.
The moment a workflow needs a flat, stakeable distance rather than a geodetic one, a second, entirely separate distortion enters the picture, and it has a name most GNSS users have heard without ever quite being told what it means.