What This Series Already Taught Us
This companion asks one question, over and over, in a dozen different disguises: grid, or ground? Answering it well depends on a small amount of groundwork this series already laid elsewhere, and rather than rebuild that foundation from scratch, this chapter simply states what to remember and where it came from.
A reference frame and a map projection were both covered at length in the second companion of this series. This chapter recaps only what the rest of this companion leans on directly, and does not re-teach either topic from the beginning.
A Coordinate Belongs to a Frame
A GNSS position is never just three numbers. It belongs to a specific reference frame, a specific realization of that frame, tied to a specific moment in time. Two receivers reporting different frames can disagree by a meter or more without either one being wrong, and no amount of precision closes that gap, only an explicit transformation does. That idea does not change anywhere in this companion. Every distance discussed here assumes both ends of the measurement already agree on frame and epoch, so that the only remaining question is the one this companion actually cares about.
If two coordinates disagree because they belong to different reference frames, that is a different problem from everything this companion covers, and it is worth ruling out first. This companion assumes that question is already settled.
A Projection Trades Curvature for Convenience
Latitude and longitude describe angles on a curved surface, and a degree of longitude does not represent a constant ground distance anywhere except the equator. For almost any practical distance, stakeout, or area calculation, that curved description has to be flattened onto a plane, and flattening a curved surface always distorts something. A map projection is the deliberate, engineered choice of where to put that distortion and how much of it to allow. State Plane and UTM are two concrete, complete answers to that same problem, built around different priorities, not two steps in a sequence.
A map projection is not an optional add-on that only some GNSS workflows use. The moment a project needs a straight, subtractable distance rather than an angle, some projection has already entered the picture, whether or not anyone chose it deliberately.
What This Companion Assumes, and What It Does Not
This companion assumes the reader already has both of those ideas in hand: a coordinate lives inside a reference frame, and a projected coordinate has already traded some accuracy in distance for the convenience of a flat, usable plane. What this companion adds is the specific, practical consequence of that second trade, the one most GNSS users encounter constantly and few have ever had explained: why a distance computed from GNSS positions and a distance measured on the actual ground do not have to be the same number, and what to do once that gap is understood rather than mistaken for an error.
Two coordinates disagree by exactly the amount a documented reference frame difference would predict. Does that belong in this companion? No. That is a frame problem, already covered elsewhere in this series, and it should be ruled out before reaching for anything in the chapters ahead.
This companion assumes a coordinate's reference frame is already settled, and that flattening a curved position onto a plane always trades some distance accuracy for usability. Everything from here forward is about the specific shape that trade takes, and what it costs a GNSS user in practice.
With that foundation restated rather than rebuilt, the next chapter puts two real points on the ground and asks the question this entire companion is organized around.