A state plane coordinate is a position on a flat map of a curved earth. The flattening is done by a projection, the projection is chosen so that the distortion stays small over a limited area, and the limited area is the zone. Everything confusing about the system follows from that one compromise: a plane grid is convenient, the earth is not a plane, and the difference has to go somewhere.
Where it goes is into the scale. A distance measured on the ground is not the same as the distance between the two grid coordinates of its endpoints, and the ratio between them is knowable to a part per million. Failing to apply it is the most common way a survey ends up internally consistent and externally wrong.
This page covers what a zone is, what the numbers in a coordinate mean, the two scale corrections and how to combine them, a worked conversion between ground and grid, and the metadata without which a coordinate is not a position at all.
Zones and projections
The State Plane Coordinate System divides the United States into zones, each with its own projection, its own defining parameters and its own coordinate values. States that are long east to west are cut into zones by longitude and mapped with a Lambert conformal conic projection; states long north to south are cut by latitude and mapped with a transverse Mercator. A few zones use an oblique Mercator where the terrain runs diagonally.
Both projections are conformal, meaning angles are preserved locally: a right angle on the ground is a right angle on the grid, and shapes survive at small scale. What is not preserved is distance. Conformality is precisely the trade - the projection may stretch, but it stretches equally in every direction at a point, so the stretch can be described by a single number.
- Zone
- A region with one projection and one set of constants. Zone boundaries follow county lines rather than meridians, because a county should not be cut in half by a coordinate system. Coordinates from two zones are not comparable and must never be mixed in one file.
- False easting and false northing
- Constants added to every coordinate so that no value in the zone is negative and so eastings and northings cannot be confused for one another. They are the reason a state plane easting is a large number such as 1,968,500 rather than a small distance from the central meridian.
Grid scale factor
The grid scale factor is the ratio of a distance on the grid to the corresponding distance on the ellipsoid. On a Lambert zone the projection cone cuts the ellipsoid along two standard parallels, so the factor is exactly one there, less than one between them, and greater than one outside. On a transverse Mercator zone the same pattern runs along two lines either side of the central meridian, where the factor is at its minimum - conventionally about 0.9999.
The variation across a zone is deliberately bounded to roughly one part in ten thousand, which is 0.1 ft per 1,000 ft. That is invisible in a single tape length and obvious on a boundary line. It varies with position, so it must be evaluated for the project, not assumed.
Elevation factor
The grid scale factor gets you between the grid and the ellipsoid. It does not get you to the ground, because the ground is above the ellipsoid and a measured distance up there subtends a longer arc than the same angle does at ellipsoid level.
elevation factor = R ⁄ (R + H + N)At a project elevation of 1,450 ft, which is 441.96 m, and a geoid height of −27.5 m, the elevation factor is 0.99993496. Sea level is not the reference; the ellipsoid is, and the geoid height is what converts between them. Dropping N introduces an error of about 4 ppm in the conterminous states and considerably more elsewhere.
The combined factor, and a worked conversion
combined factor = grid scale factor × elevation factor
grid distance = ground distance × combined factor
ground distance = grid distance ⁄ combined factor| Quantity | Value |
|---|---|
| Grid scale factor | 0.99994120 |
| Project elevation | 1,450 ft (441.96 m) |
| Geoid height | −27.5 m |
| Elevation factor | 0.99993496 |
| Combined factor | 0.99987616 |
| Departure from unity | −123.8 ppm |
A ground distance of 4,286.51 ft becomes a grid distance of 4,285.979 ft - the grid is 0.531 ft shorter. Running it the other way, a grid mile of 5,280.00 ft is 5,280.654 ft on the ground. At 124 ppm the correction is 0.12 ft per 1,000 ft, which is roughly ten times the closure tolerance of a decent boundary traverse and about a hundred times the resolution of the instrument that measured it.
The metadata a coordinate cannot do without
A pair of numbers is not a position. Six pieces of information are needed before a state plane coordinate means anything, and a deliverable that omits any of them has exported an ambiguity rather than a survey.
- The zone, by name and ideally by its numeric identifier. Adjacent zones overlap in coordinate value, so the same pair of numbers is a valid position in more than one place.
- The datum and its realization. NAD 83 (2011) and NAD 83 (1986) are both NAD 83 and differ by up to a meter; a coordinate labelled only NAD 83 has thrown that away.
- The epoch, wherever the crust moves measurably - which is most of the western states.
- The unit, and which foot. State plane coordinates in feet are overwhelmingly in US survey feet for anything published before the 2020s, and the two feet differ by 2 ppm, which is feet at state plane magnitudes.
- Whether the values are grid or ground. Modified state plane systems, scaled and translated to be usable as ground coordinates, look exactly like real state plane coordinates and are not.
- The vertical datum and geoid model, if heights are included. NAVD 88 with GEOID18 is a different height from an ellipsoid height by tens of meters.
Write those six lines on the plat and in the metadata of every file that leaves your office. The computations on this page are arithmetic and take a minute; recovering which zone and which foot a legacy dataset was in can take a day and sometimes cannot be done at all.
For higher-precision work, or where a project spans enough elevation change that a single combined factor stops being adequate, a low-distortion projection designed for the local area removes the problem rather than correcting for it. That is increasingly the modern answer, but it does not change how the millions of existing coordinates were computed, and reading those correctly is still the daily job.