Every Coordinate You Own Is About to Change

The frame under everything is retiring

Every coordinate an American surveyor has delivered for roughly forty years stands on the same two foundations: NAD83 for horizontal, NAVD88 for vertical. Every plat, every state-plane PNEZD file, every benchmark elevation in your fileroom implicitly begins with “assuming NAD83/NAVD88…”

That assumption now has an expiration date. The National Geodetic Survey is replacing both with a modernized National Spatial Reference System – NATRF2022 for horizontal (one of four new plate-fixed frames), NAPGD2022 for the vertical, built on a new gravimetric geoid rather than a network of leveled benchmarks. NGS has been rolling components out for public testing on its beta site, with the official arrival targeted by the end of 2026.

This is not a paperwork rename. NAD83’s origin has been known for decades to sit off the true center of the Earth; NAVD88 carries a measurable tilt across the continent. The new frames fix both, which means the coordinates themselves change – on the order of a meter or more horizontally in much of the lower 48, and decimeters vertically, with the exact shift depending on where you stand. The monument in the ground does not move. Every number describing it does.

What actually changes for a working firm

Every deliverable’s frame of reference. New projects will, over time, be requested in the new frames – agencies first, then everyone. For a transition period your firm will live in both worlds at once, which is precisely when labeling discipline matters most.

State plane itself. The state plane coordinate systems surveyors memorized are being redesigned alongside the datums (SPCS2022). Zone definitions, and in many states the zone layout itself, change. If one EPSG digit already separates you from half a state, the family of codes is about to get a new generation.

Time enters the coordinate. The modernized frames track plate motion explicitly – coordinates come with an epoch, and positions drift predictably over time instead of pretending the continent stands still. “Where is this point?” gains a second clause: when? Workflows and software that assume a coordinate is forever will need to grow up.

Your software stack. GIS, CAD, data collectors, and every transformation library in between need updated definitions before any of this flows through cleanly. That work is upstream of you – but checking that your stack handles the new frames correctly is not.

The part nobody prices: the archive

Here is the thesis worth sitting with: datum succession is the biggest institutional-memory event in a generation.

NGS will publish transformation tools, as it always has. Mechanically, moving a coordinate from NAD83(2011) to NATRF2022 will be a solved problem. But a transformation is only as good as the label on its input – and that label is exactly what forty years of archives are missing.

Your fileroom almost certainly holds NAD83(1986), NAD83(HARN), NAD83(NSRS2007), and NAD83(2011) work side by side – realizations that differ by tenths of a foot and look identical on paper. Some files say which they are. Many don’t: the realization lives in a project note, a CAD template, a custom scale factor somebody baked in before GPS, or the memory of the person who ran the job. Feed a mislabeled file to a perfect transformation and you get a perfectly transformed wrong answer – shifted by exactly the sub-foot difference nobody documented, now fossilized in the new frame.

The firms that migrate cleanly will be the ones that can answer, file by file, “what datum, what realization, what epoch, and how do we know?” The firms that can’t will be doing forensic geodesy under deadline, on whichever job happens to surface the problem first – and the people who could have answered from memory retire a little every year.

What to do this fall

None of this requires panic, and none of it requires transforming anything yet. It requires knowing what you have:

1. Inventory the archive’s datums. For each era of work: which NAD83 realization, which vertical datum, which units, which zone. Write it down where the file lives, not in someone’s head.
2. Verify, don’t assume. Where the label is uncertain, check a known monument – the same ten-minute habit that catches mislabeled files today settles the realization question for good.
3. Capture the tribal knowledge now. The “Smith Farm scale factor” conversations need to happen while the people who remember are still down the hall.
4. Watch the beta, don’t marry it. Exercise the new tools on test data; hold production migrations until the frames are final and your software chain is proven against them.

Do those four, and the 2026 transition becomes a scheduling exercise. Skip them, and every old file becomes a small research project at exactly the moment everyone else’s files are too.

The ground isn’t moving. Make sure your firm’s memory of it doesn’t either.