An NGS datasheet is the published record for a survey control mark. It is a dense, fixed-format text document that has changed remarkably little in decades, and it contains everything you need to decide whether to hold the mark, check it or ignore it — provided you can read it.
Most of the errors people make with datasheets come from reading the coordinate and skipping everything that qualifies it. The qualifiers are the point of the document.
The identity block
At the top the mark is identified two ways. The designation is the name stamped on the disk or assigned by the setting agency, and it is not unique — many marks share a designation. The permanent identifier, or PID, is a six-character alphanumeric code that is unique across the entire database. Always cite the PID. Citing only a designation is how the wrong mark ends up in a report.
Alongside these you get the state and county, the mark type, the setting agency and the date it was monumented. The mark type matters more than it looks: a disk in bedrock, a disk in a concrete post, a rod driven to refusal with a sleeve, and a disk in the top of a bridge abutment have very different stability characteristics, and the vertical value is only as good as the setting.
Adjusted versus scaled — the single most important distinction
The same logic applies vertically. An orthometric height marked ADJUSTED came from levelling in the national adjustment. One marked GPS OBS came from GNSS observation and a geoid model. One marked SCALED or read from a map is an approximation for indexing only. A vertical value derived from a topographic map may be off by several meters.
| Qualifier | Applies to | Trust level |
|---|---|---|
| ADJUSTED | Horizontal or vertical | Full — came from a least-squares network adjustment |
| GPS OBS | Vertical | Good — GNSS observed, converted with a geoid model; check which model |
| SCALED | Horizontal | Indexing only — read from a map, tens of meters of uncertainty |
| POSTED / VERT ORDER estimates | Vertical | Approximate — a derived or estimated value, not a levelled one |
The datum, realization and epoch
The horizontal coordinate is labelled with its datum and realization, for example NAD 83 (2011), together with the epoch the value is valid for, such as 2010.00. The vertical value is labelled with its datum, normally NAVD 88. Both labels must be carried into your project. A northing and easting copied off a datasheet with the realization left behind is the beginning of a mismatch that will surface later.
The datasheet also gives ellipsoid height where one was determined, and the geoid height from the current model at that location. Those two together with the orthometric height let you check h = H + N directly, which is a good sanity test on your own processing.
Order and class
Classical control carries an order and class — first order, second order class I or II, third order class I or II — which describe the accuracy standard the observations met when the mark was established. Higher order means tighter tolerances on the observations and the adjustment.
These labels describe the survey as performed, not the mark as it stands today. A first-order mark set in 1954 in ground that has since subsided is still labelled first order. Order tells you about the pedigree of the number; it tells you nothing about whether the disk has been hit by a grader.
GNSS-derived positions instead carry network accuracy and local accuracy figures at the 95 percent confidence level. Network accuracy describes the position relative to the datum as a whole; local accuracy describes it relative to neighbouring marks. For a project that needs internal consistency, local accuracy is often the number that matters.
Superseded values and the history
Below the current values a datasheet lists superseded survey control — the previous published coordinates for the mark, on earlier realizations. This section is genuinely useful when you are reconciling old work. If a 1990s survey used a value from this mark, it is likely in the superseded list, and finding it there explains a discrepancy in seconds that could otherwise take a day.
The description and recovery notes
The station description is a written route to the mark, often written decades ago, using landmarks that may no longer exist. Read it in reverse chronological order with the recovery notes, which record each time someone looked for the mark and what they found. A note reading recovered in good condition with a recent date is worth a great deal. A note reading not found, or destroyed, is worth even more, because it saves you the afternoon.
Recovery notes accumulate rather than replace, so the sequence tells a story: found in good condition, found disturbed, not found. Read the whole sequence, not just the latest line.
A practical reading order
- PID and designation — record both, cite the PID.
- The determination qualifiers — is the horizontal adjusted or scaled, is the vertical levelled, GNSS-derived or estimated?
- Datum, realization and epoch for the horizontal; datum for the vertical.
- Accuracy figures — network and local, or order and class for classical control.
- Superseded values — needed the moment you touch older work.
- Mark type and setting — how stable is this thing physically?
- Recovery notes, latest first, then the full sequence.
- The description, read last and read on site.
One habit is worth more than all of the above: check the mark against another mark before you hold it. Occupy two published marks, compute the inverse between them, and compare against the published values. If they agree, you have confirmed both marks and your own setup at once. If they do not, you have found the problem before it was built into the project rather than after.