Two Monuments, One Uniform Shift
Your firm's 1998 ALTA survey set two 1-inch monuments, MC-1 and MC-2, 1,215.6 feet apart, with coordinates on NAD83 derived from a static GPS session and published on the recorded map to 0.01 ft. In 2024 you return with network RTK referenced to NAD83(2011), epoch 2010.0, using three CORS with stated horizontal network accuracy of 0.03 ft. Holding the 1998 published coordinates fixed in a least-squares adjustment produces residuals at both monuments of about 0.86 ft north and 0.12 ft east, and the relative vector between the two monuments differs by only 0.04 ft from the 1998 map. The title company wants the new map on the 1998 basis so the two maps can be overlaid exactly.
1. Diagnose the 0.86 ft north / 0.12 ft east residual pattern. What is the likely cause, and what do you do with the monuments?
Read the worked answer
Read the pattern, not the number.
- A nearly identical shift at both monuments is a rigid translation of the whole network. That is characteristic of a reference-frame or coordinate problem - datum realization (e.g., NAD83(1991)/HARN versus NAD83(2011)), epoch propagation, a projection or scale-factor difference, or a single-point-constrained earlier adjustment.
- It is not characteristic of monument movement, local distortion, or a blunder, which would show up as residuals of differing magnitude and direction and as a change in the relative vector.
- The relative check confirms this: 0.04 ft over 1,215.6 ft is about 33 ppm, so internal geometry is excellent.
Monuments hold. Coordinates are subordinate to the physical monument. You do not dig up, move, or "correct" MC-1 and MC-2 to match published values, and you do not hold the 1998 coordinates as fixed control over the found monuments.
Investigate before reporting: confirm the 1998 realization and epoch, re-run the transformation (e.g., HTDP) between realizations, check the projection and combined scale factor, and check whether the 1998 coordinates were constrained to a single point or to a passive mark in a different frame. Document whichever explanation fits.
2. The title company wants the new map on the 1998 coordinate basis. What do you publish, and what is the defensible way to satisfy them?
Read the worked answer
Publish the truth with a documented bridge.
- State the basis explicitly: datum realization, epoch, projection, units, and the geoid model used for any heights. "NAD83" alone is not a basis.
- Hold the monuments. The new positions are derived from MC-1 and MC-2 as found; the map notes the relation between the 2024 coordinates and the 1998 published values (roughly 0.86 ft north, 0.12 ft east).
- Provide a transformation, not a fudge. A similarity (Helmert) transformation derived from common, identified points is a legitimate tool for overlay; document the parameters, the points used, and the residuals. Applying it to coordinates for comparison is fine; applying it to move a boundary is not.
- Check accuracy against a stated criterion. For a relative positional tolerance of 0.07 ft + 50 ppm, the allowance at 1,215.6 ft is 0.07 + 0.06 = 0.13 ft. Your measured relative difference of 0.04 ft passes comfortably; say so, and say against which standard.
- Do not certify that the new map "matches" the 1998 map beyond the stated, quantified relationship.
3. The title company's GIS vendor asks for state-plane coordinates as a shapefile so the lender can compute a 25-ft setback from your boundary line. What must you tell them about scale, units, and use?
Read the worked answer
Grid is not ground. State plane coordinates are on the projection surface, and the distance between two published coordinates is a grid distance, not a distance measured on the ground.
- With a combined scale factor of, for example, 1.00012, a ground distance of 1,215.45 ft appears as a grid distance of 1,215.60 ft - about 0.15 ft on this one line, and more on long lines or in the aggregate across a large parcel.
- A 25-ft setback computed in grid units will be short of 25 ft on the ground where the combined factor exceeds 1, and long where it is less than 1.
- Projection convergence also rotates bearings away from true north by an amount that grows with distance from the central meridian; a GIS bearing is a grid bearing, not a record bearing.
What to tell them:
- Use the coordinate file for mapping and overlay only, not for construction layout or for setting a setback.
- If setback distances must be computed, convert with the stated combined scale factor and record the factor and its basis.
- The controlling geometry is the monumented boundary shown on the recorded map; a shapefile is a derived product with its own metadata and no monument status.
- State the metadata plainly: datum, epoch, projection, units, scale factor, and the accuracy of the source positions.