CA Surveying common mistakes
CA Surveying · Understand the traps and how to avoid them.
Sets a corner on the first iron pipe or rebar located near a computed position, without testing whether it is the monument the record actually calls for.
The maxim 'monuments control over distances' gets memorized as unconditional, so any iron found close to the mathematical position is treated as the original corner - even when the type, size, or witness ties don't match, or when it is a later subdivision monument offset from the record corner.
What to do instead
Test each found monument on three axes before relying on it: character (type, size, and accessories match the record description), position (within a defensible tolerance of the record position and consistent with the other corners of the same map), and corroboration (lines of occupation, fences, old blazes, subsequent record ties). An uncalled-for monument is evidence to be weighed and explained, not a controlling call.
Goes to the field with only the subject parcel's deed or a printed assessor's map, and discovers during the survey that adjoining deeds, a recorded subdivision map, or corner records change the whole picture.
Fieldwork feels like the real work, and research feels like preparation for it; the deed in hand seems sufficient until a discrepancy forces a return trip that the exam clock does not allow.
What to do instead
Run research to exhaustion first: every deed in the chain of title, all adjoining parcels, recorded maps and parcel maps, corner records, highway and right-of-way maps, and any referenced unrecorded documents. Reduce it to a written record-vs-found schedule before any monument is set or relied on, and let that schedule drive the field procedure.
When a junior deed's calls convey more land than the parent tract contained, prorates the deficit proportionally across every line rather than identifying which conveyance is senior.
Proportional proration produces a tidy, closed mathematical figure and feels like the 'correct survey answer,' so the rules of seniority only get applied when a question literally uses the word 'senior.'
What to do instead
Establish the order of conveyances first, then hold the senior line as written and let the surplus or deficiency fall on the junior parcel. Before proration of any kind, re-read the deed for contrary intent - a call to the senior line, a call for area, or 'less' language - and state the seniority determination explicitly in the reasoning.
Writes up a record of survey or corner record with no stated basis of bearings and no ties from the monuments set to nearby record corners or existing monuments.
Documentation gets treated as boilerplate at the end of a long problem, so the parts that make the map self-sufficient are skipped in favor of finishing the computation.
What to do instead
Adopt the basis of bearings deliberately before writing anything - for example, the record line between two found monuments - state it on the face of the document, and tie every monument set or perpetuated to at least two independent references. A later surveyor must be able to reproduce the position from the document alone.
In slope staking and grade work, calls cut where the ground is actually below grade (or fill where it is above), and only catches the sign error after construction starts.
The cut/fill relation is remembered as a formula fragment rather than derived, so a flipped subtraction or a switch between a rod reading on the ground and a rod reading on a hub inverts the sign without any arithmetic alarm going off.
What to do instead
Derive it once from the instrument height: elevation = HI − rod, so grade rod = HI − design elevation and cut/fill = grade rod − ground rod, positive meaning cut. Keep one sign convention through the whole problem and check the result physically - the stake at a daylight point or on a known grade line should agree with what the ground is visibly doing.
Mixes arc length with chord length, or stationing with plain feet, in horizontal and vertical curve computations - computing a middle ordinate with the full central angle, or substituting the chord for the arc in L = RΔ.
The curve formulas look symmetric and are recalled as a block, while the exam supplies stations written as 12+45.50 and angles in degrees-minutes-seconds, so units and the half-angle factors slip without notice.
What to do instead
Sketch the curve, label Δ, T, L, C, M, E, and the PC and PT stations, and carry units explicitly through every line. Use Δ/2 in the tangent, external, and middle ordinate expressions, convert Δ to radians for arc length, and close the computation by checking that PC + L = PT and that the parts are mutually consistent.
Compares a measured or coordinate-inverse distance directly against a record distance with no transformation between ground and grid, applying no combined factor.
State plane scale factors and elevation factors sit within a few parts per million of unity, so the resulting discrepancy is small enough to be dismissed as measurement noise or, worse, mistaken for a blunder.
What to do instead
Compute the combined factor as the grid scale factor times the elevation factor (using ellipsoid height, not orthometric) and apply it in the correct direction for the comparison being made. When reconciling a record distance, do the inverse both ways - ground to grid and grid to ground - and show the factor so the magnitude of the correction is visible.
Publishes GNSS-derived elevations with no geoid model applied, or combines coordinate files of different datums and epochs - NAD27 with NAD83(2011), or mixed realizations - in one adjustment.
The controller and the point file carry a datum tag that is assumed current, and ellipsoid height versus orthometric height look interchangeable because both are labeled 'elevation' in the export.
What to do instead
Verify datum, realization, epoch, projection and zone, and units in the header of every file before anything is combined, and transform rather than ignore a mismatch. Convert ellipsoid height to orthometric with the appropriate geoid model (h = H + N) and state which height is being reported at each step.
Places contours by eye midway between spot elevations and computes the horizontal distance between contours as contour interval times slope.
Interpolation is done by visual habit, and the contour-spacing relation is recalled without a dimensional check, so a formula that should divide ends up multiplying and produces spacings that are obviously wrong on a sketch but not in a hurry.
What to do instead
Derive spacing from rise over run and check by units: horizontal distance between contours equals contour interval divided by slope in feet per foot. Interpolate linearly between the two bracketing shots, honor breaklines at ridges and swales, and apply the map-reading rules that catch errors - contours never cross, and they form a V pointing upstream at a stream.
Accepts the units and datum implied by a coordinate or point file - mixing US survey feet with international feet, or assuming an imported CSV is already in the project's datum and zone.
The difference between the two feet is about two parts per million, roughly a hundredth of a foot per mile, so it hides inside normal tolerance until it is compounded across a large project; and file headers are frequently stale or copied from a previous job.
What to do instead
Verify metadata (datum, epoch, units, projection and zone) against the source rather than the file name, apply the exact conversion where needed, and run a redundancy check before using the data - a repeat observation, a duplicate point number, or a check shot to a known monument. Treat any unit or datum field that cannot be traced to its source as unknown.