Benchmarks
Part of Surveying
How to establish and use permanent elevation reference points that anchor all vertical measurements.
Why This Matters
When you need to know whether water will flow downhill to your field, whether a building site is above flood level, or whether a canal will drain properly, you need elevation measurements. But elevation only has meaning relative to a reference point. That reference point — a fixed mark of known elevation — is called a benchmark.
In the pre-industrial world, every major construction project began by establishing a local benchmark. The Romans used sea level as their ultimate reference and built benchmarks along roads to establish elevation relationships across their empire. Medieval cathedral builders used local benchmarks to ensure their foundations were level even when built on sloping ground over many years. Without a network of benchmarks, each new project had to start from scratch, and connecting the elevations of different structures was impossible.
A community that establishes and maintains a network of benchmarks gains an enormous practical advantage. A benchmark set today, used carefully, will still be valid for your grandchildren’s construction projects. It represents accumulated knowledge made permanent in stone.
What a Benchmark Is
A benchmark is a physical mark — usually a chiseled symbol cut into rock, a set bolt, or a marked stone — whose elevation above a chosen datum is precisely known and recorded. The datum is the reference level: sea level is ideal for large-scale work, but any locally chosen point works for a community survey as long as it is used consistently.
A benchmark has two components:
- The physical mark — durable, permanent, and precisely located
- The recorded elevation — documented in field notes and referenced back to the datum
Without both components, a benchmark is useless. The mark without the elevation record tells you nothing. The elevation record without the physical mark cannot be reproduced.
Primary benchmarks are the most stable and precisely determined points in your network. They should be on solid rock or deep foundations, away from areas subject to disturbance. There should be at least two primary benchmarks in any survey area, set far enough apart that if one is disturbed, the other remains.
Secondary benchmarks (also called temporary benchmarks or TBMs) are set as needed during construction or survey work. They reference back to primary benchmarks and may be abandoned when a project is complete.
Selecting Benchmark Locations
Good benchmark locations share several characteristics:
Geological stability: The ground beneath must not settle, heave, or shift. Bedrock is ideal. Deep-driven wooden posts in compacted gravel are acceptable. Avoid soft clay, filled ground, or slopes subject to landslides.
Accessibility: A benchmark that cannot be found is useless. It should be accessible year-round, not blocked by seasonal vegetation, water, or snow.
Visibility: For benchmarks used as instrument stations, you need clear sight lines to other stations in the network.
Safety from disturbance: Road construction, building projects, flooding, and agricultural work all destroy benchmarks. Choose locations that are unlikely to be disturbed: large boulders, bedrock outcrops, and the foundations of permanent structures.
Documentation landmarks: You need at least three permanent features within 50 meters from which you can describe the benchmark’s exact location. Trees die; use stone walls, rock outcrops, or building corners as reference objects. Measure and record the distance and direction to each.
Constructing a Permanent Benchmark
Rock benchmark: The most durable option. Find a bedrock outcrop or large embedded boulder. Clean the surface. Using a cold chisel and hammer, cut a horizontal notch about 10 cm long, 2 cm wide, and 1 cm deep. Cut a small vertical line crossing the center of the horizontal notch — the intersection is the exact benchmark point. The cross shape is the traditional benchmark symbol. Fill the carved marks with paint or pigment to make them visible.
Driven post benchmark: When no suitable rock exists, drive a wooden post at least 60 cm into the ground, with the top protruding only a few centimeters. Drive a nail into the exact center of the post top. Protect the post with stones or a small cairn. This is less permanent than rock but adequate for construction-era benchmarks.
Concrete benchmark: If concrete is available, dig a pit about 60 cm deep and 30 cm across. Cast a concrete cylinder in it, with a bolt or embedded nail at the top as the reference point. Allow to cure completely before using as a benchmark. Stone aggregate with lime mortar works if true concrete is unavailable.
Determining and Recording Elevation
Every benchmark needs a known elevation. For a first benchmark in a new area, you assign it an elevation. Choose a number that ensures no point you are likely to survey comes out negative — if the lowest likely point in your area is the river, and you set your benchmark at what you estimate is 10 meters above the river, assign the benchmark elevation 20.000 meters. This 20-meter cushion ensures that river level (roughly 10.000 in this system) and any point below it still has a positive elevation.
Once your primary benchmark has an assigned elevation, all other benchmarks are determined by differential leveling — measuring the height differences between points and calculating their elevations from the known starting value.
Documenting a benchmark:
BENCHMARK BM-1
Location: On exposed granite outcrop, approximately 2 m north of the old oak tree
at the east end of the common field.
Reference ties:
- 8.4 m bearing S 22° E from BM-1 to the notch in the oak tree
- 15.7 m bearing N 85° E to the northwest corner of the stone wall
- 23.1 m bearing S 5° W to the center of the well coping
Elevation: 38.420 m (datum: local; assigned value, not referenced to sea level)
Date set: [date]
Set by: [name]
Mark description: BM cross chiseled into bedrock, painted white
Building a Benchmark Network
One benchmark is a start. A network of benchmarks is far more valuable. A network allows:
- Independent checks (run levels from BM-1 to BM-2 and verify against the known difference)
- Redundancy (if one benchmark is disturbed, others survive)
- Coverage (different parts of the community have nearby references)
Leveling runs connect benchmarks. A leveling run is a series of instrument setups between two benchmarks, measuring height differences at each setup. The measured difference should match the difference in their recorded elevations. If it does not, an error has occurred and must be found.
Loop leveling runs from a benchmark back to itself, going around a circuit. The sum of all the measured height differences should equal zero. The actual sum, which will be slightly different from zero, is the loop misclosure. An acceptable misclosure for basic surveying is 12√K mm, where K is the circuit length in kilometers. For a 1 km circuit, acceptable misclosure is 12 mm. Larger discrepancies require re-measurement.
Adjusting the network: When a leveling network has been measured, small adjustments to each benchmark elevation are computed to make the system internally consistent. The adjustments are proportional to the length of each leveling run — longer runs accumulate more uncertainty and receive larger corrections.
Using Benchmarks in Practice
On a construction site: Set a temporary benchmark (TBM) near the work area by leveling from the nearest primary benchmark. Give the TBM an elevation and clearly mark it. Use this TBM to set foundation grades, check floor levels, and verify drainage slopes. When construction is complete, the TBM can be abandoned.
For flood assessment: Level from your benchmark network down to the river or stream. Calculate the elevation of the normal water surface, the high-water mark from past floods (stain lines on rocks, debris caught in trees), and the elevation of critical infrastructure. Anything below the 100-year flood line needs protection or relocation.
For canal and drainage design: A canal must fall continuously from intake to outlet. The required grade is computed from the elevation difference between start and end, divided by the distance. Benchmarks along the route tell you the actual ground elevation, revealing whether the canal needs to be cut deep, built on an embankment, or redirected.
Protect Your Benchmarks
The moment you establish a benchmark, tell people about it. A cairn or fence with a simple sign — “Survey Mark — Do Not Disturb” — prevents accidental destruction far better than any physical protection alone. Community knowledge is the best protection for infrastructure.