1.1 Principles and classification of surveying
After this lesson you should be able to:
- State the principle of working from the whole to the part, and say why a survey that begins with loose detail and tries to “join it up later” accumulates inconsistency.
- Name the complementary principles of an independent check and of matching the accuracy of the work to the purpose of the survey.
- Distinguish plane surveying from geodetic surveying, and compute the departure of a level surface from a tangent plane, δ = d² / (2R), using a labelled mean radius and the official WGS 84 semi-major axis.
- Classify surveys by purpose and by the method or instrument used, and locate chain, compass, levelling, traversing, tacheometry, photogrammetry and satellite positioning in later lessons rather than treating this page as their manual.
Theory
Surveying is the art and the science of determining the relative positions of points on, above or below the surface of the Earth, and of representing those positions to a stated accuracy. A road, a railway, a canal, a building plot, a reservoir and a property boundary all start from points whose positions must be known in a common framework. If that framework is weak, every later measurement — however carefully made — sits in the wrong place. This lesson is about the few principles that stop that happening, and about the way surveys are classified. Errors and their adjustment are Lesson 1.2; maps, scales and coordinate systems are Lesson 1.3; the field methods themselves begin in Chapter 2.
The first principle is to work from the whole to the part. A surveyor first establishes a sparse network of control points to a higher accuracy, over the full extent of the work, and only then fills in the detail from those points. The control may be a triangle, a traverse, a baseline with two stations, or a set of points fixed from a national network. The detail — fence corners, road edges, existing structures — is then measured from the control, not from other unforced detail. The reason is simple: a small angular or linear mistake made at the start of a long chain of detail measurements is magnified as the work proceeds, and is very hard to find. A mistake in a control measurement is caught when the control is closed or checked, before the detail is observed. The complementary principle is that every measurement should have an independent check: a second bearing, a second distance, a close of a traverse, a second staff reading. A single un-checked observation is not a result, it is a hope. A third principle is that the accuracy of the work should match the purpose. A cadastral boundary and a setting-out for a millimetre-sensitive machine foundation do not need the same tightness; nor should a reconnaissance survey be over-instrumented as if it were a primary geodetic baseline. Wasting effort on needless precision is not conservative; it is a different kind of error, and it often delays the checks that actually matter.
The Earth is not a plane. A distinction is therefore drawn between plane surveying, in which the surface is treated as a plane and the curvature of the Earth is neglected, and geodetic surveying, in which that curvature — and, when the work requires it, the shape of the reference ellipsoid — is taken into account. The two are not rival methods, and they are not ranked. Plane surveying is the everyday civil method on a site of limited extent: a campus, a highway contract, a town plan. Geodetic surveying is the method of national control networks, long tunnels and any work in which the departure of a level surface from a plane is no longer negligible compared with the accuracy required. No unverified “area limit” is used in this lesson as a switch between the two. The decision is made by comparing the geometric departure with the tolerance of the work. A first, transparent measure of that departure is the sagitta of a circular arc: if two points are a distance d apart, measured along a great-circle arc on a sphere of radius R, the departure of the level surface from the tangent plane at one end is, to a close approximation, δ = d² / (2R). That expression follows from the circle: the angle is θ = d / R, and the sagitta is R (1 − cos θ) ≈ R θ² / 2 = d² / (2R). The same geometry gives a much smaller arc-to-chord difference, about d³ / (24 R²), which is not the quantity that matters when a “flat-earth” plan is compared with a level surface. In the examples, a conventional mean Earth radius R = 6 371 km is used as a labelled working value. That figure is the IUGG mean radius. An official ellipsoidal check is the WGS 84 semi-major axis, a = 6 378 137 m exactly, as given in the WGS 84 technical literature (NIMA / NGA). The two radii differ by about one part in a thousand; they must not be mixed in a single arithmetic step without a reason, and neither is a “more official” method of surveying than the other. A national geodetic framework in India is maintained so that local plane surveys can be attached to a common datum; this lesson names that role, and does not treat a site traverse as a substitute for it.
Surveys are classified in two useful ways, without ranking the classes. By purpose one speaks of engineering or construction surveys (roads, railways, canals, buildings, setting out), topographic surveys (the shape of the ground), cadastral or land surveys (boundaries and areas of parcels), hydrographic surveys (underwater relief and shorelines), mine surveys, and route or alignment surveys. Astronomical observations and satellite positioning (GNSS) are how a survey is tied to a geodetic or geographic reference; they are methods as well as purposes, and they are taken up in the Geomatics lessons. By method or instrument one speaks, in the civil undergraduate sequence, of chain and tape surveying, compass surveying, plane-table surveying, levelling, theodolite traversing, tacheometry, photogrammetry, and electronic or satellite methods (EDM, total station, GNSS). Each of those methods has its own later lesson. The classification is a map of the subject, not a licence to practise a method from this page alone. Units of length are SI metres; angles may be reported in degrees, minutes and seconds or in gons, as the project specifies. Scales and coordinate systems belong to Lesson 1.3 and are not developed here.
Derivation
- Working from the whole to the part is a statement about error propagation: a control network that is itself closed or independently checked confines a later detail error to the neighbourhood of that detail, whereas a survey that begins at an unchecked end and “builds outwards” carries every early mistake into the far end.
- The departure of a spherical level surface from the tangent plane at one end of an arc of length d is the sagitta δ = R (1 − cos θ) with θ = d / R in radians. For θ small compared with 1, 1 − cos θ ≈ θ² / 2, so δ ≈ d² / (2R). This is a geometric approximation, not a geodetic projection formula, and it is used here only to size the curvature effect.
- The difference between the arc length d and the chord is d − 2 R sin(θ/2) ≈ d³ / (24 R²). At ordinary civil distances this is orders of magnitude smaller than the sagitta, so it is not the reason a plane survey becomes inadequate.
- With the working mean radius R = 6 371 km = 6.371 × 10⁶ m, a 10 km arc has δ = (10 000)² / (2 × 6.371 × 10⁶) = 7.848 m. With the official WGS 84 semi-major axis a = 6 378 137 m, the same formula gives δ = 7.839 m. The two figures agree to about a centimetre at 10 km; they are not interchangeable in a calculation that claims official ellipsoidal precision.
- The corresponding arc-to-chord difference at 10 km with R = 6 371 km is (10 000)³ / (24 × (6.371 × 10⁶)²) = 1.027 mm.
Departure of a level surface from a tangent plane (spherical sagitta): δ = d² / (2R) (m), with d the arc length (m) and R the Earth radius (m)Exact sagitta on a sphere: δ = R (1 − cos(d / R)), with d / R in radians; the d² / (2R) form is the small-angle approximationArc-to-chord difference, small angle: d − chord ≈ d³ / (24 R²) (m)Working mean radius in the examples: R = 6 371 km (IUGG mean radius, labelled working value)Official ellipsoidal check: WGS 84 semi-major axis a = 6 378 137 m exactlyPlane surveying: curvature neglected. Geodetic surveying: curvature, and where needed the ellipsoid, taken into account. The switch is made by comparing δ with the required tolerance, not by a fixed area rule in this lesson
- The sagitta formula treats the Earth as a sphere of a stated radius. The Earth is better described as an ellipsoid; the WGS 84 semi-major axis is used only as a check on the order of δ, not as a complete ellipsoidal computation of geodetic height or of a map projection.
- The small-angle form δ = d² / (2R) is for d much smaller than R. At 10 km, θ is about 0.0016 rad, and the approximation is ample for this lesson.
- The 6 371 km figure is a labelled working mean radius (IUGG). It is not the WGS 84 semi-major axis, and the two are not to be mixed in one arithmetic step without a reason.
- No fixed area or distance is used here as a legal or code limit between plane and geodetic work. The comparison is always δ against the tolerance of the particular survey.
- Classifications by purpose and by method are teaching groups. A single project may use more than one method. The methods are not ranked.
- No Indian Standard clause is cited. Official BIS surveying-principles texts were not on the box at writing, so no clause numbers are invented.
- This lesson does not set out a traverse, does not choose a map projection, and does not copy a textbook figure or table.
Codes and standards
Indian Standards are cited by number and clause only; no clause text, tables or figures are reproduced. Other standards are listed for comparison, not ranked.
- No Indian Standard clause is used for the principles or for the plane/geodetic distinction; those are surveying practice and geometry. Official BIS texts of surveying standards were not on the box at writing.
- IUGG mean radius of the Earth, 6 371 km — labelled working value of R in the examples.
- NIMA / NGA, World Geodetic System 1984 — official semi-major axis a = 6 378 137 m exactly, used as a check on the sagitta, not as a map projection.
- Background textbooks, not ranked, editions as in the site references register: S. K. Duggal, Surveying, Vol. 1, McGraw Hill Education (India), 4th edition, 2013; B. C. Punmia, A. K. Jain, A. K. Jain, Surveying, Vol. 1, Laxmi Publications, 17th edition, 2016; C. D. Ghilani, Elementary Surveying: An Introduction to Geomatics, Pearson, 16th edition, 2022 (published 2021).
Worked examples
Problem. Two points lie 10.00 km apart, measured along a great-circle arc. Taking the labelled working mean radius R = 6 371 km, compute the departure δ of the level surface from the tangent plane at one end, using δ = d² / (2R). Then recompute δ with the official WGS 84 semi-major axis a = 6 378 137 m. Say, in one sentence, whether the two figures may be treated as the same number in a calculation that claims official ellipsoidal precision.
- d = 10.00 km = 10 000 m. R = 6 371 km = 6.371 × 10⁶ m.
- Working value: δ = d² / (2R) = (10 000)² / (2 × 6.371 × 10⁶) = 1.000 × 10⁸ / 1.2742 × 10⁷ = 7.848 m.
- With the official WGS 84 semi-major axis, δ = (10 000)² / (2 × 6 378 137) = 1.000 × 10⁸ / 1.2756274 × 10⁷ = 7.839 m.
- The two figures differ by about 9 mm at 10 km. They are of the same order, and either is enough to show that a 10 km “flat” plan sits several metres off a level surface, but they must not be mixed, or reported as identical, in a calculation that claims official ellipsoidal precision. The working 6 371 km radius is an IUGG mean value, not the WGS 84 axis.
Answer. δ = 7.848 m with R = 6 371 km; δ = 7.839 m with the WGS 84 semi-major axis. The two are of the same order but are not the same official number.
Problem. For the same 10.00 km arc and the same working radius R = 6 371 km, compute the arc-to-chord difference d³ / (24 R²), in millimetres. Then compute the sagitta for a 1.00 km arc with the same R. A site survey over 1 km has a vertical tolerance of 20 mm. Is the curvature departure, by itself, negligible compared with that tolerance, or must it be considered?
- Arc-to-chord difference at 10 km: d³ / (24 R²) = (10 000)³ / (24 × (6.371 × 10⁶)²) = 1.000 × 10¹² / 9.7415 × 10¹⁴ = 0.001027 m = 1.027 mm. This is about 7 600 times smaller than the 7.848 m sagitta, so the “flat-earth” problem at this distance is the departure of the level surface from the plane, not the difference between arc and chord.
- For d = 1.00 km = 1 000 m, δ = (1 000)² / (2 × 6.371 × 10⁶) = 1.000 × 10⁶ / 1.2742 × 10⁷ = 0.07848 m = 78.5 mm.
- The stated vertical tolerance is 20 mm. The sagitta 78.5 mm is larger than 20 mm, so the curvature departure is not negligible compared with that tolerance, even over 1 km, if the work is being compared with a tangent plane at one end. (A survey that uses a level surface, or a proper height datum, as its reference is a different matter: the sagitta is then already part of the definition of height, and the remaining plane-survey question is about horizontal position.)
- The comparison is with the tolerance of the work, not with a fixed area rule. A 20 mm tolerance and a 200 mm tolerance would lead to different decisions at the same distance.
Answer. Arc-to-chord at 10 km = 1.027 mm, much smaller than the 7.848 m sagitta. Over 1 km, δ = 78.5 mm, which is not negligible against a 20 mm vertical tolerance if the reference is a tangent plane.
Step-by-step procedure
- Write down the purpose of the survey and the tolerance that purpose requires. That tolerance is what later choices are compared with.
- Establish, or connect to, a control framework that covers the whole of the work before any detail is collected. Close or otherwise independently check the control.
- Fill in the detail from the control, not from unchecked detail. Provide an independent check on every measurement that will be used as a result.
- Decide whether curvature can be neglected by computing a sagitta δ = d² / (2R) over the longest line in the work, with a stated radius (the working 6 371 km, or an official ellipsoidal axis if one is required), and comparing δ with the tolerance. Do not apply a memorised area limit in place of this comparison.
- If the work must sit in a national or project geodetic framework, connect the local control to that framework by the method specified for the project; do not invent a local “false north” that cannot be joined later.
- Choose instruments and methods (chain, compass, level, theodolite, EDM, GNSS) from the later lessons so that they can meet the stated tolerance, and no more than is needed.
- Record the datum, the units and the checks with the observations. A survey without those notes cannot be extended or checked later.
Site notes
- A building-plot survey that starts at one corner and works around the fence, without a measured diagonal or a second baseline, is the opposite of working from the whole to the part. Put in two or three control stations first.
- A long pipeline or canal is “limited in width” but not in length. The relevant distance d for the sagitta is the length of the job, not the width of the strip.
- Do not treat a handheld GNSS reading of a fence corner as control of the same quality as a closed traverse or a connected geodetic point, unless the project has specified that method and its accuracy has been checked.
- When two contractors meet at a boundary, both must be on the same control and the same datum. “We each started from our own peg” is how a gap or an overlap appears.
- A reconnaissance can be run at a lower accuracy, but it should still be checked, and it should not be silently upgraded to setting-out accuracy in the office.
- Heights and plan positions have different curvature effects. Do not assume that a survey which is adequate in plan is automatically adequate in level, or the reverse.
- Starting the detail at one end of the site and hoping to close at the other, instead of laying down checked control first.
- Taking a single un-checked tape length or bearing as a result, or “checking” a measurement by reading the same instrument a second time without changing the set-up (that is a repeat, not an independent check).
- Applying a memorised area figure as a hard limit between plane and geodetic surveying, instead of comparing the sagitta with the tolerance of the work.
- Reporting the 7.848 m working sagitta at 10 km as if it had been computed from the official WGS 84 axis, or mixing 6 371 km and 6 378 137 m in one line of arithmetic without a reason.
- Treating the arc-to-chord difference (millimetres at 10 km) as the curvature error that matters, and ignoring the several-metre sagitta of the level surface.
- Ranking plane surveying against geodetic surveying, or ranking one textbook or one instrument class against another.
- Inventing an Indian Standard clause number that has not been read in an official text, or citing a book edition that has not been checked.
Work from the whole to the part: establish checked control first, then the detail. Give every result an independent check, and match accuracy to the purpose. Plane surveying neglects curvature; geodetic surveying takes it into account. Size the effect by δ = d² / (2R). Working R = 6 371 km (IUGG mean); official WGS 84 a = 6 378 137 m. Compare δ with the tolerance; do not use a fixed area rule here.
Summary
- Surveying determines the relative positions of points, to a stated accuracy, in a common framework. The framework is laid down first.
- Work from the whole to the part; give every result an independent check; and match the accuracy of the work to its purpose.
- Plane surveying neglects the curvature of the Earth; geodetic surveying takes that curvature, and where needed the ellipsoid, into account. The departure of a level surface from a tangent plane is, approximately, δ = d² / (2R).
- Over 10 km, that departure is 7.848 m with the working mean radius 6 371 km, and 7.839 m with the official WGS 84 semi-major axis 6 378 137 m. The arc-to-chord difference at the same distance is only about 1 mm.
- Surveys are classified by purpose and by method. The methods themselves are later lessons. No Indian Standard clause is cited in this lesson, because no official BIS text of those standards was on the box at writing.
Practice questions
Tap a question to see the answer.
1State the principle of working from the whole to the part, in one sentence.
Establish a checked control framework over the whole of the work first, and only then measure the detail from that control.
2Two points are 5.00 km apart. With the working radius R = 6 371 km, what is the sagitta δ = d² / (2R)?
d = 5 000 m; δ = (5 000)² / (2 × 6.371 × 10⁶) = 2.5 × 10⁷ / 1.2742 × 10⁷ = 1.962 m.
3What official length is the WGS 84 semi-major axis?
6 378 137 m exactly.
4Why is the arc-to-chord difference not the main reason a plane survey becomes inadequate over several kilometres?
Because that difference is only about 1 mm at 10 km, while the departure of the level surface from the tangent plane is several metres. It is the sagitta that must be compared with the tolerance.
5A topographic survey of a campus and a cadastral survey of the same ground are proposed. Which principle says they need not be observed to the same tightness?
The accuracy of the work should be matched to the purpose of the survey.
6Name three methods, in the civil undergraduate sequence, by which a survey may be classified, other than by its purpose.
Any three of: chain and tape, compass, plane table, levelling, theodolite traversing, tacheometry, photogrammetry, EDM or total station, GNSS. (Each has a later lesson.)
References
Standard texts, listed in no order of merit. Our explanations, examples and diagrams are original; nothing is copied from these books.
- S. K. Duggal, Surveying, Vol. 1, McGraw Hill Education (India), 4th edition, 2013.
- B. C. Punmia, A. K. Jain, A. K. Jain, Surveying, Vol. 1, Laxmi Publications, 17th edition, 2016.
- C. D. Ghilani, Elementary Surveying: An Introduction to Geomatics, Pearson, 16th edition, 2022 (published 2021).
- IUGG mean radius of the Earth, 6 371 km, used as the labelled working radius. NIMA / NGA, World Geodetic System 1984, official semi-major axis a = 6 378 137 m exactly.