Traversing and Control Surveys

Introduction

A control survey establishes a reliable geometric framework from which other surveys, setting-out operations, mapping, and construction measurements can be performed. A traverse is one of the most widely used forms of horizontal control. It consists of a connected sequence of survey lines whose lengths and directions are measured. Traverses are particularly valuable on construction sites because they can follow roads, property boundaries, tunnels, pipelines, and building perimeters.

The quality of a construction survey depends not only on instrument precision but also on survey design, observation procedures, mathematical computation, independent checks, and appropriate adjustment. An unchecked traverse may transfer errors into building grids, earthwork quantities, structural alignment, and as-built records. This chapter develops the principles and calculations required to design, compute, assess, and adjust open and closed traverses, and explains how horizontal and vertical control networks support construction projects.

Learning Objectives

After studying this chapter, the student should be able to:

1. Distinguish between open and closed traverses and select an appropriate form for a project.
2. Measure and reduce traverse angles, bearings, and distances.
3. Compute latitudes, departures, coordinates, and closing error.
4. Evaluate angular and linear misclosure against specified tolerances.
5. Apply Bowditch and transit adjustments appropriately.
6. Explain the establishment and use of horizontal and vertical control networks.
7. Identify field, computational, instrumental, and professional sources of error.

1. Control Surveys and Traverse Geometry

A control survey provides points whose positions and elevations are known with sufficient accuracy for subsequent work. Control may be:

• Horizontal control, defining plan position by coordinates, bearings, or azimuths.
• Vertical control, defining elevations or reduced levels relative to a datum.
• Three-dimensional control, defining both position and elevation in a common reference system.

A traverse contains stations connected by measured lines. Stations are usually marked by durable nails, pegs, bolts, concrete monuments, or instrument pillars. Each line has a horizontal distance and direction. The direction may be expressed as a whole-circle bearing or azimuth measured clockwise from north, from 0° to less than 360°.

Open traverses

An open traverse begins at one known point and ends at another point, without returning to the starting point. It is suitable for roads, canals, transmission lines, railways, and pipelines. If both the starting and ending points have known coordinates and orientation, the traverse can still be checked as a controlled link traverse. An open traverse with no independent endpoint control has limited ability to reveal errors and should not be used for high-precision setting out without additional checks.

Closed traverses

A closed traverse either:

• returns to its starting point, forming a polygon; or
• begins and ends at two known control points.

A closed polygon provides checks on both angular and coordinate closure. A link traverse between known points provides a particularly strong check because the computed endpoint can be compared with its independently known position.

Traverse design principles

Good design precedes computation. Stations should be intervisible, accessible, stable, and distributed around the area or along the route. Very short lines beside very long lines should be avoided where possible because they can weaken angular determination. Stations should not be placed where they are likely to be disturbed by excavation, traffic, or construction plant. At least one connection to existing reliable control is needed when the traverse is to be integrated into a larger project network.

2. Directions, Angles, Latitudes, and Departures

Assume an azimuth α is measured clockwise from north and a horizontal distance is L. The rectangular components of the line are its latitude and departure:

Latitude is the north–south component; it is positive toward north and negative toward south. Departure is the east–west component; it is positive toward east and negative toward west. The formula assumes α is a whole-circle azimuth and L is a horizontal distance. If a reduced bearing is used, the signs must be assigned from the quadrant.

For a line from point A with coordinates (E_A, N_A) to point B:

where E is easting and N is northing, both normally in metres. The inverse relationships are:

The function atan2 is preferred because it identifies the correct quadrant. If its result is negative, 360° is added to obtain a whole-circle azimuth.

Worked Example 1: Components and coordinates

A line AB has azimuth 120°00′00″ and horizontal length 80.000 m. Point A has coordinates E = 1000.000 m and N = 2000.000 m.

Latitude = 80.000 cos 120° = −40.000 m approximately.

Departure = 80.000 sin 120° = +69.282 m approximately.

Therefore:

E_B = 1000.000 + 69.282 = 1069.282 m

N_B = 2000.000 − 40.000 = 1960.000 m

The line therefore runs southeast: eastward by 69.282 m and southward by 40.000 m. Rounding should be controlled so that displayed coordinates remain consistent with the precision of the observations.

3. Angular Observation and Closure

Traverse angles may be measured as interior angles, deflection angles, or directions to backsights and foresights. A total station normally measures horizontal directions and vertical angles, while a theodolite may be used with separately measured distances.

For a closed polygon with n sides, the theoretical sum of interior angles is:

The angular misclosure is:

The formula applies to a simple plane polygon whose interior angles are consistently observed. For a clockwise or counterclockwise traverse, the correct angle convention must be maintained. A common approximate allocation, when all angles have equal precision, is:

This equal correction is justified only when observations have approximately equal reliability. If different angles have different standard deviations, corrections should be weighted, with more reliable observations receiving smaller corrections.

For a closed traverse expressed through direction changes, the directional convention must be stated. Interior-angle sums should not be mixed with deflection-angle sums without conversion.

Field checks for angles

Angles should be observed on both faces where precision requires it, with repetition or rounds when specified. Face-left and face-right observations help reduce collimation, vertical-axis, and circle-graduation effects. The backsight and foresight should be clearly identified, and readings should be recorded immediately rather than reconstructed from memory.

4. Coordinate Closure and Linear Misclosure

For a mathematically closed traverse, the algebraic sums of latitudes and departures should be zero:

For a traverse between known points, the sums should equal the known coordinate differences:

The linear misclosure components are therefore:

For a closed polygon, ΔN_known and ΔE_known are zero. The magnitude of linear misclosure is:

A common relative precision indicator is:

It may be reported as 1 in the relative precision denominator, for example 1:5000. This ratio is useful for comparison with specifications, but it is not a complete statistical description of survey quality. Precision standards should be stated by the project or governing authority.

Worked Example 2: Closure

Suppose a closed traverse has total length 1,250.000 m, f_N = +0.018 m, and f_E = −0.024 m.

The relative precision is approximately:

Thus the result may be expressed approximately as 1:41,700, subject to the project’s rounding convention. Whether it passes depends on the specified tolerance and on the survey class required for the work.

5. Traverse Adjustment

Adjustment distributes observed misclosure so that the final coordinates satisfy the geometric conditions of the traverse. Adjustment does not remove blunders; a gross error must first be found and corrected or the affected observation rejected under documented procedures.

Bowditch or compass rule

The Bowditch rule is commonly used when angular and distance observations are of approximately comparable relative precision. The correction to the latitude of side i is:

The correction to its departure is:

where L_i is the length of side i and ΣL is the total traverse length. The adjusted components are:

The corrections sum to −f_N and −f_E, so adjusted coordinate closure is achieved. The rule is simple and practical, but it is an approximation and should not be represented as a rigorous least-squares solution.

Transit rule

The transit rule distributes latitude corrections in proportion to the absolute observed latitudes and departure corrections in proportion to the absolute observed departures:

It is sometimes selected when angular observations are considered more precise than distance observations. The assumptions should be stated because the choice is not universally appropriate.

Least-squares adjustment

For high-order control, a least-squares adjustment is preferred. It estimates the set of coordinates or observations that minimizes the weighted sum of squared residuals while satisfying the network conditions. In matrix form, the linearized observation model is:

where v is the vector of residuals, A is the design matrix, x is the vector of unknown corrections, and l is the vector of observed-minus-computed quantities. With weight matrix P, the normal equations are:

This method requires a suitable functional model, stochastic model, datum definition, and error analysis. It is commonly implemented in professional survey software, but the surveyor remains responsible for checking observations, constraints, residuals, and units.

6. Numerical Traverse Computation Procedure

A robust computation follows this sequence:

1. Confirm station names, observation units, instrument constants, atmospheric corrections, and datum information.
2. Check raw field notes and identify possible blunders.
3. Reduce slope distances to horizontal distances where necessary.
4. Compute or verify azimuths from known control and observed angles.
5. Compute the latitude and departure of every line.
6. Sum components and calculate angular and linear misclosure.
7. Compare misclosures with project tolerances.
8. Investigate failures before applying an adjustment.
9. Apply a documented adjustment method.
10. Compute adjusted coordinates and independently verify closure.
11. Report precision, residuals, control references, equipment, date, observer, and adjustment method.

Compact computation table

Coordinates are then propagated from a known starting point. The final coordinate must be compared with the known endpoint or starting point, not merely accepted because the spreadsheet produces a result.

7. Horizontal Control Networks for Construction

A construction control network commonly contains primary, secondary, and working control. Primary points are tied to a national, regional, or site datum and are established to higher accuracy. Secondary points densify the network. Working control points are placed near construction operations and are used for setting out.

Important design decisions include:

• the coordinate reference system and geodetic datum;
• the projection and units;
• the orientation or azimuth basis;
• redundancy and independent checks;
• monument stability and protection;
• required accuracy for structural, architectural, and earthwork operations;
• procedures for replacing disturbed points.

A network should contain redundancy. For example, a point determined from only one line and one angle has little internal evidence of error. Additional observations create geometric strength and allow a network adjustment and statistical testing.

GNSS can efficiently establish broad control, but satellite visibility, multipath, reference-station quality, coordinate-system transformations, and obstruction by buildings or plant must be considered. Total stations remain highly effective for local construction networks, especially where line-of-sight observations can be planned. A combined GNSS–total-station workflow is often appropriate.

8. Vertical Control and Levelling Networks

Vertical control establishes benchmarks with known reduced levels. Differential levelling determines elevation differences from backsight and foresight readings. For a setup:

Using the height-of-instrument method:

where BS is backsight, FS is foresight, HI is height of instrument, and RL is reduced level. All readings use the same length unit, normally metres.

For a closed levelling loop, the arithmetic check is:

The levelling misclosure is the difference between the observed and known elevation difference. If an adjustment is justified, a simple proportional correction may be distributed according to sight distance or number of instrument setups, depending on the adopted specification and error model. Equal distribution by setup count is appropriate only when setups have approximately equal reliability.

Benchmarks should be stable, clearly described, and protected from construction disturbance. Temporary benchmarks must be connected to permanent or approved control before being used for critical work. Reciprocal observations and balanced sight lengths reduce the effects of collimation error and curvature/refraction in differential levelling.

9. Applied Case: Illustrative Building-Site Control Network

Background

A contractor is preparing a five-storey building site on a constrained urban parcel. Existing cadastral control is available near the southern boundary, while excavation and retaining-wall construction will alter much of the site.

Problem

The design drawings use grid coordinates, but the site has no protected internal control. A single unverified traverse would risk transferring errors into column positions and foundation offsets.

Analysis

The survey team identifies two existing control points with reliable coordinates and establishes a closed traverse around the accessible perimeter. Four permanent monuments are placed outside excavation limits, and additional working points are installed on protected walls. A separate levelling loop connects two approved benchmarks.

The total station observes angles in two faces and measures repeated distances. The traverse is processed using azimuths, latitudes, and departures. Angular and coordinate closure are checked before adjustment. The resulting relative precision satisfies the project specification. A least-squares adjustment is then performed because the network contains more observations than the minimum required.

Approach and solution

The adjusted perimeter points become secondary control. Grid lines are set out from two independent stations, and each critical column line is checked from a different setup. Elevation benchmarks are transferred by balanced differential levelling. Before excavation, all monuments are photographed, described, and entered into the project control register. Weekly checks compare selected distances and directions; after heavy plant movement, the checks are repeated.

Outcome

The building grid is established from redundant control rather than a single traverse. When one working point is later disturbed, it is re-established from protected points without changing the project datum. The control register also provides a defensible basis for resolving discrepancies between contractor and designer measurements.

Learning points

The case demonstrates that control is a management system as well as a calculation. Stable monuments, documented datums, independent checks, and suitable redundancy are as important as the traverse adjustment formula.

10. Errors, Limitations, and Professional Practice

Errors may be classified as:

• Instrumental errors, such as collimation, compensator, EDM scale, prism-constant, and circle errors.
• Personal errors, including poor centring, inaccurate pointing, incorrect target identification, or transcription mistakes.
• Environmental errors, including heat shimmer, refraction, vibration, wind, and unstable ground.
• Systematic errors, which follow a pattern and may be modelled or corrected.
• Random errors, which vary unpredictably and are reduced through repeated, well-designed observations.
• Blunders, which are gross mistakes and must not be hidden by adjustment.

A mathematically closed traverse can still be wrong if an error has been introduced consistently or if the network is poorly connected. Closure is therefore a necessary check, not proof of correctness.

Professional practice requires preserving raw observations, documenting corrections, maintaining version control for coordinate files, and reporting any rejected measurements. The surveyor must not alter control coordinates merely to force agreement with construction work. Datum transformations and units must be communicated clearly to designers, contractors, and machine-control operators.

Artificial intelligence and automated processing can assist with anomaly detection, instrument-log review, feature extraction, and quality-control dashboards. However, AI outputs require professional validation. An automated system cannot independently determine whether a disturbed monument, wrong prism constant, or incorrect coordinate datum has been used unless those conditions are represented and checked.

Best-Practice Checklist

1. Define the datum, projection, units, and orientation before observing.
2. Design stable, intervisible stations with adequate redundancy.
3. Check instrument calibration, centring, levelling, prism constants, and atmospheric settings.
4. Observe angles and distances using procedures appropriate to the required accuracy.
5. Preserve raw data and field notes.
6. Compute closure before adjustment.
7. Investigate blunders before distributing misclosure.
8. Use an adjustment method consistent with observation quality and project requirements.
9. Verify adjusted coordinates independently.
10. Protect, describe, and periodically monitor control monuments.

Chapter Summary

Traversing provides a practical framework for establishing horizontal control. Open traverses are useful for route surveys but require endpoint control or independent checks if reliable accuracy assessment is needed. Closed traverses provide angular and coordinate closure conditions. Bearings or azimuths and distances are converted into latitudes and departures, which are accumulated to compute coordinates. Angular misclosure is assessed against the theoretical polygon angle sum, while linear misclosure is obtained from the vector sum of coordinate-component errors.

Bowditch adjustment distributes misclosure in proportion to line length, whereas the transit rule uses the relative magnitudes of latitudes and departures. Both are approximate methods whose assumptions must be stated. Least-squares adjustment is more suitable for redundant, high-order control networks. Vertical control is established through benchmarks and levelling loops, with arithmetic and closure checks. Reliable construction control depends on sound design, redundancy, stable monuments, appropriate technology, documented computations, and professional quality assurance.

Key Terms

Control survey: Survey establishing reference points for subsequent measurement and setting out.

Traverse: Connected sequence of survey lines with measured lengths and directions.

Open traverse: Traverse that does not return to its starting point.

Closed traverse: Traverse that closes on its starting point or on another known control point.

Azimuth: Horizontal direction measured clockwise from north.

Bearing: Direction expressed relative to a reference meridian, commonly within a quadrant.

Latitude: North–south component of a traverse line.

Departure: East–west component of a traverse line.

Angular misclosure: Difference between observed and theoretical angular sums.

Linear misclosure: Resultant positional error from latitude and departure misclosures.

Bowditch rule: Adjustment distributing component misclosures in proportion to line lengths.

Transit rule: Adjustment distributing components in proportion to absolute latitudes and departures.

Least-squares adjustment: Weighted estimation that minimizes the sum of squared residuals subject to network conditions.

Horizontal control: Control defining plan positions.

Vertical control: Control defining elevations.

Benchmark: Stable point with a known reduced level.

Reduced level: Elevation of a point relative to an adopted datum.

Redundancy: Extra observations that permit checks and error assessment.

Datum: Reference system or surface from which coordinates or elevations are defined.

Review Questions

Short Answer Questions

1. What is the difference between an open and a closed traverse?
2. Define latitude and departure for a traverse line.
3. State the theoretical interior-angle sum for a closed polygon with n sides.
4. What is the purpose of a benchmark?
5. Why must blunders be investigated before traverse adjustment?

Long Answer Questions

1. Explain the complete procedure for computing and adjusting a closed traverse.
2. Compare the Bowditch rule, transit rule, and least-squares adjustment.
3. Discuss the design requirements of a construction-site horizontal control network.
4. Explain how differential levelling establishes and checks vertical control.
5. Describe the main sources of error in traverse surveying and methods of control.

Analytical / Application Questions

1. A traverse has total length 1,800 m, latitude misclosure +0.030 m, and departure misclosure −0.040 m. Compute the linear misclosure and relative precision.
2. A line is 125 m long with azimuth 210°. Compute its latitude and departure and state their signs.
3. A building site has two known external control points but excavation will destroy internal stations. Propose a control-network design, observation checks, monumentation plan, and adjustment strategy.

Multiple Choice Questions with Answers

Question 1

For a line of length 100 m and azimuth 0°, what are its latitude and departure?

A. 100 m north and 0 m east

B. 0 m north and 100 m east

C. 100 m south and 0 m east

D. 0 m north and 100 m west

Correct Answer: A. 100 m north and 0 m east

Question 2

The theoretical sum of interior angles of a closed five-sided traverse is:

A. 360°

B. 540°

C. 720°

D. 900°

Correct Answer: B. 540°

Question 3

If a closed traverse has f_N = 0.03 m and f_E = 0.04 m, its linear misclosure is:

A. 0.01 m

B. 0.035 m

C. 0.05 m

D. 0.07 m

Correct Answer: C. 0.05 m

Question 4

The Bowditch rule distributes latitude and departure misclosures primarily in proportion to:

A. Interior angles

B. Line lengths

C. Station elevations

D. Azimuth values

Correct Answer: B. Line lengths

Question 5

For a line with azimuth 90°, the departure is:

A. Equal to zero

B. Negative and equal to the line length

C. Positive and equal to the line length

D. Equal to half the line length

Correct Answer: C. Positive and equal to the line length

Question 6

Which observation provides the strongest basic check for a closed polygon traverse?

A. Only the first distance

B. Angular and coordinate closure

C. The largest interior angle only

D. The number of field-book pages

Correct Answer: B. Angular and coordinate closure

Question 7

In differential levelling, the rise or fall between two points is calculated as:

A. BS + FS

B. FS − BS

C. BS − FS

D. HI + FS

Correct Answer: C. BS − FS

Question 8

A benchmark is best defined as a point with known:

A. Horizontal angle only

B. Reduced level

C. Prism constant

D. Traverse perimeter

Correct Answer: B. Reduced level

Question 9

Which method is generally most appropriate for a redundant high-order control network?

A. Unchecked coordinate propagation

B. Equal rounding of all coordinates

C. Least-squares adjustment

D. Ignoring misclosure

Correct Answer: C. Least-squares adjustment

Question 10

A relative precision of 1:10,000 means that the misclosure is approximately:

A. One unit for every 10,000 units of traverse length

B. 10,000 units for every one unit of length

C. One degree for every 10,000 metres

D. 10,000 degrees per metre

Correct Answer: A. One unit for every 10,000 units of traverse length