Traverse DeskGet COGO
Closure diagnosis

How far does a bearing error move the end of a line?

Convert a small directional error into lateral displacement, then compare an individual course error with a rotation of the entire traverse.

Worked example250 m sight · 40″ direction error · roughly 48 mm lateral
On this pageCalculate the exact separation firstEstimate the scale without a calculatorSeparate one wrong course from a rotated loopDecide which distance belongs in the estimate

A direction differs by 40 arcseconds between two reductions of the same observation. That number sounds small until you attach a distance. At 250 m the two endpoints are about 48 mm apart. On a 25 m line they differ by less than 5 mm. The angle is the same in both cases. The length of the sight determines the consequence.

Calculate the exact separation first

Place a station at the origin and aim a 250.000 m line due north. The intended endpoint is N 250.000, E 0.000 m. Swing the line clockwise by 40″ while keeping its measured length. First divide 40 by 3,600 to obtain 0.0111111°. For a trigonometric function expecting radians, multiply that degree value by π/180.

The new easting is 250 sin(40″), about +0.048481 m. The northing is 250 cos(40″), about 249.999995 m. In this example almost the entire movement is transverse. The northing change is only about five micrometres. Distinguish the sideways offset from the straight-line separation between the two possible endpoints. The exact endpoint separation is 2 × 250 × sin(40″/2). At this small angle both measures round to the same millimetre.

Estimate the scale without a calculator

One radian contains about 206,265 arcseconds. Multiply the length in metres by the angular discrepancy in arcseconds, then divide by 206,265. For the long sight that gives 250 × 40 / 206,265 = 0.04848 m. This approximation replaces the sine of a small angle by the angle itself expressed in radians.

Sight length Difference in direction Approximate transverse shift
25 m 40″ 0.00485 m
100 m 40″ 0.01939 m
250 m 40″ 0.04848 m

The table expresses a sensitivity, assuming the direction difference applies to that sight. It does not assign a statistical uncertainty to the instrument. A specification for angular precision, the measured difference between two readings, and a systematic orientation bias each describe different quantities.

Separate one wrong course from a rotated loop

Use this short experiment in the closure calculator. Start with a 250 m by 80 m rectangle using azimuths 0, 90, 180 and 270°. Change only the first azimuth to 0-00-40. Keep all four distances fixed. The loop develops a roughly 0.048 m gap. The other three courses still carry the original directions, so their vectors fail to cancel the altered first leg.

Now rotate all four azimuths by the same 40″. Opposite directions are again exactly 180° apart. Equal opposite lengths cancel and the loop closes. Every bearing has moved relative to the original reference. Closure detects internal inconsistency but gives no direct test of a common rotation. A separately observed control direction can expose that error.

Decide which distance belongs in the estimate

If a single absolute bearing was mistyped, use the length of that course for a first prediction. An error in an angle used to carry directions forward can rotate a whole downstream chain instead. Its effect at the terminal point depends on the displacement from the turning station to that terminal point. A long, folded route can end close to the turning station and show a small terminal shift even while several intermediate corners move substantially.

In a review note, name the point about which the rotation occurred, the direction being compared, the sight length, and whether one row or subsequent rows inherited the change. Compute the alternate endpoint in a separate copy. Reconcile the observed change in northing and easting against the prediction before asking the crew to revisit the instrument setup.

If the shortest reliable independent tie is at an intermediate corner, run the prediction for that corner as well as for the end. A smaller final gap can coexist with a larger intermediate displacement. The shorter list of numbers makes the difference practical to discuss at the instrument.

References

Check traverses offline with COGO