Reduce two-face horizontal turns across zero
Reduce the backsight and foresight separately on each instrument face, then average two clockwise turns that cross the circle zero.
Face I · 358°10′05″ → 12°40′11″ · clockwise turn 14°30′06″On this page
Record the four circle readings with their targetsAverage the turns in secondsKnow when to unwrap before taking an averageCarry the result into a traverseA backsight reads 358° and the foresight reads 12°. Direct subtraction gives a large negative value, although the operator turned a small angle clockwise. There is another complication when the instrument is reversed. Its circle reading changes by approximately half a revolution. Reduce each pair of readings on its own face before combining results.
Record the four circle readings with their targets
This synthetic single-round example assumes the same occupied point, the same backsight and foresight, a stable setup and a horizontal circle that increases clockwise. The readings are raw circle directions. They have not already been reduced by the collector.
| Face | Backsight | Foresight |
|---|---|---|
| I | 358°10′05″ | 12°40′11″ |
| II | 178°10′09″ | 192°40′07″ |
For face I, temporarily express the foresight as 372°40′11″. Subtracting the backsight gives 14°30′06″. For face II the subtraction is direct: 192°40′07″ minus 178°10′09″ = 14°29′58″. Both are clockwise turns from the same backsight to the same foresight. Reversing both readings on face II has already cancelled the nominal 180° face difference in the subtraction.
Average the turns in seconds
The face-I result is six seconds beyond 14°30′00″. The face-II result is two seconds short of that reference. Their average offset is (+6 − 2)/2 = +2 seconds. The mean observed turn is therefore 14°30′02″. The inter-face difference is eight seconds. Retain that difference as well as the mean, since the mean alone conceals how closely the two observations agree.
Check the arithmetic by applying the mean turn to each face’s own backsight. The predicted face-I foresight is four seconds below the observed value. The predicted face-II foresight is four seconds above the observed value. Equal and opposite residuals confirm the average, although they say nothing about whether the operator aimed at the correct target.
Know when to unwrap before taking an average
Values near 0° and 360° require particular care. If one corresponding reduced direction is 359°59′58″ and another is 0°00′02″, write the second temporarily as 360°00′02″. The midpoint is 360°00′00″, equivalent to 0°. An ordinary arithmetic mean of the original numerical values gives 180°, on the opposite side of the instrument.
This unwrapping assumes the quantities describe nearly the same direction. It does not justify treating a genuinely large disagreement as a short arc. Examine the expected geometry and target IDs first, especially when the instrument or software has already converted face-II observations to a common orientation.
Carry the result into a traverse
If the known direction from the occupied point to the backsight is 70°00′00″, the outgoing azimuth for this clockwise treatment is 84°30′02″. Use that absolute bearing and the measured horizontal distance as a course in the traverse calculator. The raw circle reading of 12°40′11″ is useful in the observation ledger, but it does not carry the control orientation by itself.
Collectors can store means as separate observations and can orient a second face using earlier backsight readings. Read the export fields before repeating this manual process. Combining already averaged values with the original observations can give the same sight unintended extra weight. For work with multiple rounds, document the round membership and follow the adopted rejection and weighting procedure. Keep instrument face, sight ID, raw circle direction, reduced turn and the decision to retain each sight in a form a second observer can reconstruct.
A useful field-book annotation for this pair is “14°30′02″ mean, face spread 8″, backsight and foresight identified on both faces.” If the observations fail the applicable tolerance, the individual face readings give the crew a precise target for review. Discarding them and retaining only the mean makes it impossible to distinguish a true repeated observation from a duplicated row.