Compare Transit-rule corrections on a north, east and diagonal course
Use absolute latitude and departure weights to balance a three-sided loop and see exactly how the result differs from length weighting.
North corrections use |latitude| · east corrections use |departure|On this page
Start from observed componentsBuild two columns of weightsTest the adjusted triangle before exporting coordinatesChoose the method before comparing the answerOn a due-north course, a Transit-rule adjustment gives no easting correction when the original departure is zero. A Bowditch adjustment can give that same course an easting correction. The difference comes from the denominator used to allocate each portion of the error. A small three-course ledger makes it visible.
Start from observed components
Here is a synthetic traverse in metres, already reduced to horizontal north and east components. The vectors follow the order traveled. Positive north and positive east increase the corresponding coordinate.
| Course | Observed ΔN | Observed ΔE |
|---|---|---|
| Mill–Post | +80.000 | 0.000 |
| Post–Stone | 0.000 | +60.000 |
| Stone–Mill | −79.960 | −60.030 |
The north sum is +0.040 m. The east sum is −0.030 m. The endpoint lies 0.050 m from Mill. The corrections across the circuit must therefore total −0.040 m north and +0.030 m east. Write those targets in the header of the adjustment sheet before calculating a weight.
Build two columns of weights
Sum the absolute north components: 80.000 + 0.000 + 79.960 = 159.960 m. Sum the absolute east components separately: 0.000 + 60.000 + 60.030 = 120.030 m. Absolute values measure how much of each axis a course contributes, even if the course runs south or west.
Each course receives the target north correction times its absolute north component divided by 159.960. For east, use its absolute east component divided by 120.030. Using the tiny signed residuals as denominators would produce meaningless weights. In a geometrically closed traverse a signed sum ideally vanishes.
| Course | North correction m | East correction m |
|---|---|---|
| Mill–Post | −0.020005 | 0.000000 |
| Post–Stone | 0.000000 | +0.014996 |
| Stone–Mill | −0.019995 | +0.015004 |
The northerly first course and southerly last course both receive negative north corrections. One becomes slightly shorter in its north component and the other slightly larger in its south component. Reversing the correction sign along with the direction of the observation would defeat the allocation.
Test the adjusted triangle before exporting coordinates
Start the computation at N 0, E 0. The adjusted Post coordinate is N 79.979995, E 0. The adjusted Stone is about N 79.979995, E 60.014996. The last adjusted displacement has the opposite components, returning to Mill. If you place Mill on actual control, add its starting north and east values to those running displacements. Translation leaves every weight unchanged.
Compare one course to the Bowditch calculation. Mill–Post has a length of 80 m out of approximately 239.986 m traveled. Its share of the absolute 0.030 m easting correction under Bowditch is around 0.010 m. The Transit-rule easting correction is exactly zero. Both complete adjustments can close the same measured circuit while placing Post at different coordinates.
Choose the method before comparing the answer
A method selected because it puts a particular monument where you expected it introduces an undocumented constraint. Identify which adjustment the survey specification calls for and preserve independent ties for evaluating the result. Historical instrument assumptions can help explain why two methods exist, but no choice based only on the name of the instrument establishes the quality of this particular work.
If every course has a zero component on one axis, the corresponding absolute denominator is zero. An exactly collinear set cannot support an ordinary two-dimensional polygon adjustment. Investigate the intended geometry rather than dividing by a small substitute value. The closure method reference gives the preceding reduction from bearings and distances. Keep its unadjusted output together with the worksheet so another person can reconstruct both stages.
For this triangle, the absolute component totals exceed either individual course component even though their signed sums nearly vanish. That is an immediate sanity check on the denominator cells when you copy the method into another spreadsheet.