Walking a loop, and proving you walked it
A traverse is a chain of measured courses round a parcel. Each course carries a length and a direction, and the whole method rests on turning that pair into two components that can simply be added up.
Directions are booked as azimuths: — alpha — measured clockwise from north, from 0° to 360°. Its reverse is the back azimuth, , wrapped back into that range. Same line, looked along from the other end — and it is what the instrument is set to when the crew sights back to orient at a new station.
Each course is then split. The latitude is the north–south component: . The departure is the east–west component: . is the horizontal course length in metres and its azimuth. Cosine for latitude, sine for departure — and the reason is that azimuths are measured from NORTH, so a course pointing due north is all latitude, and . Swap the two and every plot closes beautifully in the wrong place. The signs are not decoration either: they carry south and west.
Round a closed loop the latitudes must sum to zero and so must the departures. They never quite do. The residuals and stand at right angles, so the gap between where the crew finished and the pin they started from is the hypotenuse: , the linear closure error in metres.
Then the number that actually gets reported. , where is the total perimeter walked and the precision denominator, written 1:N. A 0.20 m miss round a 1,200 m loop is 1:6,000 — good boundary work. The same 0.20 m round a 200 m loop is 1:1,000, and would be rejected. That is the whole argument for a ratio: error only means something next to the distance it accumulated over.
And read the ratio the right way round. is a denominator. One part in eight thousand is finer than one part in three thousand, so a bigger N is a better traverse. It reads backwards to the eye, and mistaking it is how a rejected loop gets signed off.