Latitudes and Departures
A closed traverse must come home, and the miss tells you how good the survey was.
Before this Lesson
A crew walks a lap around a four-sided lot, measuring every side and every direction, and finishes on the same nail they started from. Back at the office, someone replays that lap on paper, and the paper walk ends 0.20 ft from the nail. The nail didn't move. That 0.20 ft is the whole day's error, gathered into one short line. Finding it takes one question, asked of every course: how far north, and how far east?
A course is a step north and a step east
Walk a course from one station to the next, and you could have made the same trip in two straight moves: so far north or south, then so far east or west. The north-south move is the course's latitude, and the east-west move is its departure. As in Bearings and Azimuths, north and east are positive, and south and west are negative.
Those two moves and the course form a right triangle, with the course as the hypotenuse. An azimuth starts at north, so it sits between the course and the triangle's north side. That makes north the adjacent side and east the opposite side. With for the course's length:
If that looks backward, you're remembering Vectors and Components, where the cosine went with the x-axis. It's the same resolution of a vector. The only thing that changed is where the angle starts, and the cosine always belongs to the side the angle touches. The Handbook's Latitudes and Departures entry prints only the bare sign cross, with no formulas, so these two come with you to the exam.
An azimuth carries its own signs
Here's the nice part. Swing an azimuth past 90°, and the course starts heading south just as the cosine turns negative. Swing it past 180°, and the course starts heading west just as the sine turns negative. The trig functions match the sign cross in every quadrant, so there's nothing to look up.
Try 350 ft at an azimuth of 247°10′. With the calculator in degrees, use its DMS mode or convert the minutes first: 247 + 10/60 = 247.167°.
Both came out negative without being asked, and a southwest course should go south and west. That's what sketching it first is for: the sketch checks the signs before you trust them.
From a bearing: sizes from the angle, signs from the letters
A bearing's angle also starts from the north-south line, so the triangle is the same, with the cosine along the north-south side and the sine along the east-west side. The difference is that the angle never passes 90°, so its cosine and sine are always positive. They can only give you sizes. The signs come from the letters. With for the bearing's angle:
So S 35°20′ W for 240 ft gives sizes of 195.79 ft and 138.80 ft. The S makes the first negative and the W makes the second negative, so the latitude is −195.79 ft and the departure is −138.80 ft. Converting to the azimuth, 215°20′, gives the same two numbers, signs included. Take the signs from the letters or from the azimuth, but never from both, or you'll flip one that was already right.
Surveyors keep all of this in a course table: one row per course, with its direction, length, latitude, and departure. Sum the last two columns, signs included, and the totals say where the whole chain of courses ended up relative to its start.
Running the triangle backward
Hand someone the two legs of the triangle, and they can recover the course. The length is the hypotenuse. The bearing's angle sits against the north-south side, so the departure is opposite it and goes on top:
The Handbook doesn't print these in surveying terms. Its Polar Coordinate System entry has the general pair, and , with the angle measured from the x-axis, and that's why you can't just copy its ratio here.
Notice the absolute values. A course with a latitude of +300 ft and a departure of +400 ft heads northeast. One with −300 ft and −400 ft heads southwest. Their ratios are identical, so the arctangent gives 53°08′ for both. It only knows the ratio, and it only answers between −90° and 90°, so it can't tell you which way a course points. The signs can. Positive latitude is N and negative is S, positive departure is E and negative is W. So −300 ft and −400 ft make a 500 ft course running S 53°08′ W.
The closing course
A closed traverse ends where it began. Every step north is canceled by a step south, and every step east by a step west, so with perfect measurements:
That lets you find a side nobody measured, say one that runs through a pond. Sum the courses you have. Those sums say how far you are from the start, and the closing course has to undo exactly that:
The minus signs are the part that's easy to lose. The sums point from the start out to where you are, and the closing course runs the other way. Skip them, and you get the right length with a bearing exactly backward.
What's left over is the survey's error
A computed closing course closes the traverse by construction. It quietly absorbs every error in the other courses, so it can't tell you anything about them. To test a survey, you have to measure all the sides and see whether they come home.
So suppose the crew did measure DA, and got N 26°04′ W for 289.47 ft. That's a hair off the values computed above, which is what real measurements do. Add its row, and the sums no longer vanish.
| Course | Direction | Length (ft) | Latitude (ft) | Departure (ft) |
|---|---|---|---|---|
| AB | N 52°30′ E | 412.50 | +251.11 | +327.26 |
| BC | S 38°20′ E | 356.20 | −279.41 | +220.93 |
| CD | S 61°10′ W | 480.75 | −231.85 | −421.15 |
| DA | N 26°04′ W | 289.47 | +260.03 | −127.20 |
| Sum | 1,538.92 | −0.12 | −0.16 |
The paper walk ended 0.12 ft south and 0.16 ft west of the nail. Those two leftovers are the legs of one more right triangle, and its hypotenuse is the short line from where the measurements ended to where they should have ended. That's the linear misclosure:
Is 0.20 ft bad? Around a desk, it would be embarrassing. Across a county, it would be suspiciously lucky. What matters is the miss compared with the distance walked, and that's the relative precision, written 1:x:
Rounded, the lot closes at 1:7,700: about 1 ft of error for every 7,700 ft walked. The perimeter is the sum of the lengths, however many courses it took to walk them. The Handbook prints none of this. An exam question may state the relations, but plan on carrying them yourself.
The angles keep their own books
The measured angles get a check of their own. Pick one corner of a closed traverse and draw a line to every corner it isn't already joined to. An n-sided figure falls into triangles, each holding 180°, so the interior angles have to total:
The Handbook's Regular Polygon entry comes close, with one interior angle of a regular polygon, , in radians. Multiply by and convert to degrees, and you have this sum, which holds for any closed figure, regular or not.
Deflection angles follow a different rule, so (n − 2)180° is for interior angles only. Walk once around any closed traverse, and you finish facing the way you started, having turned through exactly one full circle. So the deflection angles sum to 360°, with any turn the other way, at a corner that dents inward, counted as negative.
No instrument reads perfectly, so the measured angles won't quite reach the required total. The difference is the angular misclosure, the measured sum minus the required sum. With no reason to distrust one angle over another, share it equally: every angle changes by the same , in whichever direction removes it.
How good is good enough?
A misclosure on its own only says what happened. Whether that's acceptable depends on the job, which states its minimum relative precision as a ratio. The Handbook prints no such standards, so the question has to give you one.
The ratio reads backward from most numbers. A standard of 1:5,000 allows 1 ft of miss per 5,000 ft walked, and 1:10,000 allows only half of that, so the larger x is the tighter requirement and the better survey. A traverse meets a standard when its own x is at least as large as the standard's. Our lot, at 1:7,700, passes a 1:5,000 job and fails a 1:10,000 one. Its 0.20 ft doesn't decide that alone: the same miss on a shorter lap would earn a smaller x.