A traverse is the backbone of almost every survey. Set it sloppy and every point you shoot off it inherits the error. Run it right and you get something rare and valuable: work that proves itself. That is the whole point of a closed loop. When you start and end on the same known point, the survey tells you, before you leave the site, whether you can trust it.
Why close the loop
An open traverse gives you coordinates. A closed traverse gives you coordinates and a report card. Because you return to a point whose position you already know, any error in your angles or distances shows up as a gap between where you started and where your numbers say you ended. That gap, the misclosure, is your proof. No gap you can live with, no confidence. It is the difference between "I measured it" and "I measured it and checked it."
Plan the stations before you set anything
Good closure starts in recon, not in the office. Walk the site and pick stations that give you three things:
Intervisibility. Every station has to see the one before it and the one after it, cleanly, without heat shimmer off a slab or a truck parking in your line.
Stable ground. A point in soft fill or on a slab that gets loaded tomorrow is a point that moves. Put marks where they will still be exactly where you left them.
Strong geometry. Avoid very small angles and wildly unequal leg lengths. A traverse full of skinny triangles and one 400-foot leg next to a 30-foot leg magnifies small measuring errors. Aim for stations that turn healthy, well-conditioned angles.
Occupy and measure with discipline
Set up over each station carefully. A traverse run off a rushed, half-level setup is compromised before you turn a single angle, so center and level with intent (that setup routine is its own field guide). Then, at every station:
Backsight the previous station, foresight the next. Turn your angles in sets, direct and reverse, and average them. One face of the instrument hides its own errors; two faces cancel most of them.
Measure every distance, and measure it both ways when you can. A distance shot forward and back that agrees is a distance you can trust.
Keep clean notes, station by station. The best error-catching tool on the job is a field book someone can actually read.
Check angular closure
For a closed loop, geometry gives you a free check. The interior angles of any closed polygon have to sum to (n - 2) x 180 degrees, where n is the number of sides. Add up your measured interior angles, compare to that theoretical sum, and the difference is your angular misclosure.
If that misclosure is within your spec (tighter work allows less), your angles are clean enough to trust and carry into the adjustment. If it is not within spec, you do not push it through. You find the bad angle and reshoot it, because no adjustment fixes a blunder, it only smears it across the whole loop.
Check linear closure
Angles are only half the story. Break each leg into its north-south component (latitude) and east-west component (departure). On a loop that returns to its start, all the latitudes should sum to zero and all the departures should sum to zero. Whatever is left over is your linear misclosure.
Turn that into a closure ratio: the perimeter of the traverse divided by the misclosure distance, expressed as something like 1:10,000. Bigger second number, better work. Compare it to the standard your job calls for. This ratio is your field quality check. It tells you, on site, whether the loop is good enough to trust and whether a leg is hiding a blunder you should reshoot before you pack up. What it is not is the adjustment itself.
Adjust it right: least squares
Once the loop checks out, how you adjust it matters. The old hand methods, the compass (Bowditch) rule chief among them, spread the misclosure around on a simple assumption and call it done. They work for a single loop on paper, but they treat every measurement as equally trustworthy and give you no real statistics. The professional gold standard today is a least squares adjustment.
A least squares adjustment solves the whole network at once, angles and distances together, finding the set of coordinates that best fits all of your observations. The key word is best: it weights each measurement by how good it actually is, so a tight, well-turned angle carries more say than a soft one, and a precise EDM distance more than a taped measurement. It handles redundant shots and tied-together networks the hand rules cannot, and it hands you what those rules never could: a residual on every observation, a real accuracy estimate (and error ellipse) on every adjusted point, and a statistical way to flag the one bad measurement hiding in the set.
In practice this lives in your software, in the data collector and the office package. Your job in the field is to feed it clean, redundant, well-checked observations. A least squares adjustment rewards good field discipline and exposes sloppy work. It does not rescue it.
Where it goes wrong
Almost every ugly traverse traces back to one of a few habits: a setup that was "close enough" over the point, angles turned on one face to save time, a leg measured once and never checked, or a misclosure that got balanced away instead of investigated. None of those feel like mistakes in the moment. They all show up later, in somebody else's rework.
The takeaway from the dirt
A closed traverse is a promise you make to every point you are about to shoot: this control is solid, build on it. The loop closing is how you keep that promise honestly instead of hoping. Precision downstream is wasted if the control underneath it was never checked, and it is worth remembering that your final numbers are only ever as good as the point you actually put the prism on. Good control and a clean read at the point are the same discipline, start to finish.
Shields Precision Solutions builds field-proven survey tools for crews who care about getting the number right the first time. See the lineup at shieldsprecision.com.