Survey manual 2009

It shows not just how to press buttons on a total station but why misclosures happen and how to hunt them down. Read it before your first solo job; keep it on the site desk afterwards.

1. 1Early chapters on field records

   The field book is the only witness you get

   A crew finishes a long day of levelling. The instrument is packed, the light is going, and someone asks what the reading was at the last change point. The field book answers - or it doesn't. In this manual the book is not a notebook, it is the legal and technical record of the survey. Sketches, readings, weather, who held the staff: it goes in at the moment of observation, in pencil, never copied up later from memory.

   The idea

   Memory is the least reliable instrument on site. The manual treats the field book as the survey itself - the theodolite merely feeds it. That is why erasures are forbidden and corrections are struck through with a single line so the original stays legible. Years later, when a boundary is disputed or a setting-out is queried, the book is the only thing that can speak.

   Try this: Rule a page of your notebook into the standard levelling layout and force yourself to book every reading the second you call it - no scraps of paper, no end-of-day transcription.

2. 2Chapters on theory of errors

   Every measurement carries an error budget

   Two crews measure the same baseline. One reports a distance to the millimetre; the other reports it with a stated probable error. The manual's point is that the second crew is the professional one. Systematic errors - tape too long, temperature off standard - can be corrected. Random errors can only be weighted and averaged. Blunders, the third kind, are not errors at all: they are mistakes, and the only defence is independent check.

   The idea

   Precision without knowledge of error is decoration. The book's framework separates systematic, random and gross errors because each demands a different response: calibrate, average, or repeat. What surprises newcomers is that a good survey is not the one with the smallest numbers on the page but the one whose residuals you can explain. An unexplained closure that looks perfect is more frightening than an honest small misclosure.

   Try this: Next time you quote a measurement, attach its expected error - and be ready to say which of the three error classes dominates it.

3. 3Chain surveying chapters

   Chain survey: the whole map from triangles

   No angles, no electronics - just a tape, ranging rods and discipline. The manual builds a survey by dividing the ground into triangles, because a triangle is fixed once its three sides are known. Survey lines are laid out with check lines to every station, and the draughtsman later plots each triangle from the three taped sides. The stakes are real: one unmeasured check line and a whole triangle can be plotted wrong without anyone noticing.

   The idea

   The triangle is the oldest quality-control device in surveying. Three sides overdetermine the shape, so a bad tape reading shows up as a distorted plot. The book insists on check lines and on avoiding ill-conditioned, skinny triangles - the ones where a small measurement error swings the apex wildly. It is geometry doing the work that electronics does today.

   Try this: Sketch a small site as triangles and mark a check line for each one before you measure anything - the habit of planning checks first is the whole method.

4. 4Levelling chapters

   Levelling lives or dies at the closure

   A level run leaves a benchmark in the morning and must return to one - or finish on another of known height - by evening. The staff is held on change points, the instrument moves forward, and readings accumulate page by page. At the end the sum of backsights minus the sum of foresights must equal the difference between the first and last reduced levels. When it doesn't, the run has a misclosure, and the manual shows how to trace it to a change point.

   The idea

   The arithmetic check - backsights minus foresights equals final minus initial reduced level - is free, instant and catches most booking blunders before anyone leaves the field. What it cannot catch is a staff held on the wrong point at both ends of a bay, which is why the book teaches closed loops and repeated runs. Distributing the misclosure across the run is a judgement, not a ritual.

   Try this: Add the rise-and-fall check to your next levelling run and compute it in the field, before you pack the instrument away.

5. 5Theodolite and angle measurement chapters

   The theodolite: repeat, read, resection

   Angles are measured by repetition: the instrument is sighted left, then right, then left again, accumulating the angle on the circle so that reading errors are diluted by the number of turns. Face left and face right observations are averaged to cancel collimation and trunnion-axis errors. The manual then shows the payoff - intersection of two observed angles to fix a point, and resection, where three known points fix the instrument's own position.

   The idea

   Repetition is the surveyor's version of averaging: the circle reading is coarse, but the accumulated angle over several turns is fine. Face left and face right is the same idea applied to instrument geometry - the errors are built to cancel when both faces are used. Resection is the quiet miracle: stand anywhere, sight three known points, and the map tells you where you are.

   Try this: On your next angle job, double every angle by repetition and halve the total - compare the result with your single measurement to see your own reading error.

6. 6Traverse surveying chapters

   Traverse: the framework that carries everything

   A traverse is a chain of measured lines and angles that starts and ends on known stations. The manual closes the figure, computes the algebraic sum of the interior angles against the theoretical total, and distributes the angular misclosure. Then come the latitudes and departures: the eastings and northings of each leg. If the traverse is honest, they sum to zero. When they don't, the closing error is distributed in proportion to each leg's length.

   The idea

   A traverse is a self-auditing structure. The angular check catches booking errors; the coordinate check catches distance errors; and the ratio of closing error to perimeter tells you whether the work is acceptable or must be redone. Bowditch's rule - distributing by line length - survives because it is fair: longer legs, more error. The whole edifice of a site plan hangs on this one closed loop.

   Try this: Compute the closing error as a ratio of the perimeter on your last traverse and write it on the plan - if it is worse than 1 in 5,000, plan a re-run before plotting.

7. 7Contouring chapters

   Contours: the ground made readable

   A contour is a line of constant level, and the manual treats it as both a survey product and a design tool. Spot levels are taken at grid intersections or along the lines of a survey, then contours are interpolated between them. The vertical interval is chosen to suit the ground and the map scale - close together on steep slopes, wide apart on flat. From the contour plan, engineers read cut-and-fill, drainage and gradients at a glance.

   The idea

   Contours convert thousands of point heights into a picture the eye can parse instantly. The craft is in the interpolation and in choosing a vertical interval that shows the ground's character without cluttering the sheet. What surprises people is how much design information hides in the spacing: where lines crowd, water runs and machines struggle; where they spread, the land is buildable.

   Try this: Take any contour plan and trace where the lines are closest - that is the steepest line of drainage, and the first thing to check before siting anything on the slope.

8. 8Setting-out chapters

   Setting out: from drawing to pegs in the ground

   The drawings are approved; now the building must appear on the ground at the right position, level and orientation. The manual walks through transferring the plan's lines to pegs, establishing a level from a benchmark, and checking diagonals and offsets before the excavator arrives. Curves get special treatment: tangent lengths and deflection angles set out so the curve can be checked from several points, not just one.

   The idea

   Setting out is the moment surveying stops being descriptive and becomes prescriptive - the numbers now direct machines and concrete. The book's discipline is redundancy: every peg checked from a second reference, every level closed back to the benchmark. A setting-out error discovered after foundations are poured is measured in money, not millimetres.

   Try this: Sketch the rough layout of your project on paper, then this week drive actual pegs into the ground (or mark equivalent positions with stakes/string) to transfer one drawn line from plan to physical space.

9. 9Curve ranging chapters

   Curves: geometry you can drive on

   A road or railway changes direction through a curve, and the surveyor's job is to put that curve on the ground peg by peg. The manual sets out the vocabulary - tangent length, deflection angle, radius - and then the two field methods: linear, by offsets from the tangent, and angular, by deflection angles from a theodolite. Each peg is set from the previous one, and the last one must land on the far tangent point or the curve is wrong.

   The idea

   Curve work is where traverse discipline meets construction speed. The deflection-angle method wins in the field because each sight is independent of the last - an error at one peg does not corrupt the next. The check that the final peg lands on the second tangent point is the whole method's quality control, the same closure logic as a traverse in miniature.

   Try this: On any curve you set out, book the theoretical deflection angle to the last peg and compare it with the observed one before releasing the pegs to the gang.

10. 10Computation chapters on areas and volumes

    Areas and volumes: the numbers that get paid

    The survey ends at the quantity surveyor's desk: how many square metres of land, how many cubic metres of cut and fill. The manual gives the tools in order of refinement - the trapezoidal rule for a first estimate, Simpson's rule for curved boundaries, the planimeter for irregular plots. For earthworks, cross-sections are taken at regular intervals and the volumes between them computed by the prismoidal or end-area methods.

    The idea

    These are the numbers that turn into invoices and land titles, so the book is blunt about method choice: trapezoidal overestimates on curved boundaries, Simpson's rule is exact for cubics, and the prismoidal method beats averaging end areas when the ground changes shape between sections. The professional habit is computing a quantity two ways and reconciling the difference before anyone signs.

    Try this: Compute one area both by trapezoidal and by Simpson's rule and note the gap - that gap is the price of using the cheap method on curved ground.

11. 11Tacheometry chapters

    Tacheometry: distance without the tape

    Where a tape cannot reach - across a river, up a slope, through scrub - the tacheometer does the work optically. The staff is sighted through the instrument's stadia hairs; the intercept between the hairs, multiplied by the constant, gives the distance, and the vertical angle corrects it for slope and gives the reduced level in the same observation. One sighting, three numbers: distance, height, and position.

    The idea

    Tacheometry is the ancestor of the total station's reflectorless shot, and the logic is identical: trade a little accuracy for enormous gains in speed and reach. The book's caution is that the staff must be held plumb and read carefully, because the intercept is small and a misread multiplies straight into the distance. It is the right tool for contouring and topographic detail, the wrong one for boundary lines.

    Try this: Use stadia readings for your next detail sweep and compare one tacheometric distance against a taped one - calibrate your own reading error before trusting the method.

12. 12Computation chapters

    Computations: the office half of the job

    Back from the field, the surveyor faces the tables: coordinates to compute, bearings to carry, areas to close. The manual lays out the routine - reduce the field book, compute the traverse, adjust the misclosure, plot, and check the plot against the field book. Every step has a check, and every check is cheaper than a return visit to the field.

    The idea

    The office routine is where a survey becomes a document. The book's insistence on checks at every stage - arithmetic checks, plotting checks, closure checks - reflects a hard truth: errors found in the office cost minutes, errors found on site cost days. The discipline of reducing the field book the same day, while the ground is still in memory, is the cheapest insurance in the profession.

    Try this: Adopt the same-day rule: reduce and check your field book before you sleep on it, while a wrong booking can still be retraced.

Field surveyors, site engineers and construction students who set out buildings or run levelling and traverse work and want their numbers to close. Also useful to anyone auditing survey records who needs to know what a defensible field book looks like.

FAQ