Four corrections, one order
A distance read off a tape is not the horizontal distance between two points, and four separate things stand between them. Three of them change how long the tape itself is: temperature, whether the tape is actually its nominal length, and how much it sags between whatever is holding it up. The fourth converts the corrected slope distance into a horizontal one. They have to be applied in that order, because the slope reduction operates on the true length along the ground, not on the raw reading.
Two of them can go either way. A tape hotter than its standard temperature is longer than nominal, so its marks are further apart than they claim and the true distance is larger than the reading. A tape that has had an end clip replaced can be long or short by a sixteenth or more. Sag only ever goes one way — the tape hangs in a curve longer than the straight line between the end marks, so the reading is always too big — and so does the slope reduction, which always makes the horizontal distance smaller than the slope distance.
Sizes, so you know what to bother with
| Source | Condition | On a 100 ft pull |
|---|---|---|
| Slope | 5 percent grade | 1.50 in short |
| Slope | 10 percent grade | 5.96 in short |
| Sag | 0.02 lb/ft tape, 20 lb pull, unsupported | 0.50 in long |
| Sag | Same tape at 40 lb pull | 0.125 in long |
| Temperature | Steel tape 22 F above standard | 0.17 in short in the reading |
| Standardisation | Tape 1/16 in long | 0.06 in short in the reading |
Slope dominates and it is not close. At 5 percent a hundred foot pull is an inch and a half long; at 10 percent it is nearly six inches, four times as much for twice the grade, because the cosine falls away faster than the grade climbs. It also does not average out: a site that falls one way falls that way for every pull you take on it, so the error accumulates in one direction into every dimension derived from that line. Temperature at least changes sign between a cold start and a hot afternoon.
The cheap fixes beat the arithmetic
Correcting afterwards means knowing the slope, the temperature of the steel and the tension accurately, and those are all harder to establish than they sound. Holding the tape level and plumbing down removes the slope correction entirely and needs nothing measured. Laying the tape on the ground, or supporting it at the midpoint, removes most of the sag — sag grows with the cube of the unsupported span, so halving the span cuts it to an eighth per span and a quarter overall. Pulling harder helps too, since sag falls with the square of tension, but there is a limit to what a person holds steady.
Breaking a long pull into shorter level segments, each plumbed down, is the traditional answer to sloping ground and it is still the right one. It trades one large systematic error for several small independent ones, which partly cancel instead of adding.
What corrections cannot fix
All four of these are systematic errors with a size and a sign. Blunders are a different category: hooking the tape on the wrong side of a nail, reading from a tape whose zero is a foot in from the end, letting the tape wrap over a stake, or writing 47.3 when the tape said 43.7. None of them has a formula and none of them is small. The check is measuring the line again from the other end, which is also the only thing that catches a mistake in the correction arithmetic itself.
The distances go on to be used somewhere. If the point of the measurement is a rectangle, the squaring calculator pulls diagonals and tells you what a mismatch means, and the same slope problem shows up there — a diagonal measured along the ground on a sloping site reads long and makes a square layout look racked. For converting between the ways a grade gets expressed before it goes into any of this, the slope and grade calculator does percent, ratio, degrees and fall per foot.
Questions people ask
Which correction matters most?
Slope, by a wide margin, on anything that is not flat. Five percent of grade costs an inch and a half over a hundred feet; ten percent costs almost six inches. Sag on an unsupported light tape at moderate tension is about half an inch, temperature a couple of tenths, and a tape that is off nominal by a sixteenth is about six hundredths. If you only ever correct one thing, correct for slope — or better, hold the tape level and plumb down so there is nothing to correct.
Does a hot tape read long or short?
The reading comes out short of the true distance. Steel expands when it is hot, so a tape above its standard temperature is physically longer than nominal, and its marks are further apart than they claim. Lay it along a real 100 feet and the tape will read slightly less than 100. The correction is therefore positive: add to the reading. It is a small effect for most work — 22 degrees above standard on a steel tape is about 0.17 inches per hundred feet — but the sign is worth getting right because half of people guess it backwards.
How do I reduce sag without a formula?
Support the tape. Sag depends on the cube of the unsupported span, so a single support at the midpoint turns one full span into two half spans and cuts the total correction to a quarter. Laying the tape flat on the ground removes it entirely, which is why ground measurement is preferred whenever the surface allows. Tension helps as the square, so doubling the pull also quarters the sag, but a heavier pull is harder to hold consistently and stretches the tape slightly, which is a different correction.
What if I do not know the slope?
Then hold the tape level and plumb down, which removes the need to know it. Breaking a long pull into shorter level segments is the standard way to handle sloping ground with nothing but a tape and a plumb bob, and it turns one big systematic error into several small independent ones that partly cancel. If you must measure along the slope, the vertical difference between the two ends is usually easier to establish accurately than the angle, and either input works here.
Do these corrections apply to a fibreglass or cloth tape?
The structure of the calculation does, but the numbers do not. The expansion coefficient here defaults to a figure typical of a steel tape and should be replaced with whatever the manufacturer states for the tape in your hand — different materials are not close to each other. More importantly, a cloth or fibreglass tape stretches under tension by far more than steel, and by an amount that depends on how hard you pull, which is not modelled here at all. For work where a sixteenth matters, a steel tape at a known tension is the tool.