measurement
Footprint Is Not Area, and the Difference Is the Slope Factor
Surface area is footprint times a factor that depends only on pitch. The arithmetic is one square root, and getting it wrong costs a whole order.
Look straight down at a roof and you see its footprint: the area it covers on the ground. That is not the area you are buying material for, because the roof is tilted and the tilt makes the surface longer than its shadow.
The relationship between the two is exact, it depends on nothing but the pitch, and it is one line of arithmetic.
The slope factor
Roof pitch in North America is written as rise over run with the run fixed at 12. A 6/12 roof rises six inches for every twelve horizontal.
A plane rising r over a run of 12 has a surface length, per 12 units of run, of the hypotenuse of that triangle. So the multiplier between footprint area and surface area is
`` slope factor = sqrt(12^2 + r^2) / 12 = sqrt(1 + (r/12)^2) ``
Worked out for the common pitches:
| Pitch | Slope factor | Footprint 2,000 sq ft becomes |
|---|---|---|
| 3/12 | 1.031 | 2,062 |
| 4/12 | 1.054 | 2,108 |
| 6/12 | 1.118 | 2,236 |
| 8/12 | 1.202 | 2,404 |
| 10/12 | 1.302 | 2,603 |
| 12/12 | 1.414 | 2,828 |
The practical reading of that table: on a steep roof the difference between footprint and surface is over forty percent. Pricing a 12/12 roof off its footprint underprices the material by more than a quarter of the job.
Two things are worth noticing. The factor is never less than 1, so footprint is always an underestimate. And it grows slowly at low pitches and quickly at high ones, which is why the error is modest on a ranch and severe on a Victorian.
It applies per plane, not per roof
The formula takes one pitch. Roofs have several.
A house with a 6/12 main roof and a 3/12 porch has two slope factors, and applying either one to the whole building is wrong in a direction that depends on which. The only correct method is to compute per plane and sum: each facet's footprint times that facet's own factor.
This is the single most common error in hand measurement, and it is also the reason a measurement report that gives you one total and no facet breakdown is less useful than it looks. If the report cannot show you the planes, it cannot show you how it arrived at the number, and you cannot check it.
What pitch actually gets measured
Aerial measurement derives pitch from the geometry it can reconstruct, which means pitch comes with a confidence that varies by source.
Stereo imagery and lidar give genuine elevation and a pitch derived from measurement. Single-image methods infer pitch from shadow, from roof line geometry, or from assumption, and the assumption is usually the regional convention for the building type. That is often right and it is not a measurement.
Any report should say which it did. A pitch reported to one decimal place with no indication of method is reporting precision it may not have, and the difference between an assumed 6/12 and an actual 8/12 is thirteen percent of the material on that plane.
The honest fallback, when the method is uncertain, is to state the pitch as a value with a tolerance, or to flag the plane for verification. Field crews can check a pitch in ten seconds with a level and a tape; they cannot check a number that was presented as certain.
Where the footprint itself comes from
Everything above assumes the footprint is right, which depends on the imagery underneath it.
Public imagery programmes such as the USGS National Map publish their resolution and their capture dates, and commercial providers vary widely on both. Two properties decide whether a footprint is trustworthy.
Ground sample distance, the real-world size of one pixel. At 30cm per pixel, an edge can be located to roughly a pixel, which on a 40-foot roof line is about a one percent error in one dimension. At 1m per pixel it is three times that, and small features such as a dormer start to disappear into the noise rather than being measured.
Orthorectification. A raw aerial photograph has perspective: tall things lean away from the centre of the image, so a roof edge is not where it appears to be. Orthorectified imagery has had that corrected using a terrain model, which is what makes measuring off it valid at all. Measuring off a raw oblique and scaling it is a different operation with a much larger error, and it is worth knowing which one a tool is doing.
The number to put on the order
Surface area is not the order quantity. Two more steps come after it.
Convert to squares: one roofing square is 100 square feet, so surface area divided by 100.
Then add waste, which covers cuts at hips and valleys, starter and ridge courses, and the pieces that end up in the skip. Waste is a function of the roof's complexity rather than its size: a simple gable wastes little, a roof with many hips and valleys wastes considerably more. A report that gives you surface area without a complexity indication has handed you the easy half of the calculation.
The discipline that prevents the expensive mistake is to keep the three numbers distinct and labelled all the way to the order: footprint, surface area, ordered quantity. Conflating any two of them is how a crew ends up one square short on a Friday afternoon, and the slope factor is where it usually happens.