Volumetric Accuracy in Drone Surveying Explained

Home

What is volumetric accuracy in drone surveying? (And how to evaluate it)

Propeller Aero

31st July 2026

Read Time: 4 mins

A small perfect cone of dark aggregate lit from the side against a neutral background

Volumetric accuracy is how closely a measured volume matches the true volume on the ground, and a high-precision drone workflow delivers 97–99%, tighter than the two to five percent margin common across many photogrammetry platforms. The piece separates accuracy from resolution, explains what GSD actually does for your volumes, and flags the factor most accuracy claims skip: base surface definition, which is the quiet reason two tools can measure the same pile and disagree.

Volumetric accuracy is how closely a measured volume matches the true volume of material on the ground, usually expressed as a percentage. In drone surveying, a high-precision workflow delivers volumetric accuracy of 97–99%, meaning the reported volume is within one to three percent of reality. That is tighter than the two to five percent error margin common across many photogrammetry platforms.

The number sounds simple, but it hides a few things worth understanding, because volumetric accuracy is not the same as image resolution, and the single biggest factor is one most platforms barely mention. Here is how to read an accuracy claim and judge whether it holds up.

Accuracy is not resolution

These get blurred together constantly, so it is worth separating them.

Horizontal resolution is how much detail you can see across the map, driven largely by how sharp and dense your imagery is. It tells you how crisp the picture looks.

Vertical accuracy is how correct your elevations are, up and down. This is the one that drives volume, because a volume is really a stack of height differences between two surfaces.

Volumetric accuracy is the result of all of it working together: sharp horizontal detail, correct vertical elevations, and a well-defined surface to measure against. A platform can produce a beautiful, high-resolution map and still hand you a volume that is off, because a pretty picture does not guarantee correct elevations.

What GSD actually means for your volumes

Ground sample distance, or GSD, is how much real-world ground each pixel represents. A GSD of two centimeters means every pixel covers a two-centimeter square of earth. Lower GSD means more detail.

In practice, GSD is set by how high you fly and the camera you fly. Fly too high and pile edges smear, which matters most on steep or ragged stockpiles where the edge is where the volume lives. But GSD alone does not make a volume accurate. You can have excellent GSD and still be off if your survey is not anchored to real-world coordinates, which is why high-precision ground control from AeroPoints or a PPK drone mapping workflow does the heavy lifting on accuracy, not resolution by itself.

The hidden variable: base surface definition

Here is the factor that most accuracy claims skip. A volume is the space between two surfaces, the top of the pile and the ground beneath it. Your top surface can be flawless, but if the base surface is wrong, so is the volume.

Many platforms default to a flat mathematical plane pushed under the pile. On a real site, the ground is rarely flat. Ramps, dips, old material, and uneven pads all mean a flat base overstates or understates the true volume, sometimes by more than the accuracy percentage the platform advertises. This is the quiet reason two tools can measure the same pile and disagree. Propeller lets you define a real, measured base surface on the Propeller Platform and run repeatable volume calculations against it, so the number reflects the site as it actually sits.

How to evaluate an accuracy claim

When a platform advertises a volumetric accuracy figure, these are the questions that separate a defensible number from a marketing one.

– Is the figure volumetric accuracy, or just image resolution? They are not the same
– Does it depend on high-precision ground control, and is that ground control included in the workflow?
– How is the base surface defined? Can you set a real measured surface, or are you stuck with a flat plane?
– Is the output timestamped and repeatable, so two surveys weeks apart stay comparable?
– Can the result stand up to an external auditor?

That last question is the one that matters to a surveyor whose name is on the deliverable. An accuracy claim is only useful if you can defend it, and offloading processing to a proven survey processing pipeline is how survey and engineering teams keep the number defensible without drowning in point clouds.

Where Propeller lands

Propeller delivers 97–99% volumetric accuracy by treating every part of the chain as something that has to hold up: high-precision ground control anchors the survey, processing handles denoising and quality control, and a measured base surface makes the volume reflect real ground. The result is a number tighter than the common two to five percent industry margin and, more importantly, one you can put in front of an auditor.

For survey and engineering teams, that is the whole point. Accurate volumes are only valuable if they are defensible, and defensible volumes are what keep inventory honest across aggregates and every other site where the pile is the balance sheet. Track how each surface changes over time with progress tracking and the accuracy holds from one capture to the next.

Frequently asked questions

What is a good volumetric accuracy for drone surveys?
A high-precision workflow delivers 97–99% volumetric accuracy, which is within one to three percent of true volume. Many platforms cite a wider two to five percent error margin.

What is the difference between GSD and volumetric accuracy?
GSD is how much ground each pixel represents, a measure of detail. Volumetric accuracy is how close a measured volume is to the real one. Good GSD helps, but ground control and base surface definition drive accuracy.

Why do two platforms measure the same pile differently?
Usually the base surface. A flat mathematical plane and a real measured surface can produce meaningfully different volumes for the same pile.