geoprimsField-grade geospatial math

What is ground sample distance (GSD)?

How much ground one pixel covers, how to work it out, the height for a target GSD, and why GSD is not accuracy.

For drone operators · updated 2026-09-23

Planning and education aid. Not for primary navigation. Full disclaimer

Ground sample distance (GSD), also called ground sampling distance, is the width of ground that one pixel of an aerial image covers. A GSD of 2.7 cm means each pixel spans 2.7 cm on the ground, so the smallest thing you can pick out is a few pixels, or several centimeters, across. GSD depends on the camera and on how high above the ground you fly: fly twice as high and each pixel covers twice as much ground.

GSD is the first number in almost every mapping flight plan. It sets the flying height, which sets the photo footprint, which sets the overlap spacing and the number of photos.

Why it matters

A client asks for “2 cm imagery,” or a job needs enough detail to see cracks, paint lines, or survey targets. GSD is how you turn that request into a height. It also drives the cost of the job: halving the GSD halves the footprint in each direction, so it takes about four times as many photos to cover the same area.

In the United States, Part 107 limits a small drone to 400 ft above ground level, unless it stays within a 400 ft radius of a structure and no higher than 400 ft above the structure’s top (14 CFR 107.51(b)). That caps the coarsest GSD a given camera can reach under the rule, and it is the first thing to check when the height for a target GSD comes out high.

How it is worked out

A vertical photo is a pair of similar triangles that meet at the lens. The sensor lies one focal length behind the lens, and the ground lies one flying height below it. Penn State’s GEOG 892 course writes the scale as focal length over flying height. For a digital camera the useful form is per pixel:

GSD = sensor width × height above ground / (focal length × image width in pixels)

Sensor width divided by image width is the pixel pitch, the physical size of one pixel on the chip. Turn the formula around and you get the height for a target GSD:

Height = target GSD × focal length × image width / sensor width

Three details matter:

A worked example

A drone with a 1-inch, 20 megapixel camera: a 13.2 mm × 8.8 mm sensor, 5,472 × 3,648 pixels, and an 8.8 mm lens, flying 100 m above the ground:

Step Result
GSD 2.741 cm
Footprint across track 150 m
Footprint along track 100 m

Change the height and the GSD scales with it:

Height above ground GSD Footprint across track
72.96 m 2 cm (target) 109.4 m
100 m 2.741 cm 150 m
120 m 3.289 cm 180 m
400 ft (121.9 m) 3.342 cm 182.9 m

For a 2 cm GSD this camera flies at 72.96 m. For 1 inch per pixel, the height for a GSD tool gives 92.66 m.

What GSD says about accuracy, and what it does not

GSD is resolution, not accuracy. A sharp image can still sit in the wrong place on the earth if the camera calibration, the GNSS, or the ground control is poor, and a map made from coarser pixels can be more accurate than one made from finer pixels.

Older map standards tied accuracy to scale or pixel size. ASPRS Edition 2 of its Positional Accuracy Standards for Digital Geospatial Data (Version 2, 2024) does not. Penn State’s summary of it says that accuracy measures based on map scale, film scale, and GSD no longer apply, and that the new accuracy classes are independent of GSD. Accuracy is stated as root mean square error (RMSE) measured at independent checkpoints, with a minimum of 30 checkpoints, and the accuracy of the checkpoint survey itself is counted in. With a 1.00 cm vertical fit and checkpoints surveyed to 2.0 cm, the ASPRS accuracy tool reports 2.24 cm RMSE.

So choose GSD for what you need to see, and prove accuracy with checkpoints.

Common mistakes

Where the numbers come from

The image geometry and the height formula follow Penn State’s GEOG 892 course, lesson 4, which draws on Wolf, Dewitt, and Wilkinson’s Elements of Photogrammetry. The accuracy points come from Penn State’s lesson 8 summary of the ASPRS Edition 2 standards, and the height limit from 14 CFR 107.51. The GSD tool shows each step under “How we got this.” Once you have a height, the overlap and trigger tool turns it into photo spacing and line spacing, and the Part 107 altitude tool checks it against the rule.

Try it: Ground sampling distance (GSD)

The calculator below is the real tool, running its worked example. Change any value; nothing leaves your device.

2.741cm

Each pixel covers 2.741 cm of ground, and one image spans 150 m across track.

Footprint across track
GSD along track
Footprint along track
Focal length used
Provenance
Computed by
drone.photogrammetry.gsd 1.0.0, core 0.1.0
Model
GSD = sensor width × height / (focal length × image width)
Accuracy
Exact for a nadir view over flat ground at the given height; real GSD varies with terrain and lens distortion
Notes
None
Cites
Wolf, P. R., Dewitt, B. A., and Wilkinson, B. E., McGraw-Hill, Elements of Photogrammetry with Applications in GIS

Open the full tool, with your values →

Your values

Showing an example. Change anything.
More options 4

Tools for this

Sources

Every number on this page comes from the tools above, which cite their sources on their own pages under "How we got this." Found a mistake? Use "Report a problem" at the bottom of the page.