A payload that reports a target coordinate is chaining together five independent measurements, each with its own error. The result is not a GPS fix on the target; it is a computed estimate whose accuracy degrades sharply with slant range and with how obliquely you are looking. Knowing the terms tells you when to trust the number.
Key takeaways
- Geolocation error is a chain: attitude, position, ranging and terrain assumption each contribute, and the chain is dominated by its worst link.
- Slant angle amplifies attitude error. The same payload is far more accurate looking down than looking out.
- Real-world error is worse than the budget table because the table assumes a known ground elevation that surveys rarely have.
- Report accuracy as a figure with its conditions attached, so the people acting on the coordinate know what tolerance they are working inside.
On this page
The chain
Target position = aircraft position + (rotation from aircraft attitude) + (rotation from gimbal angles) + (range along the line of sight). Each term contributes error, and the rotational terms contribute error that scales with range.
That last point is the whole story. A 0.5° attitude error contributes about 8.7 m of lateral error at 1,000 m and 87 m at 10,000 m. Position error contributes a constant offset; angular error contributes a growing one. Beyond a few hundred metres, angular terms dominate everything else.
Worked error budget
| Term | Typical error | Contribution at 300 m | At 1,000 m | At 3,000 m |
|---|---|---|---|---|
| Aircraft position (GNSS) | 1.5 m | 1.5 m | 1.5 m | 1.5 m |
| Aircraft position (RTK) | 0.03 m | 0.03 m | 0.03 m | 0.03 m |
| Carrier heading (magnetometer) | 1.0° | 5.2 m | 17.5 m | 52 m |
| Carrier heading (dual-antenna) | 0.2° | 1.0 m | 3.5 m | 10.5 m |
| Carrier pitch/roll (AHRS) | 0.3° | 1.6 m | 5.2 m | 15.7 m |
| Gimbal angle encoder | 0.05° | 0.3 m | 0.9 m | 2.6 m |
| Boresight misalignment | 0.1° | 0.5 m | 1.7 m | 5.2 m |
| Laser range | ±1.0 m | 1.0 m | 1.0 m | 1.0 m |
| RSS total, magnetometer heading | — | ≈ 5.8 m | ≈ 18.6 m | ≈ 55 m |
| RSS total, RTK + dual antenna | — | ≈ 2.2 m | ≈ 6.5 m | ≈ 19 m |
What makes it worse than the table
Oblique viewing. At a shallow grazing angle, a small vertical angle error translates into a large horizontal distance error. Looking down at 60° is far more accurate than looking out at 10°, and payload manuals that quote a single accuracy figure are implicitly assuming a favourable geometry.
No laser return. Without a measured range, the payload must intersect the line of sight with a terrain model or a flat-earth assumption at the aircraft’s take-off elevation. On sloping ground that assumption alone can produce errors of tens or hundreds of metres.
Magnetic disturbance. Heading is the largest single term in most installations, and it is the one most easily corrupted by nearby ferrous structures, high-voltage yards and the aircraft’s own motor currents.
Boresight drift. The alignment between laser, thermal and EO channels is a calibration that can shift with thermal cycling and hard landings. See EO/IR fusion and boresight alignment.
Using the number responsibly
- Treat a reported coordinate as a search centre with a radius, not as a point. Publish the radius alongside it.
- Prefer near-nadir geometry when accuracy matters more than standoff.
- Always use the laser range when available; a measured range removes the single largest non-angular assumption.
- On accuracy-critical work, use RTK positioning and dual-antenna heading. Together they cut the total error roughly threefold at long range.
- Re-verify boresight after any hard landing, payload swap or large temperature excursion.
How this shows up in our payloads
Pods with integrated ranging such as OP-90D (1,200 m) and LX-9B compute the coordinate from a measured slant range rather than a terrain assumption, which is why their published positioning accuracy is stated as a set of range-dependent values rather than a single number. That presentation is honest about the physics rather than convenient for marketing.
Related reading
Technology: laser rangefinding and target geolocation and IMU, AHRS and attitude fusion.
Field practice: why a rangefinder belongs on your payload, pointing accuracy at range and SAR search patterns.
Reporting accuracy honestly to the people who use it
The failure mode of geolocation is not technical, it is communicative. A payload reports a coordinate to six decimal places and a responder treats it as a survey point. When they arrive and the target is 40 m away, confidence in the whole system drops — not because the payload was wrong, but because nobody attached an uncertainty to the number.
The fix is to report a coordinate and a radius together, always. On a search operation that radius defines where the ground team starts a spiral; on an inspection it defines whether the finding can be tied to a specific asset or only to a span. Both are useful. A bare coordinate with implied precision is not.
Build the radius from the geometry of the shot rather than from a fixed figure. Slant range, viewing angle and whether the laser returned are all known at the moment of capture, which means the uncertainty can be computed automatically and travel with the coordinate through the workflow.
- Report every target coordinate with an accuracy radius, computed from slant range and geometry.
- Flag whether the range was laser-measured or terrain-assumed — the difference is often an order of magnitude.
- Record aircraft heading source (magnetometer or dual-antenna) in the mission log; it is the dominant term.
- For findings that will be revisited, prefer a near-nadir shot even if it costs standoff.
- Re-verify boresight after hard landings, payload swaps and large temperature excursions.
FAQ
Why does the manual say target positioning accuracy is for reference only?
Because the accuracy depends on variables outside the payload: aircraft position quality, heading source, viewing geometry, terrain and slant range. A single number cannot describe all of those, so manufacturers publish representative values and a disclaimer rather than a guarantee.
What is the single biggest source of coordinate error?
Carrier heading, in most installations. A 1° magnetometer heading error contributes about 17.5 m at 1,000 m — larger than every other term combined. Dual-antenna GNSS heading is the most effective single upgrade for geolocation accuracy.
Does the laser rangefinder fix everything?
It removes the terrain-assumption error, which is often the largest term when it applies, and it is essential over sloping or unknown ground. It does not reduce angular errors, which continue to grow with range regardless of how well you measure distance.