Infrared reflects. A polished surface shows you the temperature of whatever is reflected in it — the sun, the sky, a nearby hot object, sometimes the drone itself. Most false findings in aerial thermography are reflections, and almost all of them are eliminated by one twenty-second check.
Key takeaways
- Infrared reflects. A polished surface shows the temperature of whatever is reflected in it, not its own temperature.
- Most false findings in aerial thermography are reflections, and emissivity and reflectivity are two views of the same problem.
- A twenty-second check — move the aircraft and see whether the hot spot moves with you — separates a reflection from a real anomaly.
- Design the survey around the problem: pick angles and time windows that keep sky and sun out of the reflection path.
On this page
Emissivity and reflectivity are the same problem
A surface emits some of its own infrared and reflects the rest of what falls on it. High-emissivity surfaces — painted metal, concrete, rubber, oxidised steel — emit strongly and reflect little, so what you see is mostly their real temperature. Low-emissivity surfaces — polished aluminium, stainless steel, galvanised sheet, glass — emit weakly and reflect strongly, so what you see is mostly something else.
This has two consequences that pull in opposite directions. Low-emissivity surfaces under-read their own temperature, so a genuinely hot polished busbar can look cool. And they show reflected sources, so a cool polished surface can look hot. Both errors occur on the same materials and often in the same image.
Material risk table
| Material | Approx. emissivity | Reflection risk | How to handle it |
|---|---|---|---|
| Painted metal | 0.90–0.96 | Low | Read directly |
| Concrete, brick | 0.90–0.95 | Low | Read directly |
| Oxidised / weathered steel | 0.75–0.90 | Low–medium | Read with care |
| Roofing felt, EPDM | 0.90–0.95 | Low | Read directly |
| Anodised aluminium | 0.55–0.80 | Medium | Comparative reading preferred |
| Galvanised steel | 0.20–0.35 | High | Comparative only, or apply target patch |
| Polished stainless | 0.10–0.20 | Very high | Do not read absolutely |
| Bare / polished aluminium | 0.03–0.10 | Very high | Do not read absolutely |
| Glass | 0.85–0.95 in LWIR | High for sky reflection | Beware sky and sun in glazing |
| Water surface | 0.95–0.98 | High for sky reflection | Angle matters more than material |
The twenty-second check
Orbit the target. A genuine thermal source is a property of the object and stays in the same place on the object as your viewing angle changes. A reflection is a property of the geometry between the source, the surface and you — so it slides across the surface, changes shape, or disappears entirely.
That single manoeuvre eliminates the large majority of false hot spots, and it is fast enough that there is no excuse for skipping it. Where orbiting is not possible, a second observation from a different altitude achieves much the same thing.
Two supporting checks are worth knowing. A reflection often has softer edges and a less physically plausible shape than a real thermal source. And a real heat source usually shows conduction into surrounding material — a hot joint warms the conductor either side of it — while a reflection sits on the surface with no thermal gradient around it.
Other things that are not what they look like
- Solar loading. A south-facing surface at 15:00 is warm because the sun heated it, not because anything is wrong. Survey before sunrise or well after sunset when absolute temperature matters.
- Thermal shadows. A cool patch where something recently sat — a moved pallet, a departed vehicle — reads as an anomaly and is just history.
- Wet patches. Evaporative cooling makes damp areas read cold. Useful for finding moisture; misleading when you are looking for something else.
- Vegetation and debris. Insulating cover over an asset changes its apparent temperature without changing its actual one.
- The aircraft itself. On highly polished surfaces at close range, a warm drone can appear in its own image.
Related reading
Operations: emissivity in drone thermography, seven thermal imaging mistakes and what a report should contain.
Technology: radiometric temperature measurement.
Missions: substation inspection and roof and industrial inspection.
Designing the survey to avoid reflections in the first place
Detecting reflections is a skill; not creating them is a plan. Three planning decisions remove most of the problem before the aircraft leaves the ground.
Survey time. Pre-dawn and well after sunset eliminate solar reflection entirely and remove the solar loading that makes every surface warm for the wrong reason. For anything where absolute temperature matters, this single choice does more than any technique.
Approach geometry. Plan the flight so critical surfaces are viewed near-normal and with the sky behind you rather than reflected in front of you. On a substation or a roof, that is usually a matter of choosing which side of the asset the flight line sits on — a free decision at planning time and an expensive one to discover in post-processing.
Reference targets. Where a low-emissivity surface must be measured repeatedly, a small high-emissivity patch — tape, paint, an emissive sticker — applied once during a maintenance outage turns an unmeasurable surface into a reliable one for every future survey. Asset owners are often willing to do this when the alternative is explained.
- Survey before sunrise or after sunset when absolute temperature matters.
- Plan flight lines so critical surfaces are viewed near-normal, sky behind the aircraft.
- Apply emissive reference patches to low-emissivity assets during a planned outage.
- Capture every finding from two angles as standard practice, not only when something looks suspicious.
- Photograph the surroundings — the reflected source often identifies itself in the visible image.
FAQ
How do I tell a reflection from a real hot spot?
Orbit the target. A real thermal source stays in the same place on the object as your viewing angle changes; a reflection slides across the surface or disappears. Real sources also usually show conduction into surrounding material, while reflections sit on the surface with no gradient around them.
Why does a shiny hot component read cool?
Low emissivity. A polished surface emits little of its own infrared and reflects mostly ambient, so it under-reads its true temperature. Use comparative readings against a high-emissivity reference on the same assembly, or apply an emissive target patch where the asset allows.
Can I fix reflections by adjusting emissivity settings?
No. Emissivity correction addresses under-reading of a surface’s own emission; it does not remove a reflected source. Reflections are a geometry problem and are solved by changing viewing angle or survey time.