Capture guide · Thermal + RGB pairs

Flat roof moisture survey

Trapped moisture holds heat longer than the dry roof around it. Everything below exists to make that difference visible — which is why the timing matters more than the flying.

Before you fly

The roof has had a clear or mostly clear day beforehand, so it has taken on solar heat.
No appreciable precipitation for at least 24 hours before the survey. ASTM C1153
Roof surface free of ponded water, snow, ice and debris. ASTM C1153
Wind below 25 km/h (≈15 mph) while scanning. ASTM C1153
Temperature differential of 10°C or more between wet and dry areas.
Heating on inside the building where that is possible — it strengthens the differential.
Fly from one hour after sunset onward. Early morning, just before dawn, is best — the roof has had all night to give up its heat and the contrast is at its strongest.

The flight

Camera straight down (nadir, 90°) for the whole roof. Oblique frames only as supporting context — angled shots pick up sky reflection and distort the thermal signature.
Fly the roof in parallel passes at a constant height. Changing height mid-survey changes the apparent size of everything you find.
Fly at 20 m above roof level. On a 640×512 thermal this puts roughly 2.9 cm on each pixel, so the 12 cm anomaly the standard asks you to resolve spans about four pixels. If your thermal sensor is lower resolution than 640×512, fly lower still — and tell us the height you used either way.
Overlap of 30% forward and to the side, so there is no unsurveyed strip between passes.
Capture a matching RGB frame at every thermal position. Every thermal anomaly is cross-referenced against the visual image before it becomes a finding — without the pair it stays an anomaly and never makes the report.
Include at least one frame showing the whole roof for the site overview section of the report.
Shoot radiometric JPEG (R-JPEG) and do not change palette or level/span mid-flight.

What you must tell us

Send these with your imagery. We do not need you to fly to one fixed number — we need to know the numbers you actually flew to, so findings can be sized and assessed correctly. Anything you cannot measure, say so rather than estimating.

Drone and sensor used — make and model of both aircraft and thermal / multispectral payload
Flight height above the surface, and whether it was held constant
Date and time of flight, to the nearest 30 minutes
Ambient air temperature
Wind speed
Relative humidity
Reflected apparent temperature
Measured delta T, if you took one

What makes a set unusable

Thermal frames with no matching RGB. Findings cannot be cross-referenced, so they cannot be confirmed — and unconfirmed anomalies do not go in a report.
Flown in daylight, or too soon after sunset. Solar loading is still on the surface and masks the very thing being looked for.
Rain in the 24 hours before. Surface water masks the signal completely.
Non-radiometric JPEGs. No embedded temperature data means no spot temperatures and no differentials.
Palette or scale changed between frames. Nothing can be compared across the set.

See the flat roof moisture survey service → All capture guides

Issued 23 August 2026. Conditions, timing, wind limit and the 12 cm aerial resolution requirement follow ASTM C1153, the published standard for infrared roof moisture surveys. If your site operator imposes stricter limits than anything below, follow theirs and tell us what you used.