Landfill and waste-facility fires rarely start as flames — they start as weeks of subsurface heating, invisible from ground level. Regular thermal drone overflights turn that slow buildup into an early-warning signal, at a fraction of the cost of a single fire response. The finding is never a temperature; it is a trend.
- WeeklyBaseline survey interval
- 60–100 mTypical survey altitude
- Pre-dawnLowest-noise window
- 20–30%Required frame overlap
Key takeaways
- Decomposition heat is normal. The anomaly is a localised cell trending hotter than its surroundings across successive surveys — which means a single flight proves almost nothing.
- Lithium batteries in modern waste streams create acute risk at tipping faces: small, intense hotspots rather than slow trends.
- At 60–100 m survey altitude, NETD matters more than long detection range. You are resolving small temperature differences, not distant targets.
- Thermal reads the surface. Deep subsurface heat only appears once it has conducted upward — which is exactly why weekly trending beats occasional checking.
On this page
What thermal patrol actually detects
Decomposition heat is normal; the anomaly is a localised cell trending hotter than its surroundings across successive surveys. Flying the same grid weekly and comparing radiometric data makes the trend unambiguous.
This is the central discipline of the mission. A landfill surface is warm everywhere — biological decomposition generates heat continuously, and any single survey shows a patchwork of temperatures that means nothing in isolation. What matters is whether cell 7 was 4 K above its neighbours last week and is 9 K above them now.
Lithium battery loads in modern waste streams add acute risk at tipping faces and recycling piles — small intense hotspots that isotherm alarms flag automatically. These behave completely differently from decomposition heating: they appear suddenly, are spatially small, and can escalate within hours rather than weeks.
| Signature | Spatial pattern | Time behaviour | Interpretation | Urgency |
|---|---|---|---|---|
| Broad warm area | Whole cell, diffuse | Stable week to week | Normal decomposition | None |
| Cell trending upward | Localised, growing | Rising across 2–4 surveys | Developing subsurface heating | High |
| Small intense hotspot | < 2 m, sharp edges | Appears within hours | Battery or hot load | Critical |
| Warm linear feature | Along a boundary | Stable | Gas well or leachate line | None |
| Hot spot at gas well head | Point source | Variable | Well fault or flare | Medium |
| Warm patch after tipping | At working face | Decays over days | Fresh warm load | Low, monitor |
| Cool patch | Localised | After rain | Water infiltration | Low |
| Surface crack venting | Linear, hot | Persistent | Subsurface fire reaching surface | Critical |
Battery hotspots break the weekly model. Decomposition heating develops over weeks and weekly patrol catches it comfortably. A lithium battery in a recycling pile can escalate within a shift, which is why tipping faces and fresh piles deserve a pass on every flight regardless of the trending schedule.
Survey pattern
A fixed grid at consistent altitude and time of day, with 20–30 % overlap. Pre-dawn is ideal because solar loading is the main noise source — a surface warmed unevenly by sun produces exactly the kind of patchy anomaly the survey is trying to distinguish from real heating.
Radiometric capture lets operators re-threshold in post rather than reflying. This matters more here than on most missions: the useful threshold changes with season and with what the previous surveys showed, and a non-radiometric dataset cannot be re-examined against a new threshold.
- Fix the grid. Same flight lines, same altitude, same gimbal angle, recorded so the next survey reproduces them.
- Fly pre-dawn, at the thermal minimum, before solar loading begins.
- Maintain 20–30 % overlap so the site mosaics cleanly and no ground is missed between lines.
- Capture radiometrically so the data can be re-thresholded and compared against earlier surveys.
- Add a dedicated pass over tipping faces and fresh piles on every flight, for acute battery risk.
- Compare against the previous survey before reporting. New or rising beats merely warm.
At 60–100 m survey altitude, detection sensitivity matters far more than long-range DRI performance. You are resolving small temperature differences across a nearby surface, not detecting distant targets — see thermal sensitivity explained.
Cost and the economics of early detection
A landfill fire is one of the most expensive incidents a waste facility can experience. Subsurface fires can burn for months, resist conventional suppression, generate regulatory exposure through air quality and leachate, and in severe cases force excavation of the affected cell.
Against that, weekly aerial patrol costs a fraction of a single response — and the comparison is not close. This is the rare monitoring application where the business case does not need careful construction; one prevented cell fire funds years of surveying.
The operational cost that matters is analysis discipline rather than flight time. Weekly surveys accumulate quickly, and the programme only works if someone compares each survey against the last. Automated isotherm thresholding removes most of that burden — see isotherms and temperature alarms — but the comparison step cannot be skipped.
Where landfill monitoring goes wrong
- Treating a single survey as diagnostic. Everything on a landfill is warm; only the trend is meaningful.
- Flying in daylight. Solar loading produces patchy warm areas indistinguishable from real anomalies.
- Non-radiometric capture. Removes the ability to re-threshold or compare across surveys, which is the whole method.
- Changing grid geometry. Different altitude or lines makes week-on-week comparison unreliable.
- Ignoring tipping faces between scheduled surveys. Battery risk is acute, not gradual.
- Expecting to see deep heat immediately. Thermal reads the surface; a deep hot core appears only once it has conducted upward, which is precisely why frequency matters.
Recommended UAVThermal payloads
OP-90P — radiometric 640×512 with high and low gain ranges (−20 to 150 °C and 0 to 550 °C) for both trend monitoring and active incident work.
MV-2P — 130 g thermal at NETD below 40 mK with thermometry for routine gridded patrol on light aircraft, where sensitivity at survey altitude is the binding specification.
OP-125A — adds 30x zoom for perimeter security and site surveillance on the same flights.
Related reading
Technology: radiometric temperature measurement and how sensitivity is specified.
Field practice: isotherms and temperature alarms, NETD explained and managing survey data over time.
Adjacent missions: wildfire early detection and mining and quarry monitoring.
FAQ
How often should a landfill fly thermal surveys?
Weekly is a common baseline, tightening to daily for cells showing an upward trend or after hot-load acceptance. Tipping faces and fresh recycling piles warrant a pass on every flight regardless of schedule, because battery-related hotspots escalate within hours rather than weeks.
Can thermal see heat under cover soil?
Only once the heat has conducted to the surface. Long-wave infrared reads the outermost surface, so a deep hot core becomes visible when it has warmed the ground above it — which is exactly why frequent surveys and trend comparison outperform occasional checking.
Why fly pre-dawn?
Solar loading is the dominant noise source. A surface warmed unevenly by sun produces patchy anomalies indistinguishable from genuine subsurface heating. At the pre-dawn thermal minimum, what remains warm is warm for a reason.
Do I need radiometric capture for landfill patrol?
Yes. The method depends on comparing this survey against previous ones and on re-thresholding in post as conditions change. Non-radiometric data shows relative contrast that shifts with automatic gain control, so it cannot support either operation.