C-DRONE GUIDE · 2 SEPTEMBER 2026

The Measurement Window for Drone Thermography: Which Season, Which Hour and Which Weather Make the Flight Valid

A client calls in May to have their head office surveyed thermally. The weather is fine, the drone can fly, the contractor is available : everything is in place to produce a rigorously useless report. Thermography is not a photo shoot, it is a measurement ; and a measurement only exists under the conditions that make it possible. On a building envelope those conditions amount to a few dozen usable hours per month between November and March, usually before dawn. On a flat roof they invert : a summer evening after a sunny day. On a solar farm they invert again. This guide sets out the measurement window for what it really is to a client : a scheduling constraint, one that can be costed, enforced and written into the contract before signing.

Published on 2 September 2026, reviewed on 11 September 2026 — regulations in force as of September 2026.

Two independent gates: the drone can fly, and the measurement can still be void

The costliest confusion in this market fits in one sentence : « the weather is nice, let’s go ». Two different questions arise before a thermography flight, and they have neither the same thresholds nor the same answers.

• Flyability : can the drone take off, hold position and land safely ? This depends on mean wind and gusts, rain, icing, visibility and battery operating temperature. That is the subject of our guide on flight limits for a professional drone and rescheduling a mission.
• Measurement validity : does the phenomenon you are trying to observe exist, at the time of the flight, with an amplitude above the noise ? This depends on the temperature difference across the wall, on the thermal history of the preceding hours, on surface condition and on the sky.

The two gates are independent, and the second is almost always the more restrictive. A fine March day without a breath of wind is an ideal day to fly and a void day to survey a façade : the sun has loaded the walls, the indoor/outdoor difference collapsed at midday, and the image will show nothing but the sun’s path across the building. Conversely, an overcast, damp, cool January night — thoroughly unpleasant and poor for photography — is the best window of the year for an envelope audit.

The immediate practical consequence : a contractor offering thermography « whenever suits you » is not selling a measurement. A serious contractor will refuse certain days — and that refusal should be read as the main sign of competence, not as a lack of flexibility. The rest of this guide sets out the criteria behind that refusal, so that you can write them into your specification instead of discovering them the hard way.

The standards framework: careful, EN 13187 has been withdrawn since August 2023

Three documents govern the measurement window depending on the family of mission, and the first has changed recently without the French market catching up.

On building envelopes. EN 13187 of July 1999, « Thermal performance of buildings — Qualitative detection of thermal irregularities in building envelopes — Infrared method », quoted for twenty-five years in virtually every French thermal specification, has been withdrawn. Its AFNOR catalogue entry carries the status « standard cancelled ». It was superseded in August 2023 by EN ISO 6781-1, « Performance of buildings — Detection of heat, air and moisture irregularities in buildings by infrared methods — Part 1 : General procedures », the European transposition of ISO 6781-1:2023. The new text widens the scope to air and moisture irregularities and sets requirements on equipment, on the qualification of those involved and on report content. Part 3 of the same series, ISO 6781-3:2015, covers the qualification of equipment operators, data analysts and report writers in three classes.

Two lessons for a client in September 2026 : first, a specification or a report still invoking EN 13187 signals documentation that has not been kept up to date ; second, and this is the substantive point for the measurement window, this family of standards describes the detection of irregularities — not performance measurement. It asks you to create and document conditions in which a defect becomes visible, not to reach metrological accuracy. The figures in circulation (air temperature difference of at least 10 K, fifteen degrees to be comfortable, wind below 5 m/s, no direct sunlight during the examination or in the preceding hours, overcast sky preferable) are not thresholds of legal conformity : they are the converging conditions inherited from the old standard, from literature reviews and from practice. Treat them as such — good-practice rules to write into the contract, not clauses to invoke.

On roof moisture surveys. The reference document is American but authoritative everywhere : ASTM C1153, « Standard Practice for Location of Wet Insulation in Roofing Systems Using Infrared Imaging », in its 2023 edition. It describes night-time imaging, possible from one hour after sunset, the best images being obtained in practice three to four hours after ; it calls for wind below 25 km/h, no appreciable precipitation in the preceding 24 hours, and, under heavy overcast, a difference of at least 10 °C between the roof surface and the underside of the deck. Detailed application in our guide to flat-roof leak detection.

On photovoltaics. Technical specification IEC TS 62446-3:2017 calls for an in-plane irradiance of at least 600 W/m², low wind and limited cloud cover : the window is exactly the opposite of the building one. We do not detail it here, since our guide to solar farm thermography under IEC 62446-3 is devoted to it.

Finally, the competence of whoever interprets the images falls under separate schemes : in France, COFREND certifies the infrared thermography method (known as TT) at levels 1, 2 and 3 ; ISO 18436-7 organises machine condition-monitoring thermography into three categories. That side is covered by our guide on what a drone thermography report is worth.

The six physical variables that decide, and the decision grid

A thermal camera does not see insulation : it sees a surface temperature, the result of a balance between what the wall receives, what it stores and what it radiates. Six variables govern that balance, and each on its own can void the measurement.

• The temperature difference across the wall. This is the driver : with no flux through the fabric, an insulation defect leaves no signature. It is also what bounds the useful season in mainland France, from late October to early April depending on climate zone — longer in the east and the Massif Central, shorter around the Mediterranean.
• Thermal mass and delayed solar radiation. A sunlit façade keeps releasing stored energy long after sunset, and the heavier the material, the longer it takes. A 20 cm cast concrete wall does not become neutral again in an hour. That is why the window sits at the end of the night rather than early evening : the wall needs time to forget the day.
• Wind. Forced convection evens out surface temperatures and flattens the very differences you are looking for. Above a few metres per second, a genuine thermal bridge vanishes from the image with nothing to signal its disappearance — the most insidious false negative of the method.
• Surface moisture and recent rain. A wet wall cools as it dries, and does so unevenly. The areas drying fastest appear cold ; you then read an evaporation map while believing you are reading an insulation map. Dew and frost produce the same artefact.
• Sky condition. Under a clear sky a surface radiates towards the celestial vault, whose apparent temperature drops very low, and cools all the more where it is exposed and tilted upwards. On a vertical façade this creates gradients that have nothing to do with insulation : an overcast sky is therefore preferable for an envelope audit. On a flat roof the same effect is sought after, since it is what widens the contrast between dry and wet insulation.
• The heating regime of the preceding hours. The most frequently neglected variable, and the easiest to fix. A building must be heated steadily for at least 24 hours for its walls to be close to steady state. Heating restarted the same morning gives nothing ; a heavy wall has not finished responding.

These mechanisms are documented. The reference review by Angeliki Kylili, Paris A. Fokaides, Petros Christou and Soteris A. Kalogirou, published in 2014 in Applied Energy (vol. 134, pp. 531-549), surveys infrared thermography applications for building diagnostics and devotes a substantial part of its argument to the limitations and error sources of the method, ambient conditions foremost among them (see the study on Google Scholar).

One point deserves emphasis, because it changes how a mission should be ordered : the window is not an instant, it is a duration. Work by Matthew Fox, David Coley, Steve Goodhew and Pieter de Wilde published in 2015 in Energy and Buildings (vol. 92, pp. 95-106) tests time-lapse thermography — a series of images of the same building through the night rather than a single shot — and shows that dynamic thermal behaviour is more revealing than a snapshot ; they note that under typical conditions the differences between thermal areas from one image to the next are of the order of 0.2 K for a thirty-minute interval (see the study on Google Scholar). Operational translation : two passes spaced through the night beat a single one, and a report comparing two moments of the same night is markedly more robust than one built on a single image set.

Hence the following decision grid, to be copied into an envelope-audit specification.

Three families of missions, three opposite windows

The most common planning error is to transpose from one mission to another a rule learned for a single case. « Thermography is a winter job » is true for a façade and false for three missions out of four. The table below summarises the genuinely useful windows.

Two scheduling consequences are worth remembering. First, the envelope and flat-roof windows are incompatible : on the same building, a heat-loss audit and a water-ingress survey have to be ordered as two campaigns six months apart. A quote promising both in the same flight is a red flag. Second, the envelope window is far narrower than it looks : over a winter month, once you remove the windy nights, the rainy nights, the too-mild nights and the nights of wet frost, there are often only six to twelve genuinely usable nights. That figure is why a campaign is booked several weeks ahead, and why a firm slot has a price.

Emissivity: the surfaces no weather window can rescue

There is a category of failure the calendar cannot solve. A thermal camera converts received radiation into a temperature by assuming an emissivity — a surface’s ability to emit its own radiation rather than reflect that of its surroundings. On a high-emissivity surface, what you see comes from the surface. On a low-emissivity surface, what you see comes from somewhere else : the sky, a neighbouring building, the operator, the drone itself. No measurement window corrects this, because the problem is not in the conditions : it is in the material.

Three families of buildings are affected in practice : industrial and logistics buildings with profiled steel roofing, zinc roofs and claddings in urban fabric, and the glazed façades of commercial property. On those surfaces a displayed absolute temperature means nothing, and a night thermogram under a clear sky will mostly show the reflection of the cold sky — a uniformly « cold » roof that teaches you nothing.

The answer is not to move the date, it is to change the method : target a reference patch of known emissivity placed on the surface, reduce the angle of incidence by getting closer to perpendicular, fall back on non-metallic parts (parapets, upstands, verges, joints), or switch to a ground check with a contact probe. This is a conversation to have before the quote, because it changes the deliverable scope. Sensor fundamentals and measurement parameters are covered in our guide to the professional drone thermal camera.

Finally, the distinction that governs everything : qualitative thermography compares neighbouring areas within the same image and concludes « this area is clearly warmer than its immediate surroundings » ; quantitative thermography states an absolute temperature and commits to an accuracy. A drone flight honours the first properly and the second only with great difficulty, for reasons of distance, angle and sensor stability. The measurement window exists precisely to make the qualitative reading robust : that is its purpose, and it is not a limitation but a framing — provided the report says so, as explained in our guide on what a thermography report is worth.

Planning the campaign: the go/no-go protocol at D-30, D-2 and H-4

A thermography campaign is not ordered like a photo shoot : it is prepared like a test. Here is the sequence a client can impose, and which any competent contractor will accept without argument.

At D-30 — administrative framing. This is the sizing deadline, and it is regulatory. The measurement window for an envelope audit falls in the middle of aeronautical night, and in France, under the order of 3 December 2020 on the use of airspace by unmanned aircraft — in its version in force in September 2026, last amended by the order of 23 December 2025 — the aircraft shall not operate at night. The text provides for exceptions, in particular operating below 50 m height, with an aircraft under 8 kg, under a standard or national scenario, fitted with a light signalling device and over a secured operating area ; outside those cases, a derogation from the territorially competent prefect is required, after consulting the civil aviation and defence services. The ministerial page on open-category operations, updated on 13 July 2026, states that the request must be filed at least 30 days before the flights. A fenced, lit and guarded industrial site often ticks the exception conditions ; a city-centre head office much more rarely. Our guide to night flying with a professional drone and the prefectoral derogation details the procedure. Remember the scheduling rule : you do not decide on a night thermography survey a fortnight in advance.

At D-30 as well — technical framing. The scope has to be settled in the light of the surfaces that are actually measurable (see the section on emissivity), the number of passes per building has to be decided, and above all the most common trap in commercial property has to be dealt with : the night setback. An office building whose controls drop the setpoint from 19 °C to 16 °C between 8 pm and 6 am destroys the window at the very moment it opens. You therefore need written confirmation from the operator that the heating regime will be maintained for the 24 hours around the flight, and that instruction should be recorded. The same reasoning applies to schools at weekends, sports halls, churches and intermittently occupied buildings.

At D-2 — weather shortlisting. The contractor names one or two candidate nights on the basis of forecasts : minimum temperature, expected mean wind, cloud cover, precipitation over the preceding 24 hours. The client warns occupants, security staff and the heating operator. This is also the time to confirm site access.

At H-4 — the go/no-go decision. It is taken on observed data, not forecasts : actual outdoor temperature, wind measured on site, surface condition (dry, damp, frosted), rain in the preceding hours, cloud cover. The contractor records and time-stamps those values whether or not the flight goes ahead. A documented no-go is better than a failed flight : one costs a call-out, the other costs a report to redo and a works decision taken on false images.

During the flight — traceability. Conditions recorded at the start and at the end, indoor temperatures taken in two or three reference rooms, a visible-light photograph of each façade to remove interpretation ambiguities, and a record of camera settings. The review by Tarek Rakha and Alice Gorodetsky published in 2018 in Automation in Construction (vol. 93, pp. 252-264) on drone applications in the built environment insists precisely on formalising flight-planning procedures as a condition of reproducible inspection (see the study on Google Scholar). Across a property portfolio tracked from one year to the next, that formalisation is what makes two campaigns comparable : same trajectory, same angles, same time relative to sunrise, same settings.

In the contract — three clauses. A window clause listing the required conditions and making the flight conditional on them being observed ; a standby clause reserving a two- to three-week interval rather than a date, with 48-hour confirmation notice ; and a postponement clause setting in advance the cost of a call-out without a flight, rather than negotiating it at 4 am in a car park. Those three clauses cost a few hundred euros and avoid most of the disputes in this line of work.

2026 campaign prices, and the cost of a missed window

The figures below are ranges observed on the French market in 2026, excluding VAT, for a complete service covering preparation, the flight within the window, analysis and a report naming the measurement conditions. They are not firm prices : the actual figure depends on façade length, height, aerial accessibility of the site and the depth of analysis requested.

That leaves the invisible cost, the cost of a missed window. It is not measured by the price of the flight but by the price of the decision that follows. A campaign flown outside the window typically produces three effects : a « nothing to report » verdict on a genuinely leaky building, pushing insulation works back by one or two heating seasons ; or the opposite, false positives caused by sun and moisture that trigger a needless intrusive investigation ; or, at best, a report the project manager sets aside and which has to be redone the following winter. Set against those three outcomes, the extra cost of a weather standby is negligible — that is the argument to make internally when a buyer compares two quotes, one of which provides for no window at all. Our orders of magnitude by building type are detailed in the guide to drone thermography pricing, and the service as a whole on our drone thermography page.

Scheduling your campaign at the right time

If your need concerns heat loss, thermal bridges or an insulation defect, the question is not « when are you available ? » but « what is the first usable window, and what has to be prepared before then ? ». In practice, a winter campaign is framed in September or October : that is the lead time needed to settle the scope, secure a maintained heating regime and, where applicable, file a night-flight derogation request thirty days ahead.

Tell us about the building or the portfolio, the question you want answered and the deadline for your works decision : we will indicate the window we are aiming for, the conditions to be met and the corresponding price range. Request a drone thermography quote.

Request a free quote

Put it into practice

• Drone aerial thermal imaging: rates and cities covered from €500
• Aerial thermal imaging in Annemasse Auvergne-Rhône-Alpes
• Aerial thermal imaging in Aurillac Auvergne-Rhône-Alpes
• Aerial thermal imaging in Nice Provence-Alpes-Côte d'Azur

Also worth reading

• Olive orchards and drones: water stress and yield forecasting, price
• Prune d'Agen and drones: water stress and ripeness, price
• Asparagus and drones: spotting weeds on the ridge before emergence, price