Edition 2026.2 — economics of forest fires in France, 2022 foundation and season as of 15 September 2026

The price of fire

Preventing, fighting, rebuilding: what fires really cost France

State of the data and of the study as of 15 September 2026. The values for the current season are situation points, subject to revision, and each carries its date.

The answer

France knows what it costs to put out a fire. It does not know what a fire costs.

The mechanism, its consequence and what the study does not settle are demonstrated in the full analysis.

The criteria that would bring this thesis down were written before looking at the results; they are set out with the method.

Where do things stand today?

Public assessment of the 2022 fires: €178M
Area covered in 2025: 21,945 ha
2026 season: area covered since 1 January: about 119,000 ha

None of these values adds to another: they carry neither the same scope nor the same date, and the point for the current season is not a final tally.

Back to the answer

Reserved chapter · the economic foundation

€178M: what the official figure counts — and what it leaves out

Three items, and only three: carbon emitted in excess, commercial timber lost, and the recreational use of the forest. No destruction, no employment, no overnight stays, no insurance compensation. And the heaviest item is the average of a published range.

Reserved · full analysis

The decomposition of the only public assessment of the fires, reproduced from its parameters instead of copied: its three items, the range within which its total moves without any assumption changing, the 6 named empty items that its scope leaves out, its place in the study it comes from, the firefighting extra cost served separately, and the carbon item under the convention in force

Back to the answer

Reserved chapter · capacity

12 amphibious water bombers. How many for the next outbreak?

Between what the State owns and what can take off for a fire that has not started, four filters apply. The first three are published. The last two are not.

Reserved · full analysis

The funnel of aerial capacity — 5 stages, including 2 that carry no figure because no source publishes them — with the dated points served as observations and never as a series, the fleet type by type, seasonal rentals and operational availability

The series

What do twenty vintages of forest fires show?

One sample chart open to everyone, 4 in the full analysis. Each carries its question, what to see in it, what cannot be concluded from it and what it does not measure.

1 free chart4 premium chartsThe premium panels work — they are deliberately closed. Click one to see what it contains.

How many hectares burn each year in France?

FreeObserved 2006-2025hectares

Reading

Area covered, all vegetation
Of which area declared as forest

Source — National forest-fire database — export by vintage, all of France, forest fires. Axis: The axis starts from zero: these are areas, and a truncated base would visually exaggerate gaps that are already considerable.

What to look for

One year in twenty dominates all the rest: 2022 alone covers more than double any other year in the series.
The forest share varies strongly from one vintage to another: the same area covered does not destroy the same thing from one year to the next.
The lowest years of the series are more recent than the highest: the trend is not monotonic.

Why it matters

The only economic assessment available covers a single one of these years — the most exceptional. Everything that public debate thinks it knows about the cost of fires comes from that point.

Reading of the selected year

Hover over the chart to read a year in the tooltip, click to hold it here — or use the keyboard arrows.

What this does not mean

This is not a trend: inter-annual variance dominates any underlying movement, and twenty points allow no break to be tested.
This is not the area destroyed: inside a perimeter covered, islands may not have burned.
This is not comparable with European areas mapped by satellite, which retain only fires of about thirty hectares and more.

What this chart cannot show, and how to use it

Limits

Definitions change from 2023, after the merger of the national database with a regional one. A change that crosses this boundary mixes a season effect and a definition effect.
The sum of the areas broken down by cover is lower than the area covered over several vintages: cover detail is not entered for every fire.
The current year is not released by the national database, and therefore does not appear here.

Interactions

Hover over the plot: a tooltip gives the point's values, unit and status. Click to hold the selection — it does not disappear when the pointer moves away, so that the reading stays displayed while you read it. On the keyboard, the tab key brings focus to the plot, the arrow keys move the selection, Home and End go to the ends, and Escape closes. On a touch screen, one press selects a point, a second releases it.

Back to the answer

Reserved chapter · the initial attack

Why are some outbreaks contained, and others not?

What plays out in the first minutes decides everything else. France can describe each of these minutes; it cannot say what they are worth.

Reserved · full analysis

What plays out before a fire becomes an event: the measured distribution of fire sizes over 20 vintages, the 5 link by link chain from outbreak to outcome with what each step allows and forbids asserting, the reading of causes according to their denominator, and the economic hinge that France does not publish — what a minute gained, one more drop, a hectare spared is worth

Back to the answer

Reserved chapter · allocation

What becomes of the system when several fires start at the same time?

The chapter on capacity said what the fleet is. This one says what remains of it when several competing needs appear in the same hour.

Reserved · full analysis

What becomes of the system when several fires start at the same time: a documented episode reconstructed step by step, 3 calculated allocation states under assumptions all displayed, the availability series by type of aircraft set against the intensity of each season, and what the sources say — and do not say — about night intervention

Back to the answer

Reserved chapter · the territory

How does a fire reach the economy of a territory beyond the burned zone?

This is the chapter where French measurement stops earliest — and also where there is the most to lose.

Reserved · full analysis

How a fire reaches the economy of a territory beyond the burned zone: the three non-overlapping zones, the figure that shows 5 measured quantities facing 9 quantities that nobody publishes, the comparator of 6 dimensions between a burned establishment and a merely empty one, the six economic quantities that never add up, and the strict separation of the three insurance objects

Back to the answer

Reserved chapter · the capacity to come

Ordered today, delivered when: what does a capacity that is not yet there cost?

This chapter does not look for a culprit. It reconstructs what was known, what was decided, what was paid, and when the capacity actually arrives.

Reserved · full analysis

The double-track chronology — 15 dated events split between what was observed and what was decided, on a single time axis where the gap can be seen without a word accusing it — the price of the same aircraft according to 3 scopes that differ by a factor of two, and 9 aircraft of the industrial ecosystem with their exact status, from in service to design, with no duel

Back to the answer

Reserved chapter · the trade-off

Prevent, or pay?

The module below is arithmetic. It predicts nothing, models no effect, and asserts no causality — all its assumptions are on screen.

Reserved · full analysis

The threshold module: 36 calculated combinations of amount invested, amortisation horizon and discount rate, each rendered in one sentence — what an expenditure would have to avoid each year to break even — then set against the 4 damage layers that France can quantify. Arithmetic, assumptions on screen, with no causal model

Reserved chapter · measurement

Why do three measures of the same area not give the same number?

No series in this study is built before this table. It explains the gaps by the data, and forbids the operation that everyone would be tempted to perform.

Reserved · full analysis

The table of scopes: 5 measurement systems set against one another term by term, and 3 divergences explained by the data — with the operation that each forbids

Open to all · the live layer

The current season, as of 15 September 2026

The values of this season are situation reports. None is a balance, and the reference measure of the vintage will not exist for several months. The European system dates the end of the season; the official national figure stops on 31 July.

Quantity	Value	Producer	As of	Status
Area covered since 1 Januaryoperational record, revisable upwards as well as downwards. It is not an annual balance, and a check on 23 August 2026 had brought back in the press non-matching ministerial situation reports for the same month — checked on 15 September 2026: no later national situation report published by the producer.	about 119,000 ha	Ministry of the Interior	31 July 2026	PROVISIONAL
Area covered, consolidated vintagethis is the only measure that will become comparable with the closed series. It will not exist for several months.	not available	National forest-fire database	15 September 2026	NOT AVAILABLE

The areas covered of the operational chain and the areas mapped by the European system — EFFIS — cannot be subtracted from one another. The vintage's reference measure, that of the national database, will not exist for several months.

What these values are not

None is a final tally. All are revisable in both directions, and each will be re-queried with its producer before any publication.

Reserved · full analysis

The current season in full: all the national situation reports with their producer and their status, the labelled estimate for the day of the edition and its method, France in fire-hit Europe on a single satellite measure, and the Gironde case of the reference year

Open to all

What is said, and what the figures make of it

Each verdict is calculated from the position of the study's frozen conclusions. None is written by hand.

3 of the 9 claims in circulation, set against the 29 frozen conclusions of this study; the others are in the full study. 5 others were set aside, and the reason is written further down.

The verdict is not written: it is calculated. Each conclusion of the study is declared as going in the direction of the claim, against it, or relevant without settling — and a single rule turns these positions into a verdict. Conclusions on both sides give “misleading”; no conclusion that settles gives “cannot be settled”.

False · 2
Cannot be settled · 1

False
“France has twelve amphibious water bombers, so it can commit twelve.”
Between what the State owns and what can take off for a new outbreak, four filters apply — and the last is published nowhere.
What this verdict does not cover: the study does not say either how many aircraft are really available at a given moment: it says that nobody publishes it, and that the quantities published answer other questions.
See why— 4 conclusions of the study
Why this idea circulates: the inventory is public and the number is memorable. Nothing in the way it circulates signals that between the stock and the commitment come maintenance, the state of the day and fires already in progress.
- relevant, without settling · value withheld
  The study distinguishes the inventory of civil-security aircraft, type by type, from their capacity to be committed: an inventory is not a capacity.
- goes against it · value withheld
  The study serves the published series of technical availability of amphibious water bombers excluding annual inspection, and says what this definition neutralises.
- goes against it · value withheld
  The study serves the latest dated published point of operationally available amphibious water bombers, as an observation at a date and never as a rate.
- goes against it
  No public source publishes what remains mobilisable for an outbreak that has not started, once aircraft under maintenance and those already committed are removed. Checked on 15 September 2026.
What would overturn the verdict: a publication giving, over the same period, technical availability, operational availability and deployment, with their definitions.
Read the chapter that demonstrates it
False
“Forest fires are covered by the natural-disaster regime.”
Vegetation fires do not fall under this regime: they fall under the fire cover of the contract. The surcharge that has risen finances something else, and has nothing to do with forest fires.
What this verdict does not cover: knowing which cover a claim falls under does not say what it costs: the only charge published is an episode review by the trade federation, provisional, for the summer 2026 fires in three départements — and no public statistic follows it per financial year.
See why— 2 conclusions of the study
Why this idea circulates: the words “natural disaster” describe exactly what a victim has just lived through, and the surcharge that finances this regime appears on their premium notice. Two concordant clues, and a false conclusion.
- goes against it
  Forest and vegetation fires do not fall under the natural-disaster regime. They fall under the fire cover of contracts.
- relevant, without settling · value withheld
  The study serves the market context of home insurance, and explains why it says nothing about the fires of a later year.
What would overturn the verdict: a sector publication giving, for each financial year and not by episode, the claims charge for vegetation fires isolated from other perils.
Read the chapter that demonstrates it
Cannot be settled
“More aircraft means proportionally fewer hectares burned.”
The study does not settle it, and it explains why: resources are sent first to the most serious fires, so that the relationship between deployment and final area is inseparable from the severity that triggered the deployment. No French data allows it to be disentangled.
What this verdict does not cover: “we cannot settle it” does not mean “it is false”. The study asserts neither that aerial resources are effective, nor that they are not: it asserts that France does not publish what is needed to know.
See why— 4 conclusions of the study
Why this idea circulates: the intuition is solid, and the doctrine of attacking nascent fires reinforces it: reaching an outbreak early changes its outcome. From there to a measurable proportionality there is a step that nothing takes.
- relevant, without settling · value withheld
  The study measures, on the national database, how far burned area concentrates on a tiny share of outbreaks — and shows that this concentration varies strongly from one vintage to another.
- relevant, without settling · value withheld
  The study serves the published series of technical availability of amphibious water bombers excluding annual inspection, and says what this definition neutralises.
- relevant, without settling · value withheld
  The study sets the rate claimed by the intervention doctrine against the closest primary measure, and shows that they do not concern the same object.
- relevant, without settling · value withheld
  The study establishes that an essential economic hinge — what a hectare costs depending on the outcome of the fire — is published by no French producer.
What would overturn the verdict: a French causal identification, published with its identification strategy — exogenous variation in availability, or a documented natural experiment.
Read the chapter that demonstrates it

Five claims that this study cannot test

“Arsonist firefighters account for a significant share of outbreaks.”
no national denominator exists: neither the number of firefighters implicated nor the share of fires concerned is published by profession. A share without a denominator is not a share.
“Mandatory brush clearing returns several euros per euro spent.”
no French ratio. The values in circulation are transpositions of studies conducted on other continents, with other fire regimes, other labour costs and other forms of housing.
“One euro spent on forest defence facilities avoids twelve of damage.”
no stabilised econometric model produces this bound, in France or elsewhere. It circulates with no published method.
“A third simultaneous fire breaks the system.”
it is an internal doctrinal notion, never published with its method nor with the measurement that would underpin it. No French break threshold exists in a public source.
“Smoke from the fires caused an identifiable number of deaths in France.”
no French epidemiological study isolates the effect of smoke from that of the concurrent heatwave. Without this separation, any figure would mix two causes that occur at the same time.

Open to all

Where do these figures come from?

Each series used, with its producer, its role, its layer, its cut-off date and what it does not measure.

Series	What it contributes	Layer	What it does not measure
Ministry of Agriculture and Food Sovereignty — BDIFFas of: 2025	the national reference series: number of fires, area covered, size distribution, recorded causes	certified history	it does not measure the area seen by satellite, carries no cost, and does not release the current year
General Commission for Sustainable Development — statistical data and studies serviceas of: April 2025, on the year 2022	the only public assessment of the economic consequences of the 2022 fires, and its parameters	documentary foundation	it measures neither buildings, nor jobs, nor overnight stays, nor insurance compensation, nor firefighting
France ÉcoScan, from the definitions published by each systemas of: 2026-09-15	the table that explains why three measures of the same area differ	documentary foundation	it reconciles no gap: it explains it, and forbids subtracting it
Senate — information report r22-838 and budget opinion a25-145-13as of: 1 July 2025 for the fleet status · to be re-queried before publication	the fleet, its five availability quantities, orders, schedules and costs by scope	documentary foundation	none of these quantities says what remains mobilisable for an outbreak that has not started
Ministry of the Interior, Ministry of Agriculture, Gironde prefecture, Copernicus EFFISas of: 2026-09-15 · to be re-queried before publication	the state of the current season, each cell carrying its date and its producer	live season	none of these values is an annual balance, and none is consolidated
Gironde prefecture, regional tourism committee, national pressas of: 2026-09-15 · to be re-queried before publication	what the tourism economy allows to be measured, and the nine quantities that nobody publishes	live season	no consolidated loss of turnover, no occupancy rate, no tax revenue
Insurance Code, France Assureursas of: 2025 for the market, 28 August 2026 for the first review of the summer fires · to be re-queried before publication	the three insurance objects that public debate constantly confuses	documentary foundation	no national charge per financial year isolating vegetation fires; the only amounts are episode reviews by the trade federation — summer 2022, fires of summer 2026
Senate, BDIFF, CGDD, France Stratégieas of: 2026-09-15 · to be re-queried before publication	what was observed, what was decided, what is announced — three natures that do not mix	documentary foundation	an announced delivery date is not a fact, and a gap between alert and decision is not an attribution
Senate, manufacturers' press releasesas of: 2026-09-15 · to be re-queried before publication	the industrial ecosystem with strict statuses — in service, firm order, prototype, design, claim	documentary foundation	no announced performance is a measured performance
Ministry of Agriculture and Food Sovereignty — BDIFFas of: 2022	the concrete cases, Gironde and Brittany, drawn from the same database as the national series	certified history	neither evacuations, nor destroyed campsites, nor resources deployed: they come from other producers
France ÉcoScan, on primary documentsas of: 2026-09-15	one row per quantified claim, with its level of proof and its right of publication	documentary foundation	it measures nothing: it says what may be served, and what may not

Open to all · by necessity

What does the foundation not allow us to conclude?

This block says what this study cannot settle. Reserving it would amount to selling the only information that prevents mistakes about what one is buying.

Insurance compensation
no producer publishes, per financial year and for the whole of France, a claims charge that isolates forest and vegetation fires from other perils. The only amounts published are episode reviews by the trade federation — for summer 2022, and, provisionally, for the summer 2026 fires in three départements —, never a charge per financial year.
Why it matters: this is the item that turns destruction into money actually paid. Without it, physical destruction remains outside any economic accounting.
What would close it: a publication by France Assureurs or the Caisse centrale de réassurance giving, for each financial year and not by episode, the claims charge for vegetation fires isolated from other perils.
Consolidated tourism loss
no measure of the loss of tourist activity exists, either for 2022 or for 2026. The values in circulation are press estimates whose method is not published.
Why it matters: tourism is explicitly excluded from the only public assessment available. It is therefore, literally, a quantity that nobody counts.
What would close it: a post-season attendance survey, published at a usable territorial grain and compared with a reference season.
The cost of a hectare depending on the outcome of the fire
no French primary document publishes what a hectare treated through prevention costs compared with a hectare contested in active firefighting.
Why it matters: this is the hinge that would allow a prevention expenditure to be compared with a firefighting expenditure. Its absence is the reason why no prevention policy can be shown to pay off — nor shown to be useless.
What would close it: a publication by INRAE, ONF or the directorate-general for civil security giving a unit cost with its declared scope.
What remains mobilisable for the next outbreak
no public source says how many resources remain available at a given moment, once those under maintenance and those already committed are removed.
Why it matters: this is the only quantity that would answer the question everyone asks during a tense season. The four quantities published answer other questions.
What would close it: a publication by the directorate-general for civil security giving, over the same period, technical availability, operational availability and deployment, with their definitions.
The finances of fire and rescue services
the consolidated budgets of the fire and rescue services have not been acquired, and their nomenclature needs harmonising before any series.
Why it matters: they carry most of the public spending on firefighting. Without them, the study can say nothing about the total cost of the response.
What would close it: the accounts of the fire and rescue services, in constant euros, on a stable nomenclature.
The standardised European series
no machine access point was found for the European estimates by country. The only values usable are those that a primary document cites.
Why it matters: this is the measure that would allow France to be compared with its neighbours on a homogeneous scope.
What would close it: a data request to the European forest-fire information service, or reading its annual reports.
The end of production of the amphibious water-bomber aircraft
no primary document acquired dates or documents the end of production of the Canadair CL-415.
Why it matters: this is the fact that explains why several years separate the signing of an order from the delivery of the aircraft. Without it, the gap between decision and delivery reads as French slowness, which it is not.
What would close it: a document from the manufacturer or a parliamentary source dating the end of the line and its industrial reason.
The cost of reconstituting a burned forest
no public scale for post-fire reconstitution has been acquired. The range in circulation has no primary document.
Why it matters: this is the item that runs over the ten to forty years following the fire, and the only public assessment stops at the year of the fire.
What would close it: a public scale of post-fire reconstitution, with its scope of species and works.
Health effects of smoke
no French epidemiological study isolates the effect of wildfire smoke from that of the concurrent heatwave.
Why it matters: without this separation, any mortality figure attributed to smoke would mix two causes that occur at the same time.
What would close it: a French exposure study controlling for temperature, over a declared period and territory.

Open to all

What exactly does this study assert?

Each conclusion, with its nature, period, scope, source, formula and limits. The values of the reserved chapters appear there without their figure.

Published fact
A value published by a named producer, reproduced as is with its scope.
Derived result
A value calculated by the study from published facts, under written and reproducible conventions.
Result under a declared parameter
A value that also depends on an assumption that the study itself sets, and which is displayed with it.
Absence established
The verification that a quantity is published by no producer. It is a conclusion obtained by searching, and it carries its date.

Derived resulthigh confidence
The only public assessment of the 2022 fires is €178M, and it adds up exactly three environmental items; the study redoes it item by item from the parameters that the source publishes.
Value:
€178M
Unit:
millions of euros
Period:
year 2022
Scope:
three environmental items only, metropolitan France, under the attribution assumption declared by the source
Formula:
carbon item = 2022 emissions × (1 − 1 / area ratio) × value of climate action; timber and recreational items = annual value × share of area burned; total = sum of the three
Revision:
frozen — one-off study published in April 2025
Sources:
General Commission for Sustainable Development, “Sécheresse de 2022 : des coûts estimés à plus de 5 milliards d'euros” (2022 drought: costs estimated at more than €5 billion), April 2025
- this value assumes that all of the area burned in 2022 is attributable to the drought. It is an assumption declared by the source, not a result.
- this total adds a differential — carbon, measured as a gap to an average — and two levels pro rata to area. The operation comes from the source.
- it contains no building, no job, no overnight stay, no insurance compensation.
Derived resulthigh confidencedemonstrated in a reserved chapter
The same assessment moves within a published range depending on the bound used for the recreational value of the forest, with no other assumption changing.
Value:
reserved
Unit:
millions of euros
Period:
year 2022
Scope:
identical to the central assessment
Formula:
three exact calculations, differing only by the bound used for the recreational item
Revision:
frozen — one-off study published in April 2025, no update announced
Sources:
General Commission for Sustainable Development, “Sécheresse de 2022 : des coûts estimés à plus de 5 milliards d'euros” (2022 drought: costs estimated at more than €5 billion), April 2025
- this range is neither a margin of error nor a confidence interval: these are three exact calculations, and nothing makes the central value more likely than either bound.
- the central value is the average of a published range, not a measurement.
Published facthigh confidencedemonstrated in a reserved chapter
In the study it comes from, this assessment is a secondary line: the fires are classed there among the indirect socio-economic impacts of the drought.
Value:
reserved
Unit:
share of the study's total
Period:
year 2022
Scope:
study on the 2022 drought, five items, metropolitan France
Formula:
fires item relative to the exact sum of the five items of the study
Revision:
frozen — one-off study published in April 2025, no update announced
Sources:
General Commission for Sustainable Development, “Sécheresse de 2022 : des coûts estimés à plus de 5 milliards d'euros” (2022 drought: costs estimated at more than €5 billion), April 2025
- the sum of the five rounded labels in the source's chart gives a total different from the exact sum of the items. It is a label rounding, and the study does not correct it.
Published facthigh confidencedemonstrated in a reserved chapter
The source itself declares the quantities it leaves outside its scope and judges its overall cost undervalued; the study names them one by one.
Value:
reserved
Unit:
unitless
Period:
year 2022
Scope:
statements by the source about its own scope
Formula:
direct reading of the document
Revision:
frozen — statements by the source about its own scope
Sources:
General Commission for Sustainable Development, “Sécheresse de 2022 : des coûts estimés à plus de 5 milliards d'euros” (2022 drought: costs estimated at more than €5 billion), April 2025
- a declared exclusion does not say how much what is excluded is worth. None of these six quantities is quantified elsewhere.
Published facthigh confidencedemonstrated in a reserved chapter
The source cites separately an increase in firefighting spending for the same year, outside the assessment of the three items, and the study never adds it to them.
Value:
reserved
Unit:
millions of euros
Period:
year 2022
Scope:
outside the assessment of the three items — the body of the text introduces it with “could also be added”
Formula:
published value, not recalculated
Revision:
frozen — value published in April 2025 for a closed financial year
Sources:
General Commission for Sustainable Development, “Sécheresse de 2022 : des coûts estimés à plus de 5 milliards d'euros” (2022 drought: costs estimated at more than €5 billion), April 2025
- the source's methodological annex suggests the opposite of the body of its text. The study retains the body of the text, which the arithmetic confirms, and flags the discrepancy rather than settling it in silence.
- the sum of this amount and the assessment of the three items bears no name in this study.
Published facthigh confidencedemonstrated in a reserved chapter
The source itself acknowledges a limit of its carbon item, and the study says what this acknowledgement allows one to conclude — and what it forbids.
Value:
reserved
Unit:
unitless
Period:
note from the source, on the year 2022
Scope:
carbon item only
Formula:
direct reading of the document's note
Revision:
frozen — note from the source, no update announced
Sources:
General Commission for Sustainable Development, “Sécheresse de 2022 : des coûts estimés à plus de 5 milliards d'euros” (2022 drought: costs estimated at more than €5 billion), April 2025
- the raised value is neither on the same monetary base nor on the same gas scope. The ratio between the two is not an evolution, and substituting it would not be an update.
Published facthigh confidencedemonstrated in a reserved chapter
The study distinguishes the inventory of civil-security aircraft, type by type, from their capacity to be committed: an inventory is not a capacity.
Value:
reserved
Unit:
aircraft on the books
Period:
1 July 2025
Scope:
civil-security aircraft, excluding helicopters and excluding rentals
Formula:
published inventory
Revision:
to be queried again before publication
Sources:
Senate, budget opinion on the Civil Security programme
- an inventory is not a capacity to be committed: it says neither what flies, nor what is already committed.
Published facthigh confidencedemonstrated in a reserved chapter
The study serves the published series of technical availability of amphibious water bombers excluding annual inspection, and says what this definition neutralises.
Value:
reserved
Unit:
percentage
Period:
2016 to 2022
Scope:
contractual availability with respect to the maintenance provider, neutralising the major winter maintenance inspection
Formula:
published series, not recalculated
Revision:
frozen — the published series stops in 2022
Sources:
Senate, information report on the civil-security water-bomber aircraft fleet
- this quantity neutralises scheduled maintenance that the fire season does not neutralise.
- it is never concatenated with operational availability, which measures neither the same thing nor the same aircraft scope.
Published facthigh confidencedemonstrated in a reserved chapter
The study serves the only published points of operational availability of civil-security aircraft, against the displayed target, without making a trend of them.
Value:
reserved
Unit:
percentage
Period:
2023 and 2024
Scope:
all civil-security aircraft, not only amphibious water bombers
Formula:
published values, not recalculated
Revision:
to be queried again before publication
Sources:
Senate, budget opinion on the Civil Security programme
- two points do not make a trend, and the aircraft scope differs from that of the previous series.
Published facthigh confidencedemonstrated in a reserved chapter
The study serves the latest dated published point of operationally available amphibious water bombers, as an observation at a date and never as a rate.
Value:
reserved
Unit:
aircraft
Period:
1 July 2025
Scope:
aircraft operationally available on that date
Formula:
published value
Revision:
to be queried again before publication
Sources:
Senate, budget opinion on the Civil Security programme
- a dated point is not derived into a rate, not averaged, and not carried over to another date.
Absence establishedhigh confidence
No public source publishes what remains mobilisable for an outbreak that has not started, once aircraft under maintenance and those already committed are removed. Checked on 15 September 2026.
Value:
no source
Unit:
unitless
Period:
check of 15 September 2026
Scope:
budget documents and parliamentary reports acquired
Formula:
search in the two primary documents acquired, on the five fleet quantities
Revision:
to be queried again at each budget publication
Sources:
Senate, information report on the civil-security water-bomber aircraft fleet · Senate, budget opinion on the Civil Security programme
- an absence in two documents is not an absence in all. It is established on the foundation acquired, and the study says so.
Published factmedium confidence
Forest and vegetation fires do not fall under the natural-disaster regime. They fall under the fire cover of contracts.
Value:
outside the regime
Unit:
unitless
Period:
law in force
Scope:
France, property-damage insurance contracts
Formula:
rule of law
Revision:
stable
Sources:
Insurance Code
- the exact reference of the article that excludes vegetation fires from the scope of the regime remains to be entered in the study.
Derived resulthigh confidencedemonstrated in a reserved chapter
The study measures, on the national database, how far burned area concentrates on a tiny share of outbreaks — and shows that this concentration varies strongly from one vintage to another.
Value:
reserved
Unit:
share of area covered
Period:
2006 to 2025
Scope:
all of France, forest fires, no area threshold
Formula:
areas sorted in decreasing order, sum of the first hundredth relative to the total
Revision:
closed vintages now move only at the margin
Sources:
National forest-fire database, Ministry of Agriculture and Food Sovereignty, export of 23 August 2026
- this measure is DESCRIPTIVE. It produces no threshold, and justifies no cut-off at a round number of hectares.
- it says nothing about the causes of the concentration: weather, fuel, resources committed and chance are mixed without being separable.
Derived resulthigh confidencedemonstrated in a reserved chapter
The study measures the share of outbreaks that stop small, and compares it with the rate that the intervention doctrine claims.
Value:
reserved
Unit:
share of fires
Period:
2006 to 2025
Scope:
all of France, forest fires, no area threshold
Formula:
number of fires under one hectare relative to the number of fires of the vintage
Revision:
closed vintages now move only at the margin
Sources:
National forest-fire database, Ministry of Agriculture and Food Sovereignty, export of 23 August 2026
- this share is NOT the rate claimed by the doctrine: one counts fires brought under control, the other fires that stayed small. The two are never served for one another.
- a fire that stayed small does not prove that a resource stopped it: it may never have had what it takes to grow.
Derived resultmedium confidencedemonstrated in a reserved chapter
The study measures the share of fires whose cause is actually recorded, and shows what a share of cause becomes depending on the denominator given to it.
Value:
reserved
Unit:
share of fires whose cause is recorded
Period:
year 2022
Scope:
all of France, on only the fires whose cause is entered
Formula:
number of fires of a cause relative to the number of fires whose cause is recorded
Revision:
the entry of causes is completed after the close of the vintage
Sources:
National forest-fire database, Ministry of Agriculture and Food Sovereignty, export of 23 August 2026
- every share of cause is written with its denominator. A share calculated on the recorded fires is never presented as a share of outbreaks.
- more than half of outbreaks remain without a recorded cause: the distribution observed is not necessarily that of the whole.
Derived resultmedium confidencedemonstrated in a reserved chapter
The study sets the rate claimed by the intervention doctrine against the closest primary measure, and shows that they do not concern the same object.
Value:
reserved
Unit:
percentage
Period:
2006 to 2025 for the measure
Scope:
doctrinal declaration on one side, national database on the other
Formula:
comparison of two declared quantities that are not identical
Revision:
stable
Sources:
Senate, information report on the civil-security water-bomber aircraft fleet · National forest-fire database, Ministry of Agriculture and Food Sovereignty, export of 23 August 2026
- the two quantities are not identical, and the study does not claim that they are. The gap only forbids serving the doctrinal rate as an observation.
- the source itself words this rate in two different ways in the same document, one with an area threshold and the other without.
Absence establishedhigh confidencedemonstrated in a reserved chapter
The study establishes that an essential economic hinge — what a hectare costs depending on the outcome of the fire — is published by no French producer.
Value:
reserved
Unit:
unitless
Period:
check of 15 September 2026
Scope:
search for a French primary document on prescribed burning and on active firefighting
Formula:
targeted search, no usable result
Revision:
to be queried again at each publication of the agronomic research institute or the forestry office
Sources:
documentary search of the study
- orders of magnitude circulate, but none has a primary document acquired, and none declares its scope. The study therefore does not publish them.
Derived resulthigh confidencedemonstrated in a reserved chapter
The study shows that the price of the same aircraft doubles depending on the scope used, and names what each scope contains.
Value:
reserved
Unit:
millions of euros per aircraft
Period:
2024 contract and 2026 budget
Scope:
three distinct scopes, named by what they contain
Formula:
European share published for two aircraft; the other two scopes are divisions documented by the number of aircraft
Revision:
to be queried again at each budget year
Sources:
Senate, budget opinion on the Civil Security programme
- two of the three scopes are reconstructions by division, and not published unit prices.
- a commitment authorisation is a ceiling that the State authorises itself, not an observed purchase price.
Published factmedium confidencedemonstrated in a reserved chapter
The study reconstructs the gap between the moment a capacity is decided and the moment it is announced as available.
Value:
reserved
Unit:
announced years
Period:
from 2024 to 2033
Scope:
announced schedules, never observed deliveries
Formula:
published dates, served with their nature
Revision:
to be queried again at each budget year
Sources:
Senate, budget opinion on the Civil Security programme
- an announced delivery date is an industrial objective, not a fact.
- the end of production of the previous generation, the manufacturer's re-industrialisation rate and the international queue are not French decisions.
Published factmedium confidencedemonstrated in a reserved chapter
The study serves the only employment quantity actually measured during the season, with what it is not.
Value:
reserved
Unit:
applications and employees
Period:
1 August 2026
Scope:
Gironde, applications filed
Formula:
published values
Revision:
provisional — to be queried again before publication
Sources:
Gironde prefecture, situation report of 1 August 2026
- an application is neither an acceptance, nor a payment, nor a job loss. The number of hours compensated is not published.
- provisional value for a season in progress, revisable in both directions, to be queried again with its producer before publication.
Published factmedium confidencedemonstrated in a reserved chapter
The study serves the economic weight of tourism in the département concerned, and explains why this weight converts into no loss.
Value:
reserved
Unit:
billions of euros of spin-offs
Period:
6 August 2026
Scope:
Gironde, annual tourism spin-offs
Formula:
values published by a sector producer
Revision:
provisional — to be queried again before publication
Sources:
Regional tourism committee and tourism partners, 6 August 2026
- it is a WEIGHT, never a loss. Multiplying these spin-offs by a cancellation rate would confuse spin-offs with value added, and would assume a uniform cancellation over the year.
- spin-offs are not value added, and are added to no gross domestic product.
Published factsubject to reservation confidencedemonstrated in a reserved chapter
The study serves the only estimate of tourism loss available, with its producer, its date and its status as an estimate.
Value:
reserved
Unit:
millions of euros
Period:
16 August 2026
Scope:
Gironde and Landes, press estimate
Formula:
none — the method is not published
Revision:
provisional — to be queried again before publication
Sources:
national press, 16 August 2026
- this estimate enters no total of this study, and is compared with no value of another nature.
- provisional value for a season in progress, revisable in both directions, to be queried again with its producer before publication.
Published factmedium confidencedemonstrated in a reserved chapter
The study serves the first insurance review of the summer 2026 fires, with its producer, its date, its scope and its provisional status.
Value:
reserved
Unit:
millions of euros of estimated insured damage
Period:
fires since 21 July 2026, figures reported to mid-August, review of 28 August 2026
Scope:
Gironde, Landes, Var — claims declared to insurers
Formula:
published values
Revision:
provisional — first review liable to change, declarations open until 31 August 2026
Sources:
France Assureurs, statement of 28 August 2026 on the fires in the Gironde, the Landes and the Var
- these are estimated insured damages, not compensation paid; uninsured property and deductibles are not included.
- an episode review over three départements is compared with no national annual charge, and is added to no other quantity of this study.
- the amount includes the rehousing costs covered; business-interruption losses cover only establishments insured for this risk.
Published factmedium confidencedemonstrated in a reserved chapter
The study serves the only insurance amount published for the 2022 fires, with its producer and its summer scope.
Value:
reserved
Unit:
millions of euros of claims cost
Period:
summer 2022, 1 June – 30 September
Scope:
forest fires occurring in France, claim declarations to insurers
Formula:
value published, for comparison, in the review of the 2026 fires
Revision:
cited in 2026 for a closed summer; neither the cut-off date nor the exact scope of the covers is published
Sources:
France Assureurs, statement of 28 August 2026 on the fires in the Gironde, the Landes and the Var
- it does not enter the 2022 public assessment, which places compensation outside its scope, and is not added to it.
- a summer is not a financial year: this amount does not say the annual charge of the peril.
Absence establishedhigh confidencedemonstrated in a reserved chapter
The study establishes that nine of the quantities that would make up the economic cost of a fire are published by nobody, and names them one by one.
Value:
reserved
Unit:
unitless
Period:
check of 15 September 2026
Scope:
hotels, open-air accommodation, restaurants, leisure, occupancy, local taxation, aid, insurance charge per financial year, consolidated loss
Formula:
search for a producer for each of the nine quantities
Revision:
attendance surveys appear after the season
Sources:
documentary search of the study
- two of these quantities will exist later — occupancy and local taxation. The other seven have no production system identified.
Absence establishedhigh confidencedemonstrated in a reserved chapter
The study establishes that the insurance charge for vegetation fires is published only by episode, never as a series per financial year.
Value:
reserved
Unit:
unitless
Period:
check of 15 September 2026
Scope:
publications of the trade federation and of the public reinsurer
Formula:
search for a breakdown by peril and by financial year in sector publications
Revision:
to be queried again at each sector publication
Sources:
documentary search of the study · France Assureurs, statement of 28 August 2026 on the fires in the Gironde, the Landes and the Var
- claims are compensated every year, and two episodes are quantified — summer 2022, the fires of summer 2026: the absence concerns the SERIES per financial year, not the existence of compensation nor any amount.
Published factmedium confidencedemonstrated in a reserved chapter
The study serves the market context of home insurance, and explains why it says nothing about the fires of a later year.
Value:
reserved
Unit:
annual percentage change
Period:
financial year 2025
Scope:
French home insurance market, all perils combined
Formula:
published values
Revision:
annual — to be queried again before publication
Sources:
France Assureurs, financial year 2025
- it is a MARKET CONTEXT for a closed financial year, never an effect of the fires of a later year.
- the fire-cover claims charge covers all fires, the vast majority of which have nothing to do with forests.
Result under a declared parameterhigh confidencedemonstrated in a reserved chapter
The study calculates, under displayed assumptions, the amount of damage that an expenditure would have to avoid each year to break even — and sets it against the only layers that France can quantify.
Value:
reserved
Unit:
millions of euros per year
Period:
no period — it is an arithmetic
Scope:
expenditure assumptions entered by the reader, all displayed
Formula:
annual threshold = discounted value of costs ÷ annuity factor
Revision:
not applicable
Sources:
break-even threshold engine of the study, discounted arithmetic
- this module calculates a threshold, not a result. It does not say that an expenditure avoids this damage: it says what it would have to avoid in order to break even.
- the reduction in probability is an assumption entered, never a measured effectiveness. No French data allows it to be identified.
Published facthigh confidencedemonstrated in a reserved chapter
The study quantifies the year the fire left the South, and compares it with the normal of the previous sixteen vintages.
Value:
reserved
Unit:
fires and hectares
Period:
year 2022, compared with 2006–2021
Scope:
four Breton départements, forest fires, no area threshold
Formula:
departmental aggregation of the national database, arithmetic mean over sixteen vintages
Revision:
frozen — closed vintages of the national database
Sources:
National forest-fire database, Ministry of Agriculture and Food Sovereignty, export of 23 August 2026
- this case is a GEOGRAPHICAL CONTRAST. No cost, no compensation, no loss of activity is published for these fires: it carries no economic value.
- the average of the previous sixteen vintages is a benchmark, never a norm or a forecast.
- the areas of the national database and those of the European system do not measure the same thing and are not subtracted.

Open to all

How is this study made?

The reproduction of the public assessment, the sentences the study forbids itself, and the strings that an automatic check prevents from appearing on the page.

The calculation redone, rather than copied

The public assessment is not taken over as it stands: it is recalculated from the parameters its source publishes — the year's emissions, the area ratio, the value of climate action, the annual forest values. The calculation finds exactly the published values, and it is by redoing it that one discovers that the heaviest item is the mean of a range.

reproduction of the 2022 public assessment from its published parameters; national series aggregated fire by fire; break-even threshold by pure discounted arithmetic, with no causal model.

Back to the thesis

What would refute this thesiswritten before looking at the results, as the protocol requires

The publication of a homogeneous national accounting of wildfire damage — firefighting, destruction, activity, insurance, health, carbon, reconstitution — over a declared scope and period.
Consequence: the thesis falls entirely: France would then know what a fire costs, and the study would no longer have a purpose.
The publication, by the directorate-general for civil security, of what remains mobilisable at a given moment, with its definition.
Consequence: the fifth stage of the funnel would cease to be a question and become a measurement. Two chapters would change nature.
The publication of a cost per hectare depending on the outcome of the fire, with its scope, by a French producer.
Consequence: the missing economic hinge would exist, and the trade-off between prevention and capacity would become quantifiable instead of merely boundable.
A French causal identification of the effect of a firefighting resource on the final area, published with its identification strategy.
Consequence: the claim “more aircraft, proportionally fewer hectares” would cease to be undecidable, in either direction.

The 12 sentences this study forbids itself to writethe shortcuts a hurried reader would take

Never present the 2022 assessment as “the cost of the 2022 fires”. It values three consequences over a declared scope, under an attribution assumption that its source itself writes.
Never add the assessment of the three items and the firefighting extra cost to make a total. The body of the source's text places the second outside the first, and the arithmetic confirms it.
Never present a round area as a scientific escape threshold. The values in circulation are classification conventions or doctrinal objectives, not demonstrated breaks.
Never write that a given number of simultaneous fires breaks the system. No French break threshold is published with its method.
Never write that an additional aircraft avoids a number of hectares, nor that a euro spent avoids another. Resources are sent first to the most serious fires: no French data allows identification.
Never serve an availability rate without saying which of the quantities it measures, nor over which aircraft scope.
Never present the natural-disaster regime surcharge as funding for forest fires, nor derive a premium rise from the fires of a later year.
Never present lost turnover as a loss of value added, nor add two of the six economic quantities of the territorial chapter without a written test for double counting.
Never multiply a sector weight by a cancellation rate to produce a loss. One says what an activity represents, the other what a period cancelled.
Never subtract two area measures from different systems, nor present their gap as under-reporting.
Never write that a military transport aircraft fitted with a drop kit replaces an amphibious water bomber, nor that a unit load converts into a multiple of effectiveness.
Never infer from a technical capacity attributed to a type of aircraft that France has night-time aerial cover, nor that an aircraft to come will have it.

The 18 assertions that cannot appear on this pagean automatic check rejects them

A false assertion is not held back by an editor's vigilance. Each of the following is checked automatically on the text actually served, and its presence would make the production of the study fail.

an official cost of the 2022 fires of the order of a billion and a half euros
Why: no national accounting of fire damage exists in France
an amount of insurance compensation for the 2022 fires
Why: the only amount published for 2022 is the trade federation's summer review, served in the full study; the amount targeted here corresponds to no publication identified
a quantified tourism loss for the Gironde in 2022
Why: no measure of tourism loss exists for 2022
a number of deaths attributable to smoke from the 2022 fires
Why: no French study isolates the effect of smoke from that of the concurrent heatwave
a breakdown of the 2022 assessment into four items of 55, 65, 43 and 15 million
Why: direct reading of the primary document contradicts it on all four items
a total cost of the 2022 fires of two hundred and nine million euros
Why: this number is the addition of two distinct accounting scopes, and it also designates budget commitment authorisations that have nothing to do with the 2022 fires
the same addition, written the other way round
Why: a total named before its amount remains a total
a return on prevention expressed in euros avoided per euro spent
Why: no stabilised econometric model produces this bound, in France or elsewhere
a profitability ratio for mandatory brush clearing
Why: the values in circulation are transpositions of foreign studies
a universal scientific escape threshold at ten hectares
Why: it is a classification convention or a doctrinal objective, not a demonstrated break
an automatic breakdown of the system at the third simultaneous fire
Why: it is an internal doctrinal notion, never published with its method
an equivalence between a military transport aircraft and three water bombers
Why: the primary document writes that this aircraft is not intended to replace amphibious water bombers
a night-drop capacity attributed to the aircraft to come
Why: no document documents it
a price per aircraft of ninety-one million euros including all taxes
Why: this value was found in neither of the two budget documents acquired
an average summer technical availability rate of sixty-eight per cent
Why: this value was found in neither of the two budget documents acquired
the same value, written the other way round
Why: a rate named before its amount remains that rate
a direct cost of eight thousand euros per flight hour
Why: this value is absent from the document acquired
a national share of firefighters implicated in outbreaks
Why: neither the numerator nor the denominator is published in France

Open to all

Reading aids

Each term is also explained where it is used. This block is a reminder, never the only place where the word is defined.

Initial attack

the intervention on a nascent fire, in the minutes after its detection, before it has taken on size. French doctrine devotes most of its rapid resources to it.

What it is not: it is not the first chronological intervention on any fire: it is the intervention during the window in which the fire can still be stopped by ordinary means.

Escaped fire

a fire that the initial attack did not contain, and which now spreads according to weather, terrain and fuel rather than according to the resources committed.

What it is not: it is not a fire that exceeds a given area. No area threshold defines escape: the round values in circulation are classification conventions or objectives, not demonstrated breaks.

Fire weather indexFWI, called IFM (indice forêt-météo) in French publications

a numerical index that combines temperature, humidity, wind and recent rain to estimate the meteorological fire danger of a given day. It drives the pre-positioning of resources.

What it is not: it is not a probability of an outbreak, still less a forecast of area: it describes conditions, not an event.

Prescribed burning

a fire lit deliberately, under chosen weather conditions and with resources in place, to remove fuel before the high-risk season.

What it is not: it is not a tolerated accidental fire. It is a planned operation, and its cost per hectare is published by no French producer.

Water-bomber aircraftABE (French acronym)

an aircraft designed or modified to drop water or retardant on a fire. Amphibious models scoop in flight from a body of water; the others reload on the ground.

What it is not: a water-bomber aircraft does not put out a fire on its own: it slows a front, or protects an axis, so that ground teams can work.

Water-bomber helicopterHBE (French acronym)

a helicopter equipped to take and drop water. Heavy models chain ten to fourteen drops an hour when a water point is close.

What it is not: it is not a scaled-down version of an aircraft: rate, precision and access to difficult zones cannot be compared term by term with a mass drop.

Armed aerial watchGAAr (French acronym)

the preventive overflight of exposed zones by aircraft already loaded, so that a detected outbreak can be handled in minutes rather than tens of minutes.

What it is not: it is not mere surveillance: the aircraft is armed, that is, ready to drop without returning to its base.

Maintenance in operational conditionMCO (French acronym)

all the maintenance, repair and parts-supply operations that keep an aircraft able to fly.

What it is not: it is not an ancillary cost: it determines how many aircraft are really available on a given day, and it depends on the spare-parts chain.

European Civil Protection MechanismrescEU

the scheme through which the European Union finances and pools rescue resources, including firefighting aircraft, that can be mobilised between Member States.

What it is not: an aircraft financed by this mechanism is not an additional French aircraft at permanent disposal: it follows European rules of engagement.

Forest defence against fireDFCI (French acronym)

all the permanent installations that make a forest defensible: access tracks, water points, fuel breaks, lookout towers.

What it is not: it is not firefighting: these are facilities installed before the season, whose effect can be measured only when a fire occurs where they exist.

Legal brush-clearing obligationsOLD (French acronym)

the obligation on an owner to remove combustible vegetation around their building, within a radius set by local regulation.

What it is not: it is not land clearance: the aim is to reduce the continuity of fuel, not to remove trees.

Area covered

the total area crossed by the fire over its course, whatever the vegetation affected. It is the quantity that the French national database records.

What it is not: it is not the area destroyed: inside a perimeter covered, islands may not have burned, and severity varies strongly.

National forest-fire databaseBDIFF (French acronym)

the public database that records, fire by fire, the date, the commune, the area covered and — when established — the cause. It covers all of France and applies no size threshold.

What it is not: it does not release the current year: the most recent vintage appears several months after its close.

European Forest Fire Information SystemEFFIS

the European system that maps by satellite the perimeter of fires of about thirty hectares and more, on a homogeneous framework across countries.

What it is not: its areas cannot be compared term by term with those of the French national database: neither the size threshold, nor the vegetation retained, nor the period coincide.

Technical availability

the share of a fleet that is not grounded for maintenance. In French publications, it is measured excluding the major winter inspection.

What it is not: it is not what can take off today: it neutralises scheduled maintenance that the fire season does not neutralise.

Operational availability

the share of a fleet actually in a state to be committed on a given date, all causes of unavailability combined.

What it is not: it is not technical availability, and the two are never concatenated: they carry neither the same definition nor the same aircraft scope.

Marginal capacity available

what is actually still mobilisable for a fire outbreak that has not yet started, once aircraft under maintenance and those already committed elsewhere are removed.

What it is not: no French public source publishes it. It is not a quantity that this study estimates: it is a quantity that it declares missing, and that is one of its results.

Direct cost

what an event causes to be spent or destroyed immediately: resources committed, buildings lost, crops burned.

What it is not: it is not necessarily what is measured: in the only public assessment of the 2022 fires, no direct destruction enters the scope.

Indirect cost

what an event causes to be lost without destroying anything: activity that does not take place, a cancelled booking, revenue that does not come in.

What it is not: indirect does not mean secondary. In the study from which the only public assessment comes, the fires themselves are classed among the indirect impacts.

Externality

an economic effect borne by someone who is neither buyer nor seller of the activity that produces it — and which, for that reason, appears on no invoice.

What it is not: it is not an imaginary cost: the carbon emitted and the walk in the forest that will not take place are externalities, and these are precisely what the 2022 assessment values.

Value of climate action

the price per tonne of carbon dioxide that the public authority uses to assess its own decisions. It is not a market price: it is a public valuation convention.

What it is not: it is not a tax, and it is not what anyone pays. Two values of this convention, established at different dates, cannot be compared without checking their monetary base and gas scope.

Zone of economic impact

the zone where activity, employment or local finances are affected even without physical destruction — because a road closes, a booking is cancelled, or an image of crisis discourages a visit.

What it is not: it is NOT an administrative category nor an official perimeter: it is an analytical construction of this study. No producer delimits it, and the study therefore does not quantify it.

Natural-disaster regimeCatNat (French acronym)

a statutory compensation regime triggered by interministerial order, financed by a compulsory surcharge on property-damage and motor insurance contracts.

What it is not: it does NOT cover forest and vegetation fires, which fall under the fire cover of the contract. The surcharge that finances it is not a fire tax.

Multi-risk home insuranceMRH (French acronym)

the standard insurance contract for a home, which combines several covers including fire.

What it is not: the change in the market for these contracts over a financial year is a market context, never the effect of an event after that year.

Counterfactual

what would have happened if a decision had not been taken. It is the only way to assess a policy — and it is also what is never observed.

What it is not: a counterfactual is not a forecast: it concerns an alternative past, and it is built under displayed assumptions, never under a certainty.

What this study contains

What does this study contain?

This list is generated by the study's contract: a surface removed disappears from it, and a surface added enters it.

Open to all

The answer to the question, in one quotable sentence, and the criteria written in advance that would refute it
3 dated benchmarks of the current state — the only public assessment, the area of the last closed year and the provisional situation report of the current season — each with its producer, its status and its date
1 interactive example chart — area covered, year by year from 2006 to 2025, with the reading of the selected year and the break in definitions flagged where it falls
The situation report of the current season, with its producer, its date and its status — and the reference measure of the vintage, declared not available
The 9 quantities that the foundation does not carry — including insurance compensation, tourism loss and the cost of a hectare depending on the outcome of the fire — with, for each, what would close it
3 common beliefs of the 9 that the study sets against its frozen conclusions, each with a verdict CALCULATED from its evidence and not written by hand — and 5 claims set aside for lack of a foundation to test it
The 11 series used, each with its producer, its role, its layer, its cut-off date and what it does not measure
The 29 conclusions of the study, each with its nature — published fact, derived result, result under a declared parameter or established absence —, its period, its scope, its source, its formula and its limits; the value of 3 of them
The method for reproducing the public assessment, the 12 sentences the study forbids itself to write, and the 18 strings that a guard technically prevents from appearing on the page
The 25 reading aids — initial attack, escaped fire, marginal capacity available, zone of economic impact — each defined on the page and recalled where it is used
The reference to copy, the version, the edition and the data cut-off date — to cite this study in a note

Reserved

The decomposition of the only public assessment of the fires, reproduced from its parameters instead of copied: its three items, the range within which its total moves without any assumption changing, the 6 named empty items that its scope leaves out, its place in the study it comes from, the firefighting extra cost served separately, and the carbon item under the convention in force
The funnel of aerial capacity — 5 stages, including 2 that carry no figure because no source publishes them — with the dated points served as observations and never as a series, the fleet type by type, seasonal rentals and operational availability
What plays out before a fire becomes an event: the measured distribution of fire sizes over 20 vintages, the 5 link by link chain from outbreak to outcome with what each step allows and forbids asserting, the reading of causes according to their denominator, and the economic hinge that France does not publish — what a minute gained, one more drop, a hectare spared is worth
What becomes of the system when several fires start at the same time: a documented episode reconstructed step by step, 3 calculated allocation states under assumptions all displayed, the availability series by type of aircraft set against the intensity of each season, and what the sources say — and do not say — about night intervention
How a fire reaches the economy of a territory beyond the burned zone: the three non-overlapping zones, the figure that shows 5 measured quantities facing 9 quantities that nobody publishes, the comparator of 6 dimensions between a burned establishment and a merely empty one, the six economic quantities that never add up, and the strict separation of the three insurance objects
The double-track chronology — 15 dated events split between what was observed and what was decided, on a single time axis where the gap can be seen without a word accusing it — the price of the same aircraft according to 3 scopes that differ by a factor of two, and 9 aircraft of the industrial ecosystem with their exact status, from in service to design, with no duel
The threshold module: 36 calculated combinations of amount invested, amortisation horizon and discount rate, each rendered in one sentence — what an expenditure would have to avoid each year to break even — then set against the 4 damage layers that France can quantify. Arithmetic, assumptions on screen, with no causal model
The table of scopes: 5 measurement systems set against one another term by term, and 3 divergences explained by the data — with the operation that each forbids
The current season in full: all the national situation reports with their producer and their status, the labelled estimate for the day of the edition and its method, France in fire-hit Europe on a single satellite measure, and the Gironde case of the reference year

Open to all

Cite this study

The reference to copy, with the edition, the version and the data cut-off date.

France ÉcoScan v9.0.14, “The price of fire”, edition 2026.2 — economics of forest fires in France.

Edition:: 2026.2
Product version:: France ÉcoScan v9.0.14
Window covered:: 2006–2025 for the closed series, current season as of 15 September 2026
Data as of:: 15 September 2026

The values of the current season are provisional. They will be re-queried with their producer before any publication, and their date will be updated.