Every wildfire map is drawn for a country. This notebook asks something smaller and sharper: at the address you type in, in the month you pick, in the year you choose — what is the number, and where does it come from? We show the whole formula, we talk about the money, and we end with three things you can do before dinner.
No exact street number needed — the place you type only signs the card, and it never leaves your browser. The five choices below are what actually move the number.
✎ change anything — the dial on the right moves with you.
On days like this, a single ignition can outrun engines. Fires here are decided by embers landing on your roof, not by the flame front.
✎ tap any month — watch the season rewrite your score.
No black box. The score is an additive fire-danger index built the way the Canadian FWI and the EU's EFFIS think: a climate baseline, plus season, fuel, exposure, wind and slope, and drought — multiplied by a year scenario, minus whatever you have already done. It is not an official forecast. It is a ruler for reading official forecasts.
Climate baseline: regions ranked by multi-year burned area and fire-danger days in the Global Wildfire Information System and EFFIS [1][3]. Season curve: the month distribution of each region's normal fire season — China's runs through winter and spring, southeast Australia's from November to February, the Amazon's from August to October [1][2][6].
Fuel and distance: wildland–urban interface research keeps finding the same thing — houses are lost to wind-blown embers, not to the flame front, so distance and roof matter more than most people assume [8][9]. Year scenario: UNEP's 2022 assessment projects extreme wildfires becoming up to 14% more likely by 2030, 30% by 2050 and 50% by 2100 [4].
Limits: this is a teaching model with six variables. A real FWI computes fine-fuel moisture, drought codes, wind speed and direction hourly. Always read your number next to your own national danger rating [3][11][12][14].
Risk is not a number, it is a chain. Press “why?” to dig one layer deeper. When you hit the bottom, the chain hands you back to Step 01.
The most honest translation of fire danger is premiums, evacuations, lost work, hospital beds and rebuilding. Every figure below is a published estimate with its source and year. Numbers get revised; the logic does not.
Preliminary total economic loss estimated for the January 2025 Los Angeles fires (Palisades and Eaton) — a winter fire, at a city edge.AccuWeather preliminary estimate, 2025 [10]
Estimated cost of Indonesia's 2015 peat and forest fires — around 1.9% of national GDP that year, most of it health and productivity lost to haze.World Bank, 2016 [7]
Primary tropical forest lost in 2024 — a record, with nearly half of it fire-driven, the first time fire outranked clearing for agriculture.Global Forest Watch / WRI, 2025 [2]
Burned in Canada's 2023 season, with about 232,000 people evacuated and smoke reaching New York and Europe. Prevention has never looked cheaper.CIFFC / NRCan, 2023 [11]
✎ UNEP's “Fire Ready Formula”:
Governments spend more than half of their fire budgets on response, and often less than 1% on planning and prevention. UNEP recommends two thirds of the money — 66% — go to planning, prevention, preparedness and recovery, leaving 34% for response. That is the cheapest sentence on this page. [4]
Open a card for one thing done well, one trend forming, one gap still open — and a line on why the gap stays open.
DONE WELLThe Australian Fire Danger Rating System, live since September 2022, replaced a 1960s model with vegetation-specific science and reduced the public message to four action levels: Moderate, High, Extreme, Catastrophic. Fewer words, more obedience.
TRENDCultural burning — low-intensity, seasonal, mosaic fire led by Aboriginal practitioners — moving from demonstration project into formal land management.
MISSINGThe AS 3959 building standard only governs new houses. After the 2019–20 Black Summer burned roughly 24 million hectares, retrofitting existing homes is still mostly self-funded. Why? Because retrofits have no ribbon to cut. [14]
DONE WELLAfter the 2017 fires that killed dozens in a single afternoon, Portugal created AGIF to run integrated fire management and made fuel-management strips around buildings compulsory; the EU built the shared rescEU aerial fleet. [3][15]
TRENDGrazing and burning as infrastructure: prescribed fire plus goats and sheep returning as routine fuel management across southern Europe.
MISSINGLand fragmentation. Thousands of tiny parcels belong to absent or deceased owners, so nobody clears them, while eucalyptus monoculture keeps spreading. The law exists; the land registry does not.
DONE WELLFireSmart's Home Ignition Zone model splits a property into 0–1.5 m, 1.5–10 m and 10–30 m rings with a checkable list, and certifies whole neighbourhoods. [11]
TRENDAfter 15 million hectares burned in 2023, cross-province and cross-border crew sharing — and smoke health warnings — became permanent agenda items.
MISSINGIndigenous-led cultural burning has a proven record but still runs on short project grants and case-by-case permits. Why? Because inside a bureaucracy, not burning is always the safer signature.
DONE WELLCalifornia publishes public Fire Hazard Severity Zone maps, NFPA's Firewise USA gives well over a thousand communities one shared method, and “Zone 0” — nothing flammable within five feet of the wall — is moving into regulation. [8][9][12]
TRENDPricing has become the sharpest risk signal: major insurers stopped writing new California home policies from 2023, and the state's insurer of last resort ballooned. [13]
MISSINGWithout large-scale retrofit subsidies, risk is only transferred, never reduced. Why? Because hardening pays back over decades, while premium increases arrive this year.
DONE WELLFengyun and Gaofen satellites plus grid-based ground patrols give near-real-time hotspot detection, and most provinces run a formal fire-prevention season from October to April with tight control on festival fire use. [16]
TRENDThe Forest Chief system, rolled out nationwide in 2021, attaches forest and grassland protection to named local officials — responsibility with a surname is responsibility you can audit.
MISSINGEven-aged pine plantations on steep south-western slopes, plus valley winds, were the backdrop to the 2020 Liangshan tragedy. Building rules and fuel management at the rural–urban edge remain the weak link. Why? Because planting trees has a target; removing fuel does not.
DONE WELLAfter the 2015 catastrophe, a dedicated peatland restoration agency began rewetting drained peat by blocking canals, and concession holders were made responsible for fire inside their own boundaries. [7]
TRENDGroundwater-level monitoring wired to El Niño forecasts, so dry years are pre-positioned for rather than reacted to.
MISSINGRewetting needs maintenance, but funding usually covers construction only. Peat fires smoulder underground for months where satellites cannot see them, and that is the most expensive fire there is.
DONE WELLBrazil's 2024 integrated fire management law rewrote the mission from suppressing fire to managing it, giving prescribed and Indigenous burning a legal footing. [6]
TRENDFire forecasts read on the same map as drought and deforestation alerts, not in a separate window.
MISSINGIn February 2024 more than 130 people died at Viña del Mar, where pine and eucalyptus plantations press directly against informal settlements. The gap is not engines; it is land-use planning and escape routes at the urban edge. Why? Because planning and firefighting sit in two different ministries.
Tick what is already true at your place. Each one subtracts from the index above — not as a consolation prize, but because these are the actions the wildland–urban interface research keeps finding on the houses that survive. Then loop back to Step 01 and recalculate.
✎ note to self: preparedness lowers the odds your house burns, not the odds fire arrives. Fire still arrives.
All public, all checkable. The first four are the fastest way to look up live danger where you actually live.
✎ the index on this page is a teaching model, not an official forecast. When it matters, listen to your local fire and weather service.