Wildfire & extreme weather
Wildfire is the climate consequence most directly felt in parts of Germany, above all Brandenburg. This page makes it visible without the tempting fallacy: a burning forest is first of all location and weather. The climate statement only emerges from the long-term series.
Why a single fire is not yet a climate signal
The vast majority of fires are started by people: sparks, negligence, arson. Climate change does not change the ignition but the conditions: drought, heat, wind, the length of the season and the state of the fuel. A single fire is therefore no proof, the climate signal lies only in the decades-long trend.
Four levels: and only two carry climate statements
Wildfire can be viewed on four levels. They are kept strictly apart because they say different things, and only the last two reveal anything about the climate.
1 · Live situation
no climate statementActive heat detections (EFFIS/Copernicus, from NASA FIRMS). A snapshot, not confirmed fires, no climate statement.
2 · Danger index
no climate statementThe fire-danger index (DWD) rates the current weather (drought, wind). A weather statement, not a fire forecast and not a climate statement.
3 · Historical series
climate statementDanger days since 1961 and burnt area since 1991. Here, and only here, the climate statement emerges. Included below.
4 · Projections
climate statementFuture danger days under scenarios (DWD reference ensemble) with a mandatory uncertainty band. A climate statement about the future.
Live levels 1 and 2 are below as a toggleable map, with data running through a privacy-compliant proxy (EFFIS/Copernicus). Level 3 is included as an observed record from 2010 (BLE). Level 4 sits in the climate projections.
Current situation (levels 1 & 2)
Toggle the live layers on: active heat detections of the last 7 days and the fire-danger index. This is the current situation, not a climate statement. That lies in the historical series (level 3).
Level 3
The observed record
Two records that must be kept apart: the condition (fire weather, since 1961) and the event (burnt area, since 1991). Only here, in the decades-long trend, can a climate statement emerge at all. And comparing the two is the real lesson of this page.
Fire weather: days of high fire danger
This is the most robust climate indicator on this page, because it measures the condition, not the event: days on which the DWD fire-danger index reaches level 4 or 5, meaning high to very high danger. Unlike burnt area it does not depend on whether someone ignites a fire, how the forest is managed, or how carefully fires are reported.
Method: a danger day is an index level of 4 or 5. Fixed reference network of 0 stations with complete records from 0 to 0, because the DWD network grew from around 200 to over 480 stations; averaging over all available stations would present that expansion as a trend. Shown is the mean per station. Source: DWD Climate Data Center, recomputed forest fire danger index (woodland), CC BY 4.0.
What fire weather is made of
The danger index above is a computed value and therefore abstract. These two quantities are part of it and are more tangible: hot days and summer rainfall. Not shown is the annual mean temperature; it says almost nothing about forest fires, because a mild winter does not start them.
The event: burnt area and number of fires
What actually burned, nationwide since 1991. Data by federal state is available from 2010.
Burnt area varies strongly from year to year. Fitting a straight line over the whole record from 1991 shows no increase, but that is due entirely to the exceptional years at the start: 1992 saw 4,908 hectares burn, more than any year since. From the mid-1990s the area does rise, by roughly 29 hectares per year, and the last decade averages 1,351 hectares, almost three times the calm period 1994 to 2017 (465 hectares).
So the record is neither an all-clear nor a simple proof. It is noisy, because burnt area depends not only on weather but on whether someone ignites a fire, how the forest is structured, how quickly it is extinguished and how it is reported. That is why the condition, fire weather, sits above as the more robust indicator.
Who ignites: the causes
Reported causes -10 to 0, Germany, share of all reported fires. This distribution answers only the question of ignition, not of scale.
Caveat: The danger days are the robust part, 65 years on a consistent method. The area record goes back 35 years but is an event measure, confounded by ignition behaviour, forest condition, firefighting capacity and reporting practice. The link between fire weather and area is statistical across many years, not evidence for any single fire.
Data sources: fire-danger days from the German Weather Service (Climate Data Center, recomputed forest fire danger index, woodland, CC BY 4.0). Burnt area and number of fires from the German Federal Office for Agriculture and Food (BLE), Data licence Germany Attribution 2.0; years 1991 to 2009 from the BLE series as published by the German Environment Agency, from 2010 from the BLE open data portal. The overlap of both sources was checked and agrees.
Hazard ≠ disaster
Whether a natural event becomes a disaster depends not only on the event but on vulnerability and preparedness. The same drought hits a resilient mixed forest differently than a pine monoculture. Structurally this is the same distinction as weather/climate and as mitigation/adaptation, and it links the extreme-weather part to adaptation.
To mitigation vs. adaptationKeeping the terms clean (DRR)
To keep the explanations consistent, a short glossary of disaster risk reduction (after UNDRR):
- Hazard
- the potentially damaging event itself, e.g. drought, storm, fire weather.
- Exposure
- what is exposed to the hazard, people, forests, infrastructure in that place.
- Vulnerability
- how susceptible the exposed is, pine monoculture vs. mixed forest, sealed vs. unsealed.
- Risk
- the interplay of hazard × exposure × vulnerability, not the hazard alone.
- Resilience
- the ability to absorb an event and recover, the goal of adaptation.
Why Brandenburg?
Brandenburg is the most affected federal state: large-scale pine monocultures, dry sandy soils and low rainfall. That is less a question of ignition than of conditions and vulnerability, and this is exactly where converting to mixed forest is also fire prevention.
What this page does not count
What stands here are days and hectares: danger days, hot days, burnt area. What comes of it when the conditions discharge is counted in other units. According to the study for the federal economics ministry, floods are the costliest extreme weather events in Germany, at least 70 billion euros since the year 2000. And at least 30,000 additional deaths since 2000 are attributed to extreme weather, 99 per cent of them to heat, that is, to precisely the quantity whose shift is shown in the series above. The same rule applies there as here: no single event is attributed to climate change.
What climate change has cost so farSources: EFFIS/Copernicus (CC BY 4.0), DWD fire-danger index (GeoNutzV) and DWD reference ensemble (CC BY 4.0); NASA FIRMS as a sub-source; EM-DAT (CRED/UCLouvain) only as a cited reference (non-commercial), terms after UNDRR. Live layers and historical series are planned as data layers; this page is editorial for now. No single fire is attributed to climate change (ADR 0116/0097).