Summer 2026 Wildfires in Europe: The Season in Review

Final assessment, EFFIS rapid perimeters through 30 September 2026

Author

Pierre Beaucoral

Published

October 2, 2026

Every summer now brings the same headlines out of the Mediterranean, and the same question underneath them: is this year actually worse, or does it just feel that way? All summer, this page tracked the 2026 season week by week and refused to call it early. The season window closed on 30 September, so it is time for the verdict: how much burned, where, how 2026 ranks against the seasons before it, and what made it different.

Show code 
# -------------------- Packages --------------------
library(sf)
library(dplyr)
library(tidyr)
library(tibble)
library(purrr)
library(lubridate)
library(forcats)
library(ggplot2)
library(scales)
library(patchwork)
library(janitor)
library(rnaturalearth)
library(leaflet)
library(htmlwidgets)

# GDAL caps single GeoJSON features at ~200 MB; the 2020 snapshot has one
# giant multipolygon above that (documented in scripts/fetch_effis.R).
Sys.setenv(OGR_GEOJSON_MAX_OBJ_SIZE = "0")

# -------------------- Shared helper library --------------------
source(file.path("R", "helpers.R"))   # to_num, parse_date_any, lab_si_ha
source(file.path("R", "geo.R"))       # Europe polygons, read_effis, tagging
source(file.path("R", "cache.R"))     # cached()
source(file.path("R", "theme.R"))     # pal_lc, theme_burns()
source(file.path("R", "flags.R"))     # flag_table()
source(file.path("R", "pipeline.R"))  # season assembly (all cached)
source(file.path("R", "plots.R"))     # plot_envelope() (shared with index)
source(file.path("R", "retro.R"))     # season-end figures (ranking, countries)
source(file.path("R", "maps.R"))      # overview proportional-symbol map
source(file.path("scripts", "latest_snapshot.R"))

# -------------------- Season parameters --------------------
YEAR_CURRENT <- 2026L
HIST_YEARS   <- 2017:2025   # archive effectively starts 2016; 2016 kept out of
                            # the band (partial-quality first year), see caveats
SEASON_START_MONTH <- 6L    # season window: 1 Jun -
SEASON_END_MONTH   <- 9L    # 30 Sep

# Calendar heatmap DOES include 2016 (the archive's first, partial-quality
# year) -- unlike HIST_YEARS above, since a full-decade grid is the point of
# that chart; 2016's row is captioned as lower-confidence rather than dropped.
CALENDAR_YEARS <- 2016:2025

snap <- latest_snapshot()
stopifnot(!is.na(snap))
eu <- get_eu()

# -------------------- Core objects (all disk-cached) --------------------
envelope <- build_envelope(
  hist_years = HIST_YEARS, year_current = YEAR_CURRENT,
  snapshot_dir = snap, eu = eu,
  start_month = SEASON_START_MONTH, end_month = SEASON_END_MONTH
)

tagged_2026_full <- get_tagged_full_year(YEAR_CURRENT, snap, eu)

as_of     <- envelope$meta$as_of_date
as_of_lab <- paste(day(as_of), month.name[month(as_of)])   # English, locale-proof

# Season-window (Jun 1 - as-of) headline numbers
season_ha    <- envelope$meta$current_cum_ha
season_n     <- envelope$meta$current_n_fires
pct_vs_med   <- envelope$meta$pct_vs_median
median_ha    <- envelope$meta$median_to_date_ha

# Full-2026-to-date numbers (Jan 1 - as-of), used by maps and rankings
full_ha <- sum(tagged_2026_full$area_ha, na.rm = TRUE)
full_n  <- nrow(tagged_2026_full)

# -------------------- Season-end verdict (all computed) --------------------
# Every past season is read at the SAME day index as 2026's last mapped day,
# so the ranking stays like for like even if the last few days of September
# carry no mapped perimeter yet.
hist_at_asof <- envelope$hist_daily |>
  filter(day_idx == envelope$meta$as_of_day_idx) |>
  select(year, cum_ha)
rank_2026  <- 1L + sum(hist_at_asof$cum_ha > season_ha)
n_seasons  <- nrow(hist_at_asof) + 1L
beaten_by  <- sort(hist_at_asof$year[hist_at_asof$cum_ha > season_ha])
calmest_ha <- min(hist_at_asof$cum_ha)
rank_word  <- c("the worst", "the second-worst", "the third-worst",
                "the fourth-worst")[rank_2026]
if (is.na(rank_word)) rank_word <- sprintf("number %d", rank_2026)

# Month by month: 2026 against the median of the same month in 2017-2025
monthly <- bind_rows(envelope$hist_daily, envelope$current) |>
  mutate(month = month(ba_date)) |>
  group_by(year, month) |>
  summarise(ha = sum(area_ha), .groups = "drop")
monthly_cmp <- monthly |>
  filter(year != YEAR_CURRENT) |>
  group_by(month) |>
  summarise(med_ha = median(ha), .groups = "drop") |>
  left_join(monthly |> filter(year == YEAR_CURRENT) |> select(month, ha),
            by = "month") |>
  mutate(ratio = ha / med_ha)
mo <- function(m, col) monthly_cmp[[col]][monthly_cmp$month == m]

# Whole-year ranking (1 Jan - 30 Sep), split spring vs summer. 2016 is shown
# in the chart but left out of the ranks quoted in prose (lower-confidence year).
ytd <- build_ytd_ranking(2016:YEAR_CURRENT, snap, eu)
ytd_ranked <- ytd |>
  filter(!low_confidence) |>
  group_by(period) |>
  mutate(period_rank = rank(-ha, ties.method = "min")) |>
  ungroup()
ytd_total_rank <- ytd_ranked |> distinct(year, total_ha) |>
  mutate(r = rank(-total_ha, ties.method = "min"))
ytd_rank   <- ytd_total_rank$r[ytd_total_rank$year == YEAR_CURRENT]
ytd_n      <- nrow(ytd_total_rank)
ytd_ahead  <- sort(ytd_total_rank$year[ytd_total_rank$r < ytd_rank])
spring_rank <- ytd_ranked$period_rank[ytd_ranked$year == YEAR_CURRENT &
                                      ytd_ranked$period == "Jan–May"]
spring_ahead <- ytd_ranked |>
  filter(period == "Jan–May", period_rank < spring_rank) |>
  arrange(desc(ha)) |> pull(year)
cap_ytd <- sprintf(paste(
  "Burned area from 1 January to 30 September for each year 2016-%d, split",
  "into January-May (light) and June-September (dark); %d in orange. Years",
  "sorted by total. * 2016 is the first archive year and is lower-confidence.",
  "Perimeters are clipped to the Europe outline without country assignment, so",
  "June-September bars can differ marginally from the country-tagged totals",
  "quoted elsewhere. Source: EFFIS rapid perimeters, snapshot %s; areas from",
  "geometry in EPSG:3035, Europe-clipped."
), YEAR_CURRENT, YEAR_CURRENT, basename(snap))
and_list <- function(x) {
  if (length(x) <= 1L) return(paste(x))
  paste(paste(x[-length(x)], collapse = ", "), "and", x[length(x)])
}

# Each country against its own 2017-2025 summers (rank_cur: 1 = worst of 10)
ctry_hist <- build_country_vs_history(HIST_YEARS, YEAR_CURRENT, snap, eu)
ctry_r <- function(cty) ctry_hist[ctry_hist$name_long == cty, ]
# Prose names: countries that take an article in running English text
with_article <- function(x) {
  ifelse(x %in% c("United Kingdom", "Netherlands", "Czech Republic"), paste("the", x), x)
}
worst_ctry  <- with_article(ctry_hist$name_long[ctry_hist$rank_cur == 1L])
second_ctry <- with_article(ctry_hist$name_long[ctry_hist$rank_cur == 2L])
below_ctry  <- with_article(ctry_hist$name_long[ctry_hist$ratio_to_median < 1])
cap_ctry_hist <- sprintf(paste(
  "Burned area per country, 1 June to 30 September %d, against each country's",
  "own 2017-2025 range (grey bar: minimum to maximum; tick: median), for the",
  "%d countries with the most burned area in %d. Log10 scale; values under",
  "10 ha are drawn at the 10 ha edge. Countries sorted by %d relative to their",
  "own median. Source: EFFIS rapid perimeters, snapshot %s; areas from geometry",
  "in EPSG:3035, Europe-clipped; countries tagged by maximum overlap."
), YEAR_CURRENT, nrow(ctry_hist), YEAR_CURRENT, YEAR_CURRENT, basename(snap))

# Did the burning move? Area-weighted median latitude/longitude per summer
loc_sum <- summarise_fire_location(
  build_fire_locations(c(HIST_YEARS, YEAR_CURRENT), snap, eu)
)
loc_cur  <- loc_sum[loc_sum$year == YEAR_CURRENT, ]
loc_hist <- loc_sum[loc_sum$year != YEAR_CURRENT, ]

ord_word <- function(r) c("first", "second", "third", "fourth", "fifth",
                          "sixth", "seventh", "eighth", "ninth", "tenth")[r]

# The three biggest single days of the summer
top_days <- envelope$current |> slice_max(area_ha, n = 3, with_ties = FALSE)
gap_ha   <- season_ha - median_ha

# Land-cover machinery shared by several chunks
lc_cols <- c("broadlea", "conifer", "mixed", "scleroph", "transit",
             "agriareas", "artifsurf", "othernatlc", "otherlc")
lc_labels <- c(
  broadlea = "Broad-leaved forest", conifer = "Coniferous forest",
  mixed = "Mixed forest", scleroph = "Sclerophyllous veg.",
  transit = "Transitional woodland-shrub", agriareas = "Agricultural areas",
  artifsurf = "Artificial surfaces", othernatlc = "Other natural LC",
  otherlc = "Other LC"
)

# -------------------- Figure captions (referenced via !expr) --------------------
cap_envelope <- sprintf(paste(
  "2026 burned about %s ha between 1 June and %s, %s of the last %d summers",
  "and about %.0f%% above their median. Top: cumulative burned area since",
  "1 June. Thin grey lines are the 2017-2025 seasons; the grey band is the",
  "middle half of those seasons and the dashed line their median, so they show",
  "variation across observed years, not statistical uncertainty. Bottom: burned",
  "area per week in 2026. The shaded last 14 days (faded bars) are still being",
  "mapped: EFFIS keeps filling in recent dates for weeks (area dated 1 June to",
  "2 September grew 5%% between the 3 September and 2 October snapshots).",
  "Source: EFFIS rapid perimeters, snapshot %s; areas from geometry in",
  "EPSG:3035, Europe-clipped."
), comma(round(season_ha, -3)), as_of_lab, rank_word, n_seasons,
   pct_vs_med - 100, basename(snap))
cap_hero <- sprintf(paste(
  "Every one of the ten largest 2026 fires sits in Iberia, south-west France or",
  "the western Balkans, but the season left marks across the whole continent.",
  "Each circle is one fire mapped between 1 January and %s; circle area is",
  "proportional to burned area, and colour shows January-May (blue) versus",
  "summer (orange). Insets zoom on Iberia and south-west France (A) and the",
  "western Balkans (B), with the ten largest fires labelled by province and",
  "country. Source: EFFIS rapid perimeters, snapshot %s; geometry in EPSG:3035."
), as_of_lab, basename(snap))
cap_countries <- sprintf(
  paste(
    "Top-10 countries by Europe-clipped burned area, 1 January – %s 2026.",
    "Left: total mapped burned area. Right: the same total as a share of national",
    "land area (land area computed from the same reference polygons,",
    "mainland-Europe crop). Source: EFFIS rapid perimeters, snapshot %s; areas from",
    "geometry in EPSG:3035; countries tagged by maximum overlap."
  ),
  as_of_lab, basename(snap)
)
cap_reburn <- sprintf(
  paste(
    "2026 perimeters (1 January – %s) split into ground with no overlap with any",
    "mapped 2017–2025 EFFIS perimeter (orange-red) and ground already inside one",
    "(blue). 'No overlap' does not mean never burned: the archive starts in 2016",
    "and misses the smallest fires. Overlap",
    "computed on ~100 m-simplified geometry in EPSG:3035. Source: EFFIS rapid",
    "perimeters, snapshot %s."
  ),
  as_of_lab, basename(snap)
)
cap_calendar <- sprintf(
  paste(
    "Fire belongs to July and August in almost every year, with spring flare-ups",
    "in some. Weekly Europe-clipped burned area by year, %d–%d, in ISO calendar",
    "weeks, darker meaning more area. Fill on a square-root scale: many weeks burn",
    "exactly zero hectares (palest blue-grey cells), which a log scale cannot",
    "display. Grey cells are 2026 weeks after this final snapshot; the dashed line",
    "marks the last mapped week (%s). The near-black October 2017 cell is the",
    "Portuguese and Galician fires of mid-October 2017.",
    "2016 is the archive's first year and its coverage is less complete than",
    "later years. This chart uses burned area, not fire counts, so the",
    "2023-to-2024 detection-system change (which makes counts incomparable",
    "across years) mostly washes out. Source: EFFIS rapid perimeters,",
    "snapshots covering %d–%d; areas from geometry in EPSG:3035, Europe-clipped."
  ),
  min(CALENDAR_YEARS), YEAR_CURRENT, as_of_lab, min(CALENDAR_YEARS), YEAR_CURRENT
)
cap_natura_trend <- sprintf(paste(
  "Inside the EU, the share of burned area falling in Natura 2000 sites has no",
  "clear trend; 2026 sits just above the 2017-2025 median. Top: share of EU-27",
  "burned area inside a Natura 2000 site, 1 June to 30 September each year,",
  "computed as PERCNA2K (EFFIS's field for the share of each perimeter inside",
  "Natura 2000) weighted by Europe-clipped area. Bottom: the same in absolute",
  "hectares. Non-EU countries are excluded because Natura 2000 only exists in",
  "member states (Cyprus and Malta fall outside the Europe outline used here).",
  "Weighting by original rather than clipped perimeter area changes the share",
  "by 0.3 points or less in nine of ten years. Source: EFFIS rapid perimeters,",
  "snapshots covering 2017-%d."
), YEAR_CURRENT)
cap_natura_map <- sprintf(
  paste(
    "2026 perimeters (1 January – %s) colored by PERCNA2K, the share of each",
    "fire's own perimeter that falls inside a Natura 2000 protected site. Grey:",
    "0%% (no protected-area overlap). Bright yellow: perimeters that burned",
    "almost entirely inside protected land. Source: EFFIS rapid perimeters,",
    "snapshot %s."
  ),
  as_of_lab, basename(snap)
)

Two numbers sum up the year, and they cover different time windows. Over the summer season, 1 June to 30 September, satellites mapped about 786,000 hectares of burned land in Europe, across 4,730 separate fire outlines. That is roughly 75 times the area of the city of Paris (about 105 km², or 10,500 hectares), and about 63% more than a typical recent summer (483,000 hectares). But 2026 did not start in June. Counting from 1 January, the total reaches about 967,000 hectares, because late winter and spring had already burned 181,000 hectares before most people start watching for fire season at all. That spring was not unusual, though: it ranks fifth of the last 10 years. The chart right below sticks to the summer window, so that seasons are compared like for like; the maps after it show everything mapped in 2026.

TipThe season in four lines
  • Rank: 2026 is the second-worst summer of the last 10 (2017 to 2026), behind only 2025.
  • Shape: an average spring, a violent July, a long August, then a September that stayed above normal.
  • Where: Spain carried the year, and France, Serbia and Bosnia and Herzegovina had their worst summer of the last ten.
  • How: a few giant fires made most of the total, as they do every year.

About the data, in one line: these are rapid satellite outlines from EFFIS, measured as burned area (never fire counts), and they will still move by a few percent before official figures come out. Hectares are the unit throughout; charts abbreviate thousands of hectares as kha.

Satellites spot large fires, and Europe’s EFFIS service (part of the Copernicus programme) draws the outline of each burned area it can map. I work with those outlines. I measure each fire’s size directly from its shape, using a map projection that keeps areas honest, so hectares are comparable across countries. And I count only the parts that fall inside Europe. The method is the same as in the 2025 season analysis this project grew out of; that post remains the reference for the details.

  • Most mapped fires are small, not big. EFFIS once caught mainly fires of roughly 30 hectares and up, but sharper Sentinel-2 imagery now maps far smaller ones too, so most outlines on this page are in fact well under 30 hectares. The very smallest fires are still missed, so totals stay below official statistics.
  • The outlines are rapid estimates. They are revised as better satellite imagery arrives, even after the season ends. The totals below are final for this page, but they are not official statistics: expect EFFIS’s consolidated figures, published later, to differ by a few percent.
  • Dates are fire dates. Every date on this page is the day EFFIS records a fire as starting. The outline itself usually appears in the database about six days later, and for one fire in ten, three weeks or more later.
  • I compare burned area, not fire counts. The detection system changed in 2023 to 2024 and suddenly saw many more small fires. Counting fires across years is therefore misleading; measuring area is much safer.

How bad was 2026?

Show code 
plot_envelope(envelope)

2026 burned about 786,000 ha between 1 June and 30 September, the second-worst of the last 10 summers and about 63% above their median. Top: cumulative burned area since 1 June. Thin grey lines are the 2017-2025 seasons; the grey band is the middle half of those seasons and the dashed line their median, so they show variation across observed years, not statistical uncertainty. Bottom: burned area per week in 2026. The shaded last 14 days (faded bars) are still being mapped: EFFIS keeps filling in recent dates for weeks (area dated 1 June to 2 September grew 5% between the 3 September and 2 October snapshots). Source: EFFIS rapid perimeters, snapshot 2026-10-02; areas from geometry in EPSG:3035, Europe-clipped.

2026 burned about 786,000 ha between 1 June and 30 September, the second-worst of the last 10 summers and about 63% above their median. Top: cumulative burned area since 1 June. Thin grey lines are the 2017-2025 seasons; the grey band is the middle half of those seasons and the dashed line their median, so they show variation across observed years, not statistical uncertainty. Bottom: burned area per week in 2026. The shaded last 14 days (faded bars) are still being mapped: EFFIS keeps filling in recent dates for weeks (area dated 1 June to 2 September grew 5% between the 3 September and 2 October snapshots). Source: EFFIS rapid perimeters, snapshot 2026-10-02; areas from geometry in EPSG:3035, Europe-clipped.

Read this like a race against the past. Each thin grey line is one of the last nine seasons, day by day; the grey band holds the middle half of them, and the dashed line is the middle of the pack. The orange line is 2026, and it now runs the full course, to 30 September. The bars underneath show how much burned each week, which is where the pace of the season is easiest to see. It finishes about 63% above the median and high in the band: the second-worst of the last 10 summers, with only 2025 ahead of it. For scale, the calmest of those summers burned about 110,000 hectares over the same window, roughly 7 times less.

The more interesting story is the shape of the line, because the verdict this page gave changed a lot over the summer. Here is the season month by month, each against the typical (median) amount for that same month in 2017 to 2025:

Table 1: Burned area by month, 2026 against the median of the same month over 2017–2025 (Europe-clipped, EFFIS rapid perimeters, snapshot 2026-10-02).
Month 2026 burned Typical month 2026 vs typical
June 35,000 ha 19,000 ha 1.9×
July 389,000 ha 113,000 ha 3.4×
August 270,000 ha 166,000 ha 1.6×
September 92,000 ha 34,000 ha 2.7×

July is where 2026 broke away. By 26 July it was running at close to four times a typical season, and this page called it a lonely outlier. August then burned heavily too, but August is also when an ordinary European season does much of its own burning, so the median line climbed steeply and closed part of the gap. That is why the multiple shrank over the summer even though the fire never really let up: the yardstick moved, not the fire. September did not bring the usual slowdown, and the line kept climbing into the finish.

The biggest single day of the summer, 22 July, added about 87,000 hectares, when very large fires broke out in Ávila and in the Gironde within hours of each other. In late July that one day accounted for most of the distance between 2026 and an ordinary year. By the end of September it was only about 29% of that gap: 2026 was not one catastrophic day, it was a summer that kept producing large fires.

Counting the whole year

The summer window keeps seasons comparable, but fire does not read calendars. The chart below counts everything from 1 January to 30 September, with each year split into its spring (lighter) and its summer (darker).

Show code 
plot_ytd_ranking(ytd, YEAR_CURRENT)

Burned area from 1 January to 30 September for each year 2016-2026, split into January-May (light) and June-September (dark); 2026 in orange. Years sorted by total. * 2016 is the first archive year and is lower-confidence. Perimeters are clipped to the Europe outline without country assignment, so June-September bars can differ marginally from the country-tagged totals quoted elsewhere. Source: EFFIS rapid perimeters, snapshot 2026-10-02; areas from geometry in EPSG:3035, Europe-clipped.

Burned area from 1 January to 30 September for each year 2016-2026, split into January-May (light) and June-September (dark); 2026 in orange. Years sorted by total. * 2016 is the first archive year and is lower-confidence. Perimeters are clipped to the Europe outline without country assignment, so June-September bars can differ marginally from the country-tagged totals quoted elsewhere. Source: EFFIS rapid perimeters, snapshot 2026-10-02; areas from geometry in EPSG:3035, Europe-clipped.
Show code 
# Gap to the year just ahead of 2026, as a share of 2026's own total: below
# the ~5% a month of late mapping measured above, the order is not settled.
ytd_tot <- ytd_total_rank |> arrange(desc(total_ha))
next_up <- ytd_tot[ytd_tot$r == ytd_rank - 1L, ]
tie_pct <- if (nrow(next_up)) 100 * (next_up$total_ha - ytd_tot$total_ha[ytd_tot$year == YEAR_CURRENT]) / ytd_tot$total_ha[ytd_tot$year == YEAR_CURRENT] else NA_real_

Counted this way, 2026 is third of the last 10 years (behind 2022 and 2025). But the gap to 2022 is only about 1,700 hectares, 0.2% of 2026’s total, well within the late mapping described below: call it a dead heat, and do not be surprised if 2026 ends up ahead of 2022 once EFFIS publishes consolidated figures. The bars also show how differently seasons are built: 2020 and 2019 did much of their burning before June, while 2026, like 2017 and 2025, was made in summer.

NoteHow final is “final”?

EFFIS maps fires as usable satellite imagery arrives, so the most recent days are the least complete and keep filling in afterwards. This is measurable. The early-September edition of this page, built from the 3 September snapshot, reported about 669,000 hectares burned between 1 June and 2 September. This final snapshot, looking at those very same dates, reports about 702,000: roughly 33,000 hectares, or 5%, showed up a month after the fact. So the last days of September on this page are still a floor, not a finished measurement, and the whole-season total will most likely creep up by a few percent before EFFIS publishes consolidated figures. That is far too small to change the summer ranking above, though it could reorder the whole-year race with 2022.

The chart below splits everything the last month of mapping added, matching each EFFIS fire outline by its identifier between the two snapshots. “Added since the last update” is not the same as “burned since the last update”.

Show code 
# Compares the 3 September snapshot with the final one. Re-rendering this page
# needs DATA/snapshots/2026-09-03 on disk (or its cached clip in DATA/cache).
mapping <- build_mapping_additions(file.path("DATA", "snapshots", "2026-09-03"), snap, eu)
mt <- setNames(mapping$table$ha, c("old", "new_dates", "late", "revised", "removed", "final"))
cap_mapping <- sprintf(paste(
  "The final month of mapping added about %s ha to the summer total, %.0f%% of it",
  "from fires dated after 2 September. Mapped burned area, 1 June to 30 September",
  "2026, from the 3 September snapshot to the final one, split by EFFIS perimeter",
  "identifier: new outlines dated after 2 September, new outlines dated earlier",
  "(late additions), area changes to outlines present in both snapshots, and",
  "outlines withdrawn. Source: EFFIS rapid perimeters, snapshots 2026-09-03 and",
  "%s; areas from geometry in EPSG:3035, Europe-clipped."
), comma(round(mt["final"] - mt["old"], -3)),
   100 * mt["new_dates"] / (mt["final"] - mt["old"]), basename(snap))
Show code 
plot_mapping_waterfall(mapping$table)

The final month of mapping added about 117,000 ha to the summer total, 72% of it from fires dated after 2 September. Mapped burned area, 1 June to 30 September 2026, from the 3 September snapshot to the final one, split by EFFIS perimeter identifier: new outlines dated after 2 September, new outlines dated earlier (late additions), area changes to outlines present in both snapshots, and outlines withdrawn. Source: EFFIS rapid perimeters, snapshots 2026-09-03 and 2026-10-02; areas from geometry in EPSG:3035, Europe-clipped.

The final month of mapping added about 117,000 ha to the summer total, 72% of it from fires dated after 2 September. Mapped burned area, 1 June to 30 September 2026, from the 3 September snapshot to the final one, split by EFFIS perimeter identifier: new outlines dated after 2 September, new outlines dated earlier (late additions), area changes to outlines present in both snapshots, and outlines withdrawn. Source: EFFIS rapid perimeters, snapshots 2026-09-03 and 2026-10-02; areas from geometry in EPSG:3035, Europe-clipped.

Most of the month’s 117,000 extra hectares are genuinely new burning, dated after 2 September. But about 14,000 hectares are whole fires that had already burned by then and were simply mapped late, and another 21,000 came from existing outlines being redrawn larger. That is why the last weeks of any chart on this page are a floor.

Does Europe only burn in summer?

Mostly, and 2026 was no exception. The chart below stacks eleven years, calendar week by calendar week, so a season’s full rhythm is visible at once.

Show code 
weekly_area <- build_weekly_area(
  years = CALENDAR_YEARS, current_year = YEAR_CURRENT,
  current_tagged = tagged_2026_full, snapshot_dir = snap, eu = eu, lc_cols = lc_cols
)

cal_grid <- prepare_calendar_grid(weekly_area, CALENDAR_YEARS, YEAR_CURRENT, as_of)
cal_cutoff_week <- min(lubridate::isoweek(as_of), 52L)

plot_calendar_heatmap(cal_grid, YEAR_CURRENT, cal_cutoff_week)

Fire belongs to July and August in almost every year, with spring flare-ups in some. Weekly Europe-clipped burned area by year, 2016–2026, in ISO calendar weeks, darker meaning more area. Fill on a square-root scale: many weeks burn exactly zero hectares (palest blue-grey cells), which a log scale cannot display. Grey cells are 2026 weeks after this final snapshot; the dashed line marks the last mapped week (30 September). The near-black October 2017 cell is the Portuguese and Galician fires of mid-October 2017. 2016 is the archive’s first year and its coverage is less complete than later years. This chart uses burned area, not fire counts, so the 2023-to-2024 detection-system change (which makes counts incomparable across years) mostly washes out. Source: EFFIS rapid perimeters, snapshots covering 2016–2026; areas from geometry in EPSG:3035, Europe-clipped.

Fire belongs to July and August in almost every year, with spring flare-ups in some. Weekly Europe-clipped burned area by year, 2016–2026, in ISO calendar weeks, darker meaning more area. Fill on a square-root scale: many weeks burn exactly zero hectares (palest blue-grey cells), which a log scale cannot display. Grey cells are 2026 weeks after this final snapshot; the dashed line marks the last mapped week (30 September). The near-black October 2017 cell is the Portuguese and Galician fires of mid-October 2017. 2016 is the archive’s first year and its coverage is less complete than later years. This chart uses burned area, not fire counts, so the 2023-to-2024 detection-system change (which makes counts incomparable across years) mostly washes out. Source: EFFIS rapid perimeters, snapshots covering 2016–2026; areas from geometry in EPSG:3035, Europe-clipped.
Show code 
peak_2026 <- cal_grid |>
  dplyr::filter(year == YEAR_CURRENT, !is.na(area_ha)) |>
  dplyr::slice_max(area_ha, n = 1, with_ties = FALSE)
peak_2026_date <- iso_week_start(YEAR_CURRENT, peak_2026$iso_week)
peak_2026_lab  <- paste(day(peak_2026_date), month.abb[month(peak_2026_date)])

Two patterns jump out. First, most years share the same silhouette: a dark block across July and August and comparatively little else. Second, the exceptions are spring flare-ups, and they are not unique to 2026: several years show bright weeks between February and April, often from fires set to clear pasture and scrub before the growing season. 2026’s spring is one of them, but 2022, 2020, 2025 and 2019 burned more between January and May. Through much of the summer this page called that spring unusual; with the full archive side by side, it sits in the middle of the pack. The single busiest week of 2026 was a summer one, the week of 20 Jul, when about 187 kha burned in that single week alone. The grey band on the right marks the autumn weeks after this final snapshot: the row simply stops there, rather than being padded with zeros.

NoteReading this chart honestly
  • The fill is burned area, not fire counts. The detection system changed in 2023 to 2024, which would make later years look artificially busier if counted by number of fires; area is far less affected.
  • 2016 is the archive’s first year. Its coverage, especially outside summer, is less complete than later years, so read that row with extra caution.
  • The color scale is a square root, not a log. Many weeks genuinely burn zero hectares, and a log scale cannot show a true zero.

Where did it burn?

With the timing settled, the next question is geographic: where, across the continent, did all of this actually happen?

Show code 
plot_overview_map(tagged_2026_full, eu)

Every one of the ten largest 2026 fires sits in Iberia, south-west France or the western Balkans, but the season left marks across the whole continent. Each circle is one fire mapped between 1 January and 30 September; circle area is proportional to burned area, and colour shows January-May (blue) versus summer (orange). Insets zoom on Iberia and south-west France (A) and the western Balkans (B), with the ten largest fires labelled by province and country. Source: EFFIS rapid perimeters, snapshot 2026-10-02; geometry in EPSG:3035.

Every one of the ten largest 2026 fires sits in Iberia, south-west France or the western Balkans, but the season left marks across the whole continent. Each circle is one fire mapped between 1 January and 30 September; circle area is proportional to burned area, and colour shows January-May (blue) versus summer (orange). Insets zoom on Iberia and south-west France (A) and the western Balkans (B), with the ten largest fires labelled by province and country. Source: EFFIS rapid perimeters, snapshot 2026-10-02; geometry in EPSG:3035.

At this scale most fire outlines would be specks, so each fire is drawn as a circle whose area matches the area it burned. The map makes two things obvious. First, the familiar geography: the Iberian Peninsula and the Mediterranean arc dominate, and every one of the ten largest fires sits in Iberia, south-west France or the western Balkans (the two zoomed panels). Second, the blue circles are the spring burning: about 181,000 of the 967,000 hectares were mapped between January and May, much of it in small fires across northern Spain and the Balkans. Keep that in mind when reading the summer chart at the top: it deliberately leaves those months out.

Want to find a particular fire? The interactive explorer shows every 2026 outline on a map you can pan and zoom, with its date, size, country and the vegetation it burned.

Which countries were hit hardest?

Individual fires are dramatic on their own; country totals show where the season landed hardest overall.

Show code 
ctry_2026 <- tagged_2026_full |>
  st_drop_geometry() |>
  group_by(name_long) |>
  summarise(burned_ha = sum(area_ha, na.rm = TRUE), n_fires = n(), .groups = "drop") |>
  left_join(country_land_area(eu), by = "name_long") |>
  mutate(pct_land = 100 * burned_ha / land_area_ha)

top10 <- ctry_2026 |>
  slice_max(burned_ha, n = 10)

flags <- flag_table(top10$name_long, eu$poly)
top10 <- top10 |>
  left_join(flags, by = "name_long") |>
  # factor AFTER the join: joining a factor to a character column silently
  # coerces name_long back to character, which would both scramble the bar
  # order (alphabetical) and break the numeric y positions for the flags
  mutate(name_long = fct_reorder(name_long, burned_ha))

max_x <- max(top10$burned_ha)

# Flag PNGs drawn with base annotation_raster() (no extra image package
# needed): one raster layer per country, placed just right of its bar.
# Discrete y positions are the factor level codes (1..10).
flag_df <- top10 |>
  filter(!is.na(flag_path)) |>
  mutate(y_pos = as.integer(name_long))
flag_layers <- lapply(seq_len(nrow(flag_df)), function(i) {
  x0 <- flag_df$burned_ha[i] + 0.05 * max_x
  annotation_raster(
    png::readPNG(flag_df$flag_path[i]),
    xmin = x0, xmax = x0 + 0.08 * max_x,
    ymin = flag_df$y_pos[i] - 0.28, ymax = flag_df$y_pos[i] + 0.28
  )
})

p_raw <- ggplot(top10, aes(x = burned_ha, y = name_long)) +
  geom_col(fill = "grey90") +
  geom_point(aes(color = col), size = 3) +
  scale_color_identity() +
  flag_layers +
  # direct value labels, placed just past each flag
  geom_text(aes(x = burned_ha + 0.15 * max_x,
                label = paste(comma(round(burned_ha / 1000)), "kha")),
            hjust = 0, size = 3.6, colour = "grey20") +
  scale_x_continuous(labels = lab_si_ha, expand = expansion(mult = c(0, 0.36))) +
  labs(x = "Burned area (ha)", y = NULL) +
  theme_burns(base_size = 12) +
  coord_cartesian(clip = "off")

p_pct <- ggplot(top10, aes(x = pct_land, y = name_long)) +
  geom_segment(aes(x = 0, xend = pct_land, yend = name_long, color = col),
               linewidth = 1.1, show.legend = FALSE) +
  geom_point(aes(color = col), size = 3.2) +
  geom_text(aes(label = sprintf("%.2f%%", pct_land)),
            hjust = -0.35, size = 3.6, colour = "grey20") +
  scale_color_identity() +
  scale_x_continuous(labels = label_percent(scale = 1),
                     expand = expansion(mult = c(0, 0.22))) +
  labs(x = "Share of national land area", y = NULL) +
  theme_burns(base_size = 12) +
  theme(axis.text.y = element_blank())

# plot_spacer() inserts a gap column between the two panels so the left
# panel's rightmost x-axis label (e.g. "150 kha") and its flags do not collide
# with the right panel's leftmost label ("0.0%").
p_raw + plot_spacer() + p_pct +
  plot_layout(widths = c(3, 0.25, 2))

Top-10 countries by Europe-clipped burned area, 1 January – 30 September 2026. Left: total mapped burned area. Right: the same total as a share of national land area (land area computed from the same reference polygons, mainland-Europe crop). Source: EFFIS rapid perimeters, snapshot 2026-10-02; areas from geometry in EPSG:3035; countries tagged by maximum overlap.

Top-10 countries by Europe-clipped burned area, 1 January – 30 September 2026. Left: total mapped burned area. Right: the same total as a share of national land area (land area computed from the same reference polygons, mainland-Europe crop). Source: EFFIS rapid perimeters, snapshot 2026-10-02; areas from geometry in EPSG:3035; countries tagged by maximum overlap.

Spain leads the raw ranking with about 313,000 hectares for the year. But raw hectares flatter big countries. The right panel asks a fairer question: how much of the country’s own land has burned? Small countries can be hit much harder relative to their size, and that is where the order reshuffles.

Raw hectares, even divided by land area, still compare countries with each other. The fairer question for a final verdict is how each country did against its own past: a summer that is ordinary for Spain would be a catastrophe for Belgium. The chart below puts each country’s 2026 summer (orange dot) on top of the range of its previous nine summers (grey bar), with the typical summer marked by a tick.

Show code 
plot_country_vs_history(ctry_hist, YEAR_CURRENT)

Burned area per country, 1 June to 30 September 2026, against each country’s own 2017-2025 range (grey bar: minimum to maximum; tick: median), for the 14 countries with the most burned area in 2026. Log10 scale; values under 10 ha are drawn at the 10 ha edge. Countries sorted by 2026 relative to their own median. Source: EFFIS rapid perimeters, snapshot 2026-10-02; areas from geometry in EPSG:3035, Europe-clipped; countries tagged by maximum overlap.

Burned area per country, 1 June to 30 September 2026, against each country’s own 2017-2025 range (grey bar: minimum to maximum; tick: median), for the 14 countries with the most burned area in 2026. Log10 scale; values under 10 ha are drawn at the 10 ha edge. Countries sorted by 2026 relative to their own median. Source: EFFIS rapid perimeters, snapshot 2026-10-02; areas from geometry in EPSG:3035, Europe-clipped; countries tagged by maximum overlap.

The horizontal axis is logarithmic again, so each gridline is ten times the one before. Measured against itself, France, Serbia and Bosnia and Herzegovina had their worst summer of the last ten, and the United Kingdom, Romania and Croatia their second-worst. The dots sitting at the right end of their bars are the real story of 2026: the season was severe across a wide band of the continent, from the Atlantic to the Balkans, not only in the usual Mediterranean hotspots. At the other end, Bulgaria finished below its own typical summer.

Did the fire move, then? Not measurably. Half of 2026’s summer burned area lay south of 41.5°N, inside the range of the previous nine summers (39.8 to 41.7°N), and the same holds east to west. Severe years differ in where in the Mediterranean they burn, not yet in a visible march north: one season is far too few to see a slow shift like that.

France and Spain, each against its own past

Two countries carried most of 2026, and it is worth seeing each on its own terms, against its own recent history rather than Europe’s. The two charts below repeat the race-against-the-past idea from the top of the page, once for France and once for Spain: the thin grey lines and band are that country’s own 2017 to 2025 seasons, and the orange line is 2026.

Show code 
fs_df <- tagged_2026_full |> st_drop_geometry()

# Gironde (the departement around Bordeaux)
gir     <- fs_df |> dplyr::filter(!is.na(province), province == "Gironde")
gir_big <- gir |> dplyr::slice_max(area_ha, n = 1, with_ties = FALSE)

# Provinces ringing Madrid (Sierra de Gredos / Guadarrama). "vila$" matches
# "Avila" without depending on the accent byte encoding of the province field.
madrid_ring <- fs_df |> dplyr::filter(
  !is.na(province), grepl("vila$|Guadalajara|^Madrid$|Segovia", province)
)
madrid_ring_ha <- sum(madrid_ring$area_ha, na.rm = TRUE)
navaluenga <- fs_df |> dplyr::filter(grepl("Navaluenga", dplyr::coalesce(commune, ""))) |>
  dplyr::slice_max(area_ha, n = 1, with_ties = FALSE)

# Largest fire inside the Comunidad de Madrid itself. Selected by PROVINCE, not
# by commune name: EFFIS re-attributes communes between snapshots as perimeters
# are refined (the 2026-07-23 fire was mapped under San Martin de Valdeiglesias
# in the 26 July snapshot and under Navas del Rey in the 20 August one), which
# silently collapsed a commune-name filter to a 35 ha fire and rounded the
# prose figure to "0 hectares". Province is the stable unit here.
madrid_big <- fs_df |> dplyr::filter(!is.na(province), province == "Madrid") |>
  dplyr::slice_max(area_ha, n = 1, with_ties = FALSE)

# Season-wide largest fire, so the prose never hardcodes which fire leads.
biggest_fire <- fs_df |> dplyr::slice_max(area_ha, n = 1, with_ties = FALSE)

# English date labels, locale-proof (see as_of_lab in the setup chunk)
date_lab <- function(d) paste(lubridate::day(d), month.name[lubridate::month(d)])

# Where 2026 finishes within a country envelope's own 2017-2025 history, read
# at the same day index (same logic as the Europe-wide verdict in setup).
env_rank <- function(env) {
  h <- env$hist_daily |> dplyr::filter(day_idx == env$meta$as_of_day_idx)
  ahead <- sort(h$year[h$cum_ha > env$meta$current_cum_ha])
  list(rank = 1L + length(ahead), ahead = ahead,
       ha = env$meta$current_cum_ha, ratio = env$meta$pct_vs_median / 100)
}
ahead_lab <- function(r) {
  if (r$rank == 1L) "the worst summer in its 2017 to 2026 record"
  else sprintf("behind only %s in its 2017 to 2026 record", paste(r$ahead, collapse = " and "))
}

cap_env_fr <- sprintf(paste(
  "Cumulative burned area in France since 1 June: the full 2026 season against",
  "France's own 2017-2025 seasons (thin grey lines; band: middle half of those",
  "seasons; dashed line: their median), with 2026's weekly burned area",
  "underneath. The shaded last 14 days are still being mapped. The 2026 line",
  "stops at its last mapped perimeter (%s). Source: EFFIS rapid perimeters,",
  "Europe-clipped, countries tagged by maximum overlap."
), as_of_lab)
cap_env_es <- sprintf(paste(
  "Cumulative burned area in Spain since 1 June: the full 2026 season against",
  "Spain's own 2017-2025 seasons (thin grey lines; band: middle half of those",
  "seasons; dashed line: their median), with 2026's weekly burned area",
  "underneath. The shaded last 14 days are still being mapped. The 2026 line",
  "stops at its last mapped perimeter (%s). Source: EFFIS rapid perimeters,",
  "Europe-clipped, countries tagged by maximum overlap."
), as_of_lab)
Show code 
env_fr <- build_envelope(
  HIST_YEARS, YEAR_CURRENT, snap, eu,
  SEASON_START_MONTH, SEASON_END_MONTH, as_of_date = as_of, country = "France"
)
plot_envelope(env_fr, note = FALSE)

Cumulative burned area in France since 1 June: the full 2026 season against France’s own 2017-2025 seasons (thin grey lines; band: middle half of those seasons; dashed line: their median), with 2026’s weekly burned area underneath. The shaded last 14 days are still being mapped. The 2026 line stops at its last mapped perimeter (30 September). Source: EFFIS rapid perimeters, Europe-clipped, countries tagged by maximum overlap.

Cumulative burned area in France since 1 June: the full 2026 season against France’s own 2017-2025 seasons (thin grey lines; band: middle half of those seasons; dashed line: their median), with 2026’s weekly burned area underneath. The shaded last 14 days are still being mapped. The 2026 line stops at its last mapped perimeter (30 September). Source: EFFIS rapid perimeters, Europe-clipped, countries tagged by maximum overlap.
Show code 
fr_r <- env_rank(env_fr)

France’s summer finished at about 73,000 hectares, 16.0 times its own typical summer and the worst summer in its 2017 to 2026 record. A single fire is about half of that, and it is close to home for me. As a French reader I have followed this Atlantic coast for years, and for the first seven weeks of the season the Gironde, the département around Bordeaux, had not recorded a single mapped fire. Then, on 22 July, the Médoc pine forest at Le Porge burned: about 37,000 hectares in one fire, more than the whole of Gironde’s catastrophic 2022 season put together. A burn that size on that coast is not something I can look past: it is the fire I will remember 2026 for.

Show code 
env_es <- build_envelope(
  HIST_YEARS, YEAR_CURRENT, snap, eu,
  SEASON_START_MONTH, SEASON_END_MONTH, as_of_date = as_of, country = "Spain"
)
plot_envelope(env_es, note = FALSE)

Cumulative burned area in Spain since 1 June: the full 2026 season against Spain’s own 2017-2025 seasons (thin grey lines; band: middle half of those seasons; dashed line: their median), with 2026’s weekly burned area underneath. The shaded last 14 days are still being mapped. The 2026 line stops at its last mapped perimeter (30 September). Source: EFFIS rapid perimeters, Europe-clipped, countries tagged by maximum overlap.

Cumulative burned area in Spain since 1 June: the full 2026 season against Spain’s own 2017-2025 seasons (thin grey lines; band: middle half of those seasons; dashed line: their median), with 2026’s weekly burned area underneath. The shaded last 14 days are still being mapped. The 2026 line stops at its last mapped perimeter (30 September). Source: EFFIS rapid perimeters, Europe-clipped, countries tagged by maximum overlap.
Show code 
es_r <- env_rank(env_es)

Spain is the heavier story. Its summer finished at about 274,000 hectares, 5.5 times its own typical summer and behind only 2022 and 2025 in its 2017 to 2026 record, and it came in two separate surges rather than one. The first sits in the mountains ringing Madrid: the Sierra de Gredos and the ranges north of the capital. About 42,000 hectares burned at Navaluenga in Ávila on 22 July, and Navas del Rey, inside the Comunidad de Madrid itself, added about 20,000 hectares the next day. Taken together, the provinces around Madrid account for roughly 107,000 hectares this year.

The second surge came in August, and it moved the record. The largest single fire of the 2026 season was not in the Madrid ring at all: about 44,000 hectares at Niebla in Huelva, mapped on 6 August, in the Atlantic southwest of Andalusia rather than the central mountains. Aragón followed four days later, with another large fire in the Huesca pre-Pyrenees. Spain’s 2026 burning did not stay in one place, and no single region explains its total.

How big is a fire, really?

Here is the surprise hiding inside every map above: most of these fires are small, and yet a handful of them make the season. The curve below ranks the summer’s fires from largest to smallest and asks how much of the burned area the biggest ones hold.

Show code 
conc <- build_concentration(YEAR_CURRENT, snap, eu)
top1  <- top_share(conc, 0.01)
top10 <- top_share(conc, 0.10)
n_top1 <- ceiling(0.01 * conc$n_fires[1])
cap_conc <- sprintf(paste(
  "A few big fires carried the 2026 season: the largest 1%% of fires hold %.0f%%",
  "of burned area and the largest 10%% hold %.0f%%. Cumulative share of burned",
  "area held by the largest fires, 1 June to 30 September 2026, fires ranked",
  "largest first on a log axis. Not compared across years: the number of small",
  "fires mapped jumped with the 2023-2024 detection change, which alone would",
  "make recent years look more concentrated. Source: EFFIS rapid perimeters,",
  "snapshot %s; areas from geometry in EPSG:3035, Europe-clipped."
), 100 * top1, 100 * top10, basename(snap))
Show code 
plot_concentration(conc, YEAR_CURRENT)

A few big fires carried the 2026 season: the largest 1% of fires hold 55% of burned area and the largest 10% hold 85%. Cumulative share of burned area held by the largest fires, 1 June to 30 September 2026, fires ranked largest first on a log axis. Not compared across years: the number of small fires mapped jumped with the 2023-2024 detection change, which alone would make recent years look more concentrated. Source: EFFIS rapid perimeters, snapshot 2026-10-02; areas from geometry in EPSG:3035, Europe-clipped.

A few big fires carried the 2026 season: the largest 1% of fires hold 55% of burned area and the largest 10% hold 85%. Cumulative share of burned area held by the largest fires, 1 June to 30 September 2026, fires ranked largest first on a log axis. Not compared across years: the number of small fires mapped jumped with the 2023-2024 detection change, which alone would make recent years look more concentrated. Source: EFFIS rapid perimeters, snapshot 2026-10-02; areas from geometry in EPSG:3035, Europe-clipped.

Of 4,730 summer fires, the largest 48 (1% of them) hold 55% of all the area that burned, and the largest 10% hold 85%. The rest, thousands of fires, share what is left. The next chart shows the same thing from the other side, by size.

The chart below splits the 2026 fires two ways on the same size axis. That axis is logarithmic, so each step to the right multiplies the size tenfold rather than adding a fixed amount, which is the only way to show fires of two hectares and fires of thirty thousand on one line.

Show code 
size_df <- tagged_2026_full |>
  st_drop_geometry() |>
  filter(is.finite(area_ha), area_ha > 0)

share_fires_small <- mean(size_df$area_ha < 30)
share_area_small  <- sum(size_df$area_ha[size_df$area_ha < 30]) / sum(size_df$area_ha)
median_fire_ha    <- median(size_df$area_ha)

cap_fire_sizes <- sprintf(
  paste(
    "Every 2026 fire (1 January – %s), split by size on a shared logarithmic",
    "axis. Top: how many fires fall in each size band. Bottom: how many hectares",
    "those same fires burned. The dashed line marks 30 hectares, the rough floor",
    "of the older MODIS-era mapping. Source: EFFIS rapid perimeters, snapshot %s;",
    "areas from Europe-clipped geometry in EPSG:3035."
  ),
  as_of_lab, basename(snap)
)
Show code 
plot_fire_sizes(tagged_2026_full)

Every 2026 fire (1 January – 30 September), split by size on a shared logarithmic axis. Top: how many fires fall in each size band. Bottom: how many hectares those same fires burned. The dashed line marks 30 hectares, the rough floor of the older MODIS-era mapping. Source: EFFIS rapid perimeters, snapshot 2026-10-02; areas from Europe-clipped geometry in EPSG:3035.

Every 2026 fire (1 January – 30 September), split by size on a shared logarithmic axis. Top: how many fires fall in each size band. Bottom: how many hectares those same fires burned. The dashed line marks 30 hectares, the rough floor of the older MODIS-era mapping. Source: EFFIS rapid perimeters, snapshot 2026-10-02; areas from Europe-clipped geometry in EPSG:3035.

The two panels lean opposite ways. The top one counts fires, and its weight sits on the left: the large majority are well under 30 hectares, the size EFFIS could barely see a few years ago. The bottom panel counts hectares instead, and it tips to the right, where a handful of big fires hold nearly all the burned area. Both things are true at once. By number, small fires dominate; by area, a few giants do. About 75% of 2026’s mapped fires are under 30 hectares, yet they account for only about 7% of the burned area, and the typical fire is just 9 hectares. That gap is exactly why this page measures burned area instead of counting fires: one large fire says more about the season than a thousand tiny ones, and it is those giants that the next section puts under the microscope.

What does a giant fire actually look like?

Numbers like “forty thousand hectares” are hard to picture on their own. Below are the ten largest fires of 2026, each drawn at the exact same scale, so size differences are visible rather than just stated. A single grey circle, equal in area to the city of Paris, sits alongside them for reference.

Show code 
# Built here (not in the shared setup chunk) because it only depends on
# tagged_2026_full, which the setup chunk already produced; the caption
# below needs gallery$half_side, so it is composed in this data chunk and
# only REFERENCED (via !expr) by the figure chunk that follows -- fig-cap
# options resolve before a chunk's own body runs, so a caption cannot depend
# on a value the same chunk is about to compute.
gallery <- build_gallery_scars(
  tagged_2026_full, n = 10L, lc_cols = lc_cols, lc_labels = lc_labels
)

cap_gallery <- sprintf(
  paste(
    "The ten largest 2026 fires by Europe-clipped area, each drawn at the same",
    "metric scale (a single shared window, about %.0f km across) so their true",
    "relative size is directly comparable; fill shows each fire's dominant",
    "land-cover class. The grey circle is not a fire: it is a circle of equal",
    "area to Paris intra-muros (about 105 km2, 10,500 ha), drawn for scale",
    "rather than traced from an official city boundary. Place names are EFFIS",
    "communes, or the province or country when EFFIS gives no commune; rapid",
    "perimeters are provisional and",
    "subject to revision. All panels share one scale; the black bar is 10 km.",
    "Source: EFFIS rapid perimeters, snapshot %s."
  ),
  gallery$half_side * 2 / 1000, basename(snap)
)
Show code 
plot_gallery_scars(gallery)

The ten largest 2026 fires by Europe-clipped area, each drawn at the same metric scale (a single shared window, about 53 km across) so their true relative size is directly comparable; fill shows each fire’s dominant land-cover class. The grey circle is not a fire: it is a circle of equal area to Paris intra-muros (about 105 km2, 10,500 ha), drawn for scale rather than traced from an official city boundary. Place names are EFFIS communes, or the province or country when EFFIS gives no commune; rapid perimeters are provisional and subject to revision. All panels share one scale; the black bar is 10 km. Source: EFFIS rapid perimeters, snapshot 2026-10-02.

The ten largest 2026 fires by Europe-clipped area, each drawn at the same metric scale (a single shared window, about 53 km across) so their true relative size is directly comparable; fill shows each fire’s dominant land-cover class. The grey circle is not a fire: it is a circle of equal area to Paris intra-muros (about 105 km2, 10,500 ha), drawn for scale rather than traced from an official city boundary. Place names are EFFIS communes, or the province or country when EFFIS gives no commune; rapid perimeters are provisional and subject to revision. All panels share one scale; the black bar is 10 km. Source: EFFIS rapid perimeters, snapshot 2026-10-02.

Fires like these are not typical of the season, but as the curve above showed, they do most of the damage. The biggest one on the sheet, in Niebla, Spain, is bigger than the Paris-sized circle next to it.

A grey circle is only a stand-in for Paris. On the interactive explorer you can drop each of these ten fires, with its true ragged outline, onto the real map of Paris, one at a time.

Is fire returning to old scars?

Some land burns again and again. To see how much of 2026 falls on ground that already burned recently, I merged all the fire outlines from 2017 to 2025 into one big historical footprint, then checked how much of this year’s burned area lands inside it.

Show code 
footprint <- get_historical_footprint(HIST_YEARS, snap, eu, tol_m = 100)

reburn <- cached(
  sprintf("reburn_2026_asof%s_tol100", format(as_of, "%Y%m%d")),
  compute_reburn(tagged_2026_full, footprint),
  version = 1
)

reburn_layers <- rbind(
  st_sf(kind = "No overlap with mapped 2017–2025 scars", geometry = reburn$new_geom),
  st_sf(kind = "Overlaps a mapped 2017–2025 scar", geometry = reburn$reburn_geom)
)

ggplot() +
  geom_sf(data = eu$poly, fill = "grey95", color = "grey70", linewidth = 0.15) +
  geom_sf(data = reburn_layers, aes(fill = kind, color = kind),
          linewidth = 0.1, alpha = 0.65) +
  scale_fill_manual(values = c(
    "No overlap with mapped 2017–2025 scars" = "#D64A05",
    "Overlaps a mapped 2017–2025 scar" = "#1F78B4"
  ), name = NULL) +
  scale_color_manual(values = c(
    "No overlap with mapped 2017–2025 scars" = "#D64A05",
    "Overlaps a mapped 2017–2025 scar" = "#1F78B4"
  ), name = NULL) +
  theme_burns(base_size = 11, map = TRUE) +
  theme(legend.position = "bottom")

2026 perimeters (1 January – 30 September) split into ground with no overlap with any mapped 2017–2025 EFFIS perimeter (orange-red) and ground already inside one (blue). ‘No overlap’ does not mean never burned: the archive starts in 2016 and misses the smallest fires. Overlap computed on ~100 m-simplified geometry in EPSG:3035. Source: EFFIS rapid perimeters, snapshot 2026-10-02.

2026 perimeters (1 January – 30 September) split into ground with no overlap with any mapped 2017–2025 EFFIS perimeter (orange-red) and ground already inside one (blue). ‘No overlap’ does not mean never burned: the archive starts in 2016 and misses the smallest fires. Overlap computed on ~100 m-simplified geometry in EPSG:3035. Source: EFFIS rapid perimeters, snapshot 2026-10-02.

The map locates the overlap; the bars below compare it country by country.

Show code 
reburn_ctry <- build_reburn_by_country(tagged_2026_full, footprint, snap, tag = "full2026")
rb <- function(cty) reburn_ctry$reburn_share[reburn_ctry$name_long == cty]
rb_top <- reburn_ctry |> filter(total_ha >= 20000) |> slice_max(reburn_share, n = 1)
cap_reburn_ctry <- sprintf(paste(
  "The share of each country's 2026 burned area that overlaps mapped 2017-2025",
  "scars ranges from about %.0f%% in Spain to %.0f%% in %s. Countries with the",
  "most 2026 burned area (1 January to %s); labels give overlapping hectares of",
  "the country total. Overlap measured against a footprint simplified at",
  "100 m; countries assigned by maximum overlap. 'No overlap' means no overlap",
  "with mapped scars, not that the land never burned. Source: EFFIS rapid",
  "perimeters, snapshot %s; areas from geometry in EPSG:3035, Europe-clipped."
), 100 * rb("Spain"), 100 * rb_top$reburn_share, rb_top$name_long, as_of_lab,
   basename(snap))
Show code 
plot_reburn_by_country(reburn_ctry)

The share of each country’s 2026 burned area that overlaps mapped 2017-2025 scars ranges from about 7% in Spain to 49% in Montenegro. Countries with the most 2026 burned area (1 January to 30 September); labels give overlapping hectares of the country total. Overlap measured against a footprint simplified at 100 m; countries assigned by maximum overlap. ‘No overlap’ means no overlap with mapped scars, not that the land never burned. Source: EFFIS rapid perimeters, snapshot 2026-10-02; areas from geometry in EPSG:3035, Europe-clipped.

The share of each country’s 2026 burned area that overlaps mapped 2017-2025 scars ranges from about 7% in Spain to 49% in Montenegro. Countries with the most 2026 burned area (1 January to 30 September); labels give overlapping hectares of the country total. Overlap measured against a footprint simplified at 100 m; countries assigned by maximum overlap. ‘No overlap’ means no overlap with mapped scars, not that the land never burned. Source: EFFIS rapid perimeters, snapshot 2026-10-02; areas from geometry in EPSG:3035, Europe-clipped.

Spain, by far the largest burner, landed mostly outside any mapped scar: only about 7% of its burned area overlaps one, against 24% in France, 29% in Portugal and 31% in Italy. Spain’s August fires in Huelva and Aragón reinforced that contrast rather than softening it. The highest shares are in the Balkans, where Montenegro returned to old scars for about 49% of its burned area.

Show code 
aude_rows <- tagged_2026_full |>
  st_drop_geometry() |>
  filter(!is.na(province), grepl("Aude", province, fixed = TRUE))
aude_n  <- nrow(aude_rows)
aude_ha <- sum(aude_rows$area_ha, na.rm = TRUE)

fr_rows <- tagged_2026_full |> st_drop_geometry() |> filter(name_long == "France")
fr_n  <- nrow(fr_rows)
fr_ha <- sum(fr_rows$area_ha, na.rm = TRUE)

The headline: about 22% of 2026’s burned area (roughly 216,000 of 967,000 hectares) sits on land that already burned at least once since 2017. That is not automatically bad news. Mediterranean shrubland and many pine forests are fire-adapted: they regrow quickly after a burn, and that fresh growth is often exactly the kind of fine, dry fuel that carries a fire well again a few years later. Repeated burning on a short cycle has long been part of how these landscapes work. What the share does not tell us is whether any single re-burn was more or less severe than the fire before it, or whether the gap between fires is shrinking: it is a description of overlap, not a verdict on land condition. Treat the exact figure as an estimate too. Both layers are rapid mapping products, and I simplified the shapes to keep the computation tractable.

I check on Aude every season now, since it is where this whole project started: the Aude département in France does show 2026 activity: 4 mapped fires totalling about 1,000 hectares. France as a whole ends the year at about 108,900 hectares across 1,712 mapped fires.

What is actually burning?

Beyond where fire keeps coming back, there is a second question worth asking: did this season eat into different vegetation than usual? To compare fairly, I use the same window every year, 1 June to 30 September, and I compare 2026 with a typical year rather than with an average that big years would dominate.

Show code 
# Kept: the Natura trend below still reads cutoff_md.
cutoff_md <- format(as_of, "%m-%d")

lc_cmp <- build_landcover_compare(HIST_YEARS, YEAR_CURRENT, snap, eu, lc_cols, lc_labels)
lc_up   <- lc_cmp |> slice_max(diff_pp, n = 2, with_ties = FALSE)
lc_down <- lc_cmp |> slice_min(diff_pp, n = 1, with_ties = FALSE)
# classes where 2026 falls outside the middle half of the nine previous summers
lc_outside <- lc_cmp |> filter(share_cur > share_q75 | share_cur < share_q25)

cap_landcover <- sprintf(paste(
  "%s and %s made up a larger share of 2026's burned area than in a typical",
  "season. Share of burned area by land-cover class, 1 June to 30 September:",
  "2026 (orange dot) against the median and interquartile range of each year's",
  "own shares in 2017-2025 (grey dot and whisker) and the pooled, area-weighted",
  "2017-2025 composition (hollow diamond), which large seasons dominate.",
  "Right-hand labels give 2026 minus the typical-year median in percentage",
  "points. Shares weight each perimeter's land-cover percentages by its",
  "Europe-clipped area. Source: EFFIS rapid perimeters, snapshot %s; areas from",
  "geometry in EPSG:3035, Europe-clipped."
), lc_up$label[1], tolower(lc_up$label[2]), basename(snap))
Show code 
plot_landcover_compare(lc_cmp, YEAR_CURRENT)

Coniferous forest and mixed forest made up a larger share of 2026’s burned area than in a typical season. Share of burned area by land-cover class, 1 June to 30 September: 2026 (orange dot) against the median and interquartile range of each year’s own shares in 2017-2025 (grey dot and whisker) and the pooled, area-weighted 2017-2025 composition (hollow diamond), which large seasons dominate. Right-hand labels give 2026 minus the typical-year median in percentage points. Shares weight each perimeter’s land-cover percentages by its Europe-clipped area. Source: EFFIS rapid perimeters, snapshot 2026-10-02; areas from geometry in EPSG:3035, Europe-clipped.

Coniferous forest and mixed forest made up a larger share of 2026’s burned area than in a typical season. Share of burned area by land-cover class, 1 June to 30 September: 2026 (orange dot) against the median and interquartile range of each year’s own shares in 2017-2025 (grey dot and whisker) and the pooled, area-weighted 2017-2025 composition (hollow diamond), which large seasons dominate. Right-hand labels give 2026 minus the typical-year median in percentage points. Shares weight each perimeter’s land-cover percentages by its Europe-clipped area. Source: EFFIS rapid perimeters, snapshot 2026-10-02; areas from geometry in EPSG:3035, Europe-clipped.

Each row is one kind of land. The grey dot is a typical summer and its grey whisker spans the middle half of the last nine; the orange dot is 2026. Read across a row: when the orange dot sits off the whisker, 2026 burned an unusual amount of that kind of land. 2026 leaned towards forest: coniferous forest made up 12% of the burned area against 7% in a typical year, and mixed forest 6% against 2%, while agricultural areas fell to 18% from 22%. That fits the season’s geography: the Médoc pines, the Ávila and Madrid forests and the Balkan woodland. The hollow diamonds show why the yardstick matters: an average pooled across all nine years is pulled towards the biggest seasons, and for some classes it tells a different story from a typical year. Read the dots as what burned, not as what burns easily: they describe the landscape the fires happened to cross.

  • Broad-leaved forest: deciduous trees (oaks, beeches).
  • Coniferous forest: needle-leaf trees (pines, spruces, firs), including plantations.
  • Mixed forest: broad-leaved and coniferous stands together.
  • Sclerophyllous vegetation: fire-adapted Mediterranean shrubland (maquis, garrigue); evergreen, waxy leaves.
  • Transitional woodland-shrub: open, regenerating or degraded woodland, shrub mosaics, young stands.
  • Agricultural areas: arable land, permanent crops, pastures.
  • Artificial surfaces: urban and built-up areas, infrastructure.
  • Other natural LC: natural cover not listed above (heath, sparse vegetation, dunes, rocky ground).
  • Other LC: residual or unknown classes.

How much of this is happening in protected nature?

One more lens on what burned: how much of it sits inside land specifically protected for nature?

NoteWhat is Natura 2000?

Natura 2000 is the European Union’s network of protected natural sites, covering habitats and species the EU has agreed are worth safeguarding. It is not a fire ban. Many Natura 2000 landscapes, especially Mediterranean scrub and some forest types, are fire-adapted ecosystems where occasional burning is part of their natural cycle. A fire inside a protected site is not automatically a disaster, but tracking how much burning happens there is still useful context.

EFFIS tags every perimeter with the share of its own area that falls inside a Natura 2000 site. The chart below turns that into a season-by-season series, using the same 1 June to 30 September window every year so the comparison is fair. It counts EU countries only: Natura 2000 does not exist outside the EU, so including the Balkans or the United Kingdom would make the share drop every time fire moves there, whatever happens inside protected land.

Show code 
natura_trend <- build_natura_trend(
  hist_years = HIST_YEARS, year_current = YEAR_CURRENT,
  current_tagged = tagged_2026_full, snapshot_dir = snap, eu = eu, cutoff_md = cutoff_md
)

plot_natura_trend(natura_trend, YEAR_CURRENT)

Inside the EU, the share of burned area falling in Natura 2000 sites has no clear trend; 2026 sits just above the 2017-2025 median. Top: share of EU-27 burned area inside a Natura 2000 site, 1 June to 30 September each year, computed as PERCNA2K (EFFIS’s field for the share of each perimeter inside Natura 2000) weighted by Europe-clipped area. Bottom: the same in absolute hectares. Non-EU countries are excluded because Natura 2000 only exists in member states (Cyprus and Malta fall outside the Europe outline used here). Weighting by original rather than clipped perimeter area changes the share by 0.3 points or less in nine of ten years. Source: EFFIS rapid perimeters, snapshots covering 2017-2026.

Inside the EU, the share of burned area falling in Natura 2000 sites has no clear trend; 2026 sits just above the 2017-2025 median. Top: share of EU-27 burned area inside a Natura 2000 site, 1 June to 30 September each year, computed as PERCNA2K (EFFIS’s field for the share of each perimeter inside Natura 2000) weighted by Europe-clipped area. Bottom: the same in absolute hectares. Non-EU countries are excluded because Natura 2000 only exists in member states (Cyprus and Malta fall outside the Europe outline used here). Weighting by original rather than clipped perimeter area changes the share by 0.3 points or less in nine of ten years. Source: EFFIS rapid perimeters, snapshots covering 2017-2026.
Show code 
natura_2026       <- natura_trend |> dplyr::filter(year == YEAR_CURRENT)
natura_hist_median <- stats::median(natura_trend$share[natura_trend$year != YEAR_CURRENT], na.rm = TRUE)

Inside the EU, about 36% of 2026’s summer burned area sits inside a Natura 2000 site, roughly 213,000 hectares, against a 33% median for 2017-2025. The share is ordinary; the hectares are not, because so much more burned overall. Only 2025 put more protected land through fire in this record. The map below shows exactly which 2026 scars carry a high protected-area share; grey outlines mark fires with none.

Show code 
plot_natura_map(tagged_2026_full, eu)

2026 perimeters (1 January – 30 September) colored by PERCNA2K, the share of each fire’s own perimeter that falls inside a Natura 2000 protected site. Grey: 0% (no protected-area overlap). Bright yellow: perimeters that burned almost entirely inside protected land. Source: EFFIS rapid perimeters, snapshot 2026-10-02.

2026 perimeters (1 January – 30 September) colored by PERCNA2K, the share of each fire’s own perimeter that falls inside a Natura 2000 protected site. Grey: 0% (no protected-area overlap). Bright yellow: perimeters that burned almost entirely inside protected land. Source: EFFIS rapid perimeters, snapshot 2026-10-02.

Explore the season yourself

Everything above is a summary. This section is for readers who want to watch the season happen, or to look at the fires one by one.

The cumulative chart at the top compresses four months into one frame, which is exactly what makes a step in the line easy to read past. So here is the same data at its own pace: the map and the line locked to one clock, one frame per day. A fire flares bright on its fire date (the day EFFIS records it starting), glows for two more, then settles to a scar, and the ring around it is sized by area, so a megafire announces itself rather than becoming one pixel among thousands. The animation holds for a beat on the three biggest days.

Watch 22 July. Two rings open almost together, one over the Gironde and one over Ávila: about 42,000 hectares in Spain and 37,000 in France on the same day. The line jumps by roughly 87,000 hectares in one step, the single biggest day of the summer. Then watch 6 August, much further southwest, when a single fire in Huelva adds about 44,000 more.

Figure 1: The 2026 season played back day by day, 1 June to 30 September; use the player controls to pause or scrub. Left: mapped perimeters accumulating, coloured by age (bright on the fire date, then fading to a muted scar); dot and ring radius scale with the square root of burned area, so marker size is proportional to area. Right: the same cumulative burned-area chart as above, drawn to the same date, against the 2017-2025 min-max band (grey) and median (dashed). The header reports the running total as a percentage of the historical median for that calendar date. Playback holds on the three largest burn days, which are annotated with the fires that drove them (any fire contributing at least 10% of that day’s area). The map viewport is framed on the 1st-99th percentile of fire locations, so a few outlying Nordic and Irish fires sit near or just beyond the edge. Source: EFFIS rapid perimeters, snapshot 2026-10-02; geometry in EPSG:3035.

What 2026 leaves behind

In July this page asked whether 2026’s early pace would hold or ease. Four months of data later, here is what the season leaves behind.

It was a severe year, and not a lonely one. 2026 finished as the second-worst summer of the last 10, at about 63% above a typical one. The only summer ahead of it is the one just before it: two of the worst seasons in this record are now back to back, 2025 and 2026. Two years do not make a trend, but they are exactly the kind of pair worth watching in 2027.

It was made in summer. The spring that looked strange in May turned out to be an average one. What set 2026 apart was July, and above all a handful of days, with 22 July the biggest of them, when several very large fires broke out at once in different countries.

It was wide. The usual hotspots burned, but so did countries that rarely burn this much: France, Serbia and Bosnia and Herzegovina had their worst summer in ten years. For France, the Médoc fire at Le Porge alone was bigger than Gironde’s whole 2022 season.

It was, as always, a story of giants. Most fires were small, but the largest 1% of them held about 55% of everything that burned over the summer. A season’s total is decided less by how many fires start than by the few that escape control on the worst days.

This page now stops updating, and it stays here as the record of the 2026 season, next to the 2025 one. The tracker will start again in June 2027.

NoteA final word of caution

Everything on this page is a rapid satellite estimate, not an official statistic. Full methods and caveats are on the About page.