e-nature.org

European heat monitor

Summer heatwaves over Europe from ERA5 daily maximum temperature: extent, statistically significant hot clusters, and events tracked in space and time. Every number is against a stated 1991–2020 baseline.

updated 12 hours ago

Latest processed day — 2026-09-30provisional

Land in heatwave
11.65% of land
In significant hot clusters
14.87% of land
Events this season
25
Active on this day
1

ERA5 daily maximum 2 m temperature over land, against the 1991–2020 climatology (p90 in a centred 15-day window). A heatwave cell is Tx above p90 on at least 3 consecutive days.

The most recent days are ERA5T (preliminary) and the last 2 carry a provisional flag: their heatwave status can still change as later days arrive. ERA5 daily statistics lag real time by about 6 days, so the newest row here is not today.

Outlook — blocking probabilityforecast

Share of the 31-member NOAA GEFS v12 ensemble with a persistent block — 5 days or longer — over western Europe, by longitude. The quantity is Tibaldi-Molteni blocking at 500 hPa.

5° W
0% of members · unlikely
0°
0% of members · unlikely
5° E
0% of members · unlikely
10° E
0% of members · unlikely
15° E
0% of members · unlikely

Each bar below is one forecast day at that longitude, left to right; the height is the share of members blocked. Tall bars early mean the ensemble already agrees, tall bars late mean it agrees about something it cannot yet see.

5° W

11-1211-27

0°

11-1211-27

5° E

11-1211-27

10° E

11-1211-27

15° E

11-1211-27

All longitudes, all forecast days

Blocking probability by longitude and forecast day
Two panels, two different questions.

Top — where and when. Each row is one forecast day, each column a longitude from 30° W to 40° E. Colour is the share of the ensemble with a block at that place on that day: pale is nobody, dark is everybody. The cyan lines mark the five longitudes summarised above.

Read it downwards as confidence draining away. The top rows are nearly black or nearly blank because the members have barely diverged yet — they all start from the same analysis, so a dark band there is a ridge that is already present, not a bold forecast. Further down the members disagree and the field turns mottled. That mottling is the forecast: a solid patch at day 10 is a real signal, a speckled one is the ensemble saying it does not know. A band drifting left to right down the figure is a ridge moving east; one that holds the same longitude for several rows has stalled — and a stalled ridge in summer is what puts a heat dome underneath it.

Bottom — how likely a lasting block is, anywhere in the period. This is not the average of the column above it, and the difference matters. For each longitude it counts the members in which the block lasted at least 5 consecutive days. A member blocked on five scattered days contributes nothing; one blocked on five days in a row contributes. So the top panel can look busy while this stays low — which means blocking that comes and goes, not blocking that sits. Persistence is the part that matters, because a ridge has to stall before heat builds beneath it.

The dashed line is 60%, the point above which the ensemble is calling it rather than hedging. The five numbers at the top of this section are this panel, read off at those five longitudes.

Probability that the CIRCULATION blocks, not that a heatwave occurs. A block is a necessary ingredient of a European heat dome, not a sufficient one. Skill falls with lead time; what the ensemble adds is the spread — a tight 90% is worth acting on, a 40–60% split means the models genuinely disagree. Cycle 2026-11-12T00:00Z, computed 2026-11-12 10:45:50 UTC.

Active on the latest processed day

EventSinceDaysPeak areaMax anomalyCentroid drift
#242026-09-11201.15 M km²+13.9 °C1758 km

"Active" means the event is still running on the last day the pipeline has processed — not that it is happening now. ERA5 daily statistics lag real time by about 6 days.

Season 2026 — tracked events day-by-day evolution →

#StartEndDaysPeak areaArea·daysPeak dayMax anomMean anomDrift
12026-05-012026-05-10101.27 M km²6.63 M km²·d2026-05-04+12.5 °C+8.7 °C1748 km
22026-05-112026-05-24141.26 M km²7.31 M km²·d2026-05-20+18.4 °C+10.8 °C1904 km
32026-05-162026-05-249104k km²447k km²·d2026-05-21+14.6 °C+9.0 °C166 km
42026-05-202026-06-06181.94 M km²14.75 M km²·d2026-05-27+15.5 °C+8.7 °C406 km
52026-05-222026-06-0212178k km²1.07 M km²·d2026-05-25+16.5 °C+9.5 °C148 km
62026-06-012026-06-1313282k km²1.46 M km²·d2026-06-05+15.8 °C+9.5 °C677 km
72026-06-072026-06-137297k km²1.09 M km²·d2026-06-11+10.4 °C+7.5 °C238 km
82026-06-112026-06-188230k km²919k km²·d2026-06-13+16.5 °C+9.8 °C210 km
92026-06-122026-07-26452.16 M km²42.57 M km²·d2026-06-28+19.8 °C+8.9 °C687 km
102026-06-212026-06-277180k km²882k km²·d2026-06-23+16.9 °C+10.4 °C157 km
112026-06-242026-06-296104k km²367k km²·d2026-06-27+14.6 °C+8.8 °C346 km
122026-07-022026-07-2423183k km²1.61 M km²·d2026-07-08+13.1 °C+7.6 °C342 km
132026-07-082026-07-2619227k km²2.86 M km²·d2026-07-17+13.4 °C+7.0 °C300 km
142026-07-082026-07-1710366k km²1.59 M km²·d2026-07-15+12.5 °C+8.1 °C297 km
152026-07-172026-07-226147k km²631k km²·d2026-07-19+8.8 °C+5.6 °C538 km
162026-07-262026-08-19251.81 M km²22.88 M km²·d2026-08-05+15.7 °C+7.7 °C1008 km
172026-08-082026-08-1811180k km²775k km²·d2026-08-12+15.2 °C+8.1 °C48 km
182026-08-152026-08-195112k km²340k km²·d2026-08-17+10.1 °C+6.1 °C404 km
192026-08-172026-09-12271.33 M km²13.57 M km²·d2026-09-07+14.7 °C+7.1 °C291 km
202026-08-222026-09-0313173k km²1.90 M km²·d2026-08-26+14.0 °C+6.8 °C218 km
212026-08-222026-08-276123k km²418k km²·d2026-08-23+8.1 °C+4.8 °C248 km
222026-09-052026-09-106138k km²506k km²·d2026-09-07+8.0 °C+5.2 °C736 km
232026-09-102026-09-145138k km²456k km²·d2026-09-12+10.2 °C+6.9 °C451 km
242026-09-112026-09-30201.15 M km²12.28 M km²·d2026-09-27+13.9 °C+7.8 °C1758 km
252026-09-202026-09-289487k km²2.02 M km²·d2026-09-24+14.9 °C+9.1 °C764 km

Largest: #9, 2026-06-12 → 2026-07-26, peak 2.16 M km², max anomaly +19.8 °C.

ERA5 is a smooth reanalysis and under-states extreme station maxima. Extent, duration, structure and tracks come from here; absolute peak temperatures should be quoted from E-OBS or station records, not from this page.

Figures

Extent through the 2026 season

Extent through the 2026 season
Share of European land in heatwave, and in significant hot clusters, day by day.

Significant clusters, latest processed day

Significant clusters, latest processed day
Local Moran's I quadrants; only cells passing the FDR threshold are shaded.

Event footprints and tracks, 2026

Event footprints and tracks, 2026
Each tracked event's footprint, with the centroid path across its lifetime.

Daily bulletin

europe-heat-pipeline — 2026-11-12 12:15 UTC — off-season

Method and baselines

ERA5 daily maximum 2 m temperature over land, against the 1991–2020 climatology (p90 in a centred 15-day window). A heatwave cell is Tx above p90 on at least 3 consecutive days.

Significant clusters are local Moran's I on the daily anomaly, k-NN weights (k = 8), with p-values from the normal approximation to the permutation distribution and Benjamini–Hochberg FDR across land cells each day. That choice is deliberate and conservative; it is stated because the test used changes what 'significant' means.

ERA5 is a smooth reanalysis and under-states extreme station maxima. Extent, duration, structure and tracks come from here; absolute peak temperatures should be quoted from E-OBS or station records, not from this page.

Global Moran's I is logged for monitoring only. Reanalysis fields are smooth, so it is almost always significant, and it is not a finding.

The most recent days are ERA5T (preliminary) and the last 2 carry a provisional flag: their heatwave status can still change as later days arrive. ERA5 daily statistics lag real time by about 6 days, so the newest row here is not today.

This site computes nothing. Every value above was produced by the europe-heat-pipeline run recorded below and is served unchanged.