UK & Europe, summer 2026 · daily max temperature anomaly vs 1991–2020
Each region is coloured by how far its daily maximum temperature sat above or below what is normal for that place at that time of year, averaged over 1991–2020. Deep red does not mean “hot” — it means hotter than summer itself normally is, by as much as 15–19 °C in the worst-affected regions. Scrub the timeline to watch the heatwave build, peak and break down, and click any region for its daily figures against the local norm.
The purpose of this visualisation is specific: it shows the severity of temperature extremes within a single season, not gradual average warming. A warming climate does not simply nudge every day a degree or two upward — it disproportionately inflates the extremes, making events that were once vanishingly rare arrive harder and more often. Heat is the clearest example, but the same physics amplifies other extremes too: longer droughts, heavier rainfall and flooding, and — through a destabilised jet stream — even severe cold outbreaks. This map documents one such extreme in detail.
In the final week of June 2026, a heat dome settled over Western Europe and delivered what World Weather Attribution analysis described as the most severe heatwave ever recorded over the region — the second major heat event of the year, arriving barely a month after a record-shattering late-May episode. On 23 June, France logged its hottest day since national measurements began in 1947: 44.3 °C at Pissos, with 42.1 °C in Bordeaux and a June record of 40.9 °C in Paris.
In the UK, the Met Office issued a red extreme-heat warning on three consecutive days for the first time in the warning system’s history, alongside UKHSA red heat-health alerts across six English regions. The heat peaked on 26 June at 37.7 °C at Lingwood, Norfolk — a provisional England and UK June record, beating a mark that had stood since 1976 — and over 150 individual stations broke their own June records. Nights brought little relief: Cardiff never fell below 23.5 °C on the 25th, a UK June record, and London Ambulance Service logged its busiest day ever on the 26th. As the dome drifted east on 27–28 June, Germany set an all-time national high and Germany, Czechia, Poland and Hungary all recorded their hottest days on record within about 24 hours; national June records also fell in the Netherlands (39.4 °C) and Switzerland (39.0 °C at Basel).
The month closed as England’s warmest June on record and the UK’s second-warmest. The human cost is still being counted: across the UK’s May and June heatwaves, researchers estimate more than 2,700 excess deaths. World Weather Attribution’s rapid analysis concluded it was “virtually impossible” to explain the event’s intensity relative to the benchmark 1976 heatwave without accounting for human-caused climate change. UK national records remain provisional pending Met Office verification.
The comparison already accounts for that. Every value here is measured against the 1991–2020 average for the same region and time of year — against what summer normally delivers. What you are seeing is the excess above summer itself.
Individual heatwaves are indeed as old as weather, and 1976 was exceptional for its time. The question science asks is not whether heatwaves happen, but whether their frequency, intensity and extent are changing — and on that, the trend data are unambiguous: events of a given severity now recur far more often, and new records are set by larger margins, across a larger area. Rapid attribution studies (such as those by World Weather Attribution) estimate specifically how much more likely climate change made an individual event like this one.
Correct — and this map does not claim otherwise. A single season is weather. This visualisation describes one event precisely; the climate context comes from decades of observations and the attribution literature, not from this map alone. The two together are the point: this is what the statistics look like when they arrive in one place, in one month.
Both, deliberately. The gridded values are from ERA5, the European reanalysis, which blends millions of quality-controlled observations — weather stations, ships, aircraft, satellites — with a physics model to produce a consistent record across the whole continent. Because each grid cell averages ~31 km, local peaks are smoothed: station records (like the 44.3 °C at Pissos) typically exceed the values mapped here. In other words, this map is conservative. National records shown in event markers are labelled provisional until formally verified by the relevant weather services.
Urban heat islands are real but cannot explain what you see here: these are region-scale averages spanning countryside, coast and mountains, not city-centre thermometers — and the reanalysis grid is far coarser than any urban footprint. The anomaly pattern sweeps across rural France and the North Sea coast just as it does across cities.
It is the current World Meteorological Organization standard reference period. Note that it is itself the warmest normal period on record — measured against the older 1961–1990 baseline, every anomaly on this map would be larger. The choice here is the conservative one.
No — it is expected. Warming shifts the whole temperature distribution while weather keeps its variability, so cold extremes still occur; they just become rarer while hot extremes become more frequent. The ratio of new hot records to new cold records has been widening for decades, which is precisely the fingerprint of a shifting distribution rather than a constant climate.
Data: ERA5 (Copernicus/ECMWF), Open-Meteo, Eurostat/GISCO boundaries — see full attributions and licences. Built as an open, reproducible pipeline; the aggregation code and this site’s source are public.