NASA POWER · city estimates + Copernicus ERA5 regional benchmark · through 31 August 2026
Were summer nights unusually warm in 2026?
—
Updated 13 September 2026 · City temperature data through 31 August · Copernicus bulletin published 10 September
Selected sample: 28 European city locations, including 11 in Portugal, plus 10 North American cities. Regional averages are simple unweighted means.
Median night-time anomaly
—
Across the selected cities, average daily minimum, °C
Cities above baseline
—
2026 average daily minimum above 2000–2025
Warmest nights
—
—
Largest night anomaly
—
—
Where were 2026 nights warmer?
Two focused panels keep the city labels readable. Each dot shows the change in average daily minimum temperature during 1 May–31 August, compared with that city’s 2000–2025 average.
Europe
—Portugal
—North America
—cooler than baselinenear baselinewarmer than baselinelabels show the exact anomaly
The map uses a gridded reanalysis estimate at each city coordinate, not a thermometer in the city centre. Use the city comparison below for the monthly and year-by-year detail.
Western Europe’s summer set a regional record
Copernicus ERA5 · June–August 2026 · relative to 1991–2020
Summer anomaly versus the previous regional record
2026Previous record
This is a regional ERA5 area average for a complete June–August season. The city panels above use NASA POWER point estimates through 31 August, so the two views answer different questions.
Sea-surface context: extra-polar global SST reached 21.07°C in August, a record for the month (previous record 20.98°C in 2023).
Source: Copernicus Climate Change Service, 10 September 2026. Dry soils and low river flows were also reported in the companion hydrology bulletin across parts of western and central Europe.
The question behind the numbers
People often notice a hot afternoon first. The harder question is what happens after sunset, when homes should start to cool. This dashboard tests that impression with a 2026 snapshot across selected European and North American cities.
The comparison uses NASA POWER’s daily 2-m temperature estimates for the same dates in every year from 2000 to 2026. It tracks the daily mean temperature and the daily minimum temperature, then asks how far 2026 sits above or below the 2000–2025 average. August is matched to 1–31 August in every year, and the 2026 snapshot runs through 31 August.
—
How to read it: a positive anomaly means the 2026 value is warmer than that city’s 2000–2025 average for the same dates. The data show how unusual this season has been in the selected sample. One summer cannot establish a long-term trend or attribute a cause on its own.
What the averages hide
Averages describe the season. Threshold counts show how often people encountered persistent heat. The cards use the matched 1 May–31 August 2026 window.
—
Article context: —
Which cities saw the most persistent heat?
Bars show 2026 counts. The line under each city compares it with that city’s 2000–2025 average for the same window.
0—
Show all threshold counts by city
| City | Days ≥35°C | Days ≥40°C | Nights ≥20°C | Nights ≥25°C | Longest ≥20°C streak |
|---|
—
Every city, month by month
Anomaly in average daily minimum temperature. Positive cells mean warmer nights than the same month in 2000–2025. August covers 1–31 August in all years.
| City | May | June | July | August 1–31 | Season |
|---|
Swipe horizontally to see June–August and the season column.
This table is the quickest screenshot of the cross-city result. It shows temperature differences, not a ranking of climate risk.
Portugal in focus
A closer look at 11 Portuguese locations, covering the mainland and islands. The bars show 2026 versus each location’s 2000–2025 average for the matched 1 May–31 August period.
—
Show the detailed Portugal location table
| Location | Mean Δ | Minimum Δ | 2026 | 2000–2025 |
|---|
The Portugal average is an unweighted mean of the 11 selected locations. Inland, coastal and island climates are deliberately shown together so their differences remain visible.
2026 versus recent summers, city by city
Ranked horizontal bars show how far 2026 sits above or below the selected historical period. The two panels separate daily mean temperature from average daily minimum temperature.
Show the detailed city table
| City | Mean · 2026 | Mean · 2000–25 | Mean Δ | Min · 2026 | Min · 2000–25 | Min Δ |
|---|
Bars are sorted from the largest positive difference to the largest negative difference. The exact 2026 values, historical averages and signed differences remain in the table below. August is matched to 1–31 August in every year.
Minimum temperatures: how did the city’s nights compare?
The primary comparison is the average of each day’s lowest temperature. Use the second button to switch to daily mean temperature.
Was 2026 the warmest matched season since 2000?1 May–31 August · 27 matched seasons
20262000–2025 averagehistorical 10th–90th percentile
—
Top 3 warmest years, 2000–2026 2026 enters the podium when it ranks #1–3 · daily mean and average daily minimum
Dots show 2026. The blue dot is the 2000–2025 average; the grey whisker spans the historical 10th–90th percentile for that month. “Minimum” is the average of daily lows, not the single coldest night.
Was 2026 warmer than recent summers?
For each month, compare 2026 with the average for the previous 5 years (2021–2025), 10 years (2016–2025), and 26 years (2000–2025). The same dates are used in every period, including 1–31 August.
Average daily minimum
The overnight signal, in °C
20262021–252016–252000–25
Daily mean temperature
The whole-day signal, in °C
20262021–252016–252000–25
—
Bars show absolute temperatures. The useful comparison is the vertical gap between the orange 2026 bar and each historical bar. A positive gap means 2026 was warmer for that month.
The same-season trend
The matched 1 May–31 August value for every year from 2000 to 2026. The dashed line is the 2000–2025 average.
MetricView
—
Is the sea warmer too?
Daily sea-surface temperature at selected European coastal points during 1 May–31 August 2026, compared with the same dates in 2000–2025.
NOAA OISST · 0.25° grid
—
Map period
cooler than baselinewithin ±0.25°Cwarmer than baselineDots show exact selected-cell values. The shaded field is a visual guide.
Ranked coastal points
Seasonal sea-surface temperature, 2026 versus the 2000–2025 average.
| Point | 2026 | Baseline | Δ |
|---|
Monthly sea-surface temperature
—
20262000–2025 averagelabels show the anomaly
Was 2026 the warmest sea-surface temperature year by month?
The ranking includes 2026, alongside the completed years from 2000 to 2025.
| Month | 2026 | 2026 rank | Warmest year | Best before 2026 | 2026 vs best |
|---|
—
Heat and excess deaths in Europe: what the data shows
A separate mortality check: comparable country data through September 2025, with a dated EuroMOMO snapshot and newer national reports for 2026.
What this can answer: whether unusual summer temperatures and all-cause excess mortality moved together in recent years. The national 2026 figures are provisional estimates from separate systems, with different windows and methods. The comparable country-by-country series still ends in 2025.
—
The pandemic shock faded, then summer excess varied
Annual all-cause excess mortality, 2020–2025. The blue line is the median across 31 countries; highlighted lines show selected countries.
31-country medianPortugal, Spain, France and Germany
2025 summer excess: a country ranking
June–August average all-cause excess mortality, latest comparable year.
Do hotter summers coincide with more excess deaths?
Country-years with comparable data, 2020–2025. Each point compares a June–August temperature anomaly with all-cause summer excess mortality.
Zero means the country-year was at its expected mortality or temperature baseline
—
Country comparison: latest summer, peak year and temperature context
All countries in the reviewed Eurostat extract. Summer means June–August; temperature columns use the selected European city composites.
| Country | 2025 summer excess | Peak summer year | Peak excess | Positive summers | 2025 mean temp Δ | 2025 minimum temp Δ |
|---|
—
Sources and method
Daily temperature values come from the NASA POWER Daily API, requested in UTC for the city coordinates shown here. The meteorological fields are derived from NASA GMAO’s MERRA‑2 assimilation model. T2M is the daily average; T2M_MIN is the minimum of the 24 hourly 2-m temperatures.
- Current snapshot: 1 May–31 August 2026, the latest complete range available in this pull.
- Baseline: the same dates in 2000–2025. City anomalies are 2026 minus that city’s historical average.
- The city-by-city comparison also offers matched 2021–2025 and 2016–2025 averages, labelled as the previous 5 and 10 years.
- “All European cities” is a simple unweighted average of the European city series shown in this dashboard.
- NASA POWER is a gridded reanalysis estimate. It is consistent across cities, but it does not capture every station, street or urban heat-island effect.
- Extreme-heat counts use NASA POWER T2M_MAX and T2M_MIN at each selected coordinate. They are city-point counts, not population exposure. The BBC/AFP regional figures use the European Drought Observatory’s gridded data and should be read separately.
- NOAA GHCN Daily and Europe’s ECA&D are the next validation layer for station-level observations. ECA&D’s public current series did not yet cover this summer in the source check.
- Sea-surface temperatures use NOAA NCEI’s OISST v2.1 daily 0.25° grid. The final product supplies 2000–2025; the near-real-time product supplies 2026. Markers sit on selected offshore grid cells, so they describe open-water SST rather than water at a particular beach or depth.
- Copernicus C3S uses ERA5 area averages for June–August. Its Western Europe region is 11°W–15°E and 37°–55°N; the regional benchmark is separate from the city point estimates above.
- Excess-mortality context uses the EuroMOMO weekly bulletin and the reviewed Eurostat demo_mexrt monthly extract. The comparable country series ends in September 2025; 2026 EuroMOMO country values are not bundled in this page.
The dashboard describes observed/reanalysed temperature differences. It does not claim that a single city’s 2026 summer was caused by climate change. A longer trend analysis should use a homogeneous station series or gridded climate product over several decades.