Methods

Every number on this site comes from official data, processed by published, repeatable rules. Sentences such as "Odds favour a drier-than-normal December–February" are assembled from the numbers by fixed rules, not written freely by a person or a machine.

El Niño status

• Status and synopsis: copied by hand, word for word, from NOAA's monthly ENSO diagnostic discussion (second Thursday of each month).
• Strength: NOAA's Oceanic Niño Index (ONI). An El Niño episode is five or more overlapping 3-month seasons with ONI ≥ +0.5 °C. Strength bands: weak 0.5–0.9, moderate 1.0–1.4, strong 1.5–1.9, very strong ≥ 2.0. Seasons are named by their three months (DJF = December–February) and dated by the middle month.

Seasonal outlook

1. We download the Copernicus Climate Change Service (C3S) seasonal forecast from ECMWF's SEAS5 model (system 51): every ensemble member, monthly means of 2 m temperature and total precipitation.
2. We also download the same model's re-forecasts (hindcasts) for 1993–2016, for the same start month.
3. Both are averaged over each 3-month season (for a forecast issued in November: December–February and February–April) and over each country, weighting each grid cell by how much of it lies inside the country and by its area.
4. "Normal" is the middle third of all hindcast values for that country and season. The share of forecast members below, inside and above that range gives the odds of drier/normal/wetter (or cooler/normal/warmer).
5. Headline rules: "likely" when one side has ≥ 50% and leads the other by ≥ 20 points; "leaning" when one side has ≥ 40% and leads; otherwise "no clear signal".

Known weaknesses. One model only (more will be added). Countries smaller than a model grid cell (about 1°) use the nearest cell. Seasonal forecasts have real skill in the tropics during strong El Niño years, but limited skill in mid-latitudes such as Europe. Because the climate has warmed since 1993–2016, "warmer than normal" is the expected result almost everywhere.

Past El Niños

1. ERA5 monthly reanalysis (C3S), 1979 to last year, 2 m temperature and precipitation, averaged to 1° and then to each country as above.
2. Overlapping 3-month seasonal means for each year.
3. The straight-line trend over all years is removed, so that warming alone does not make recent El Niños look hot.
4. El Niño seasons (from the ONI episode list, including the current event) are compared with all other seasons: the average difference, and how many El Niño seasons were above the median.
5. Signal rules: "usually wetter/warmer" when at least 70% of El Niño seasons were above the median; "usually drier/cooler" when at most 30% were; otherwise "no consistent pattern". Rainfall differences are shown as % of the long-term average.

Known weaknesses. Around 9–15 El Niño seasons per calendar season is a small sample. Country averages can hide opposite effects within a country (for example Peru's wetter north coast and drier southern highlands).

Air quality

• Copernicus Atmosphere Monitoring Service (CAMS) global forecast, 00 UTC run, PM2.5 every 3 hours for 5 days, at about 40 km resolution.
• Cities: each country's capital and up to four of its largest cities (Natural Earth populated places). The nearest model grid point is used.
• Daily averages are for complete UTC days. The value is truncated to 0.1 µg/m³ and converted to the US EPA Air Quality Index for 24-hour PM2.5 (2024 revision): Good 0–9.0, Moderate 9.1–35.4, Unhealthy for Sensitive Groups 35.5–55.4, Unhealthy 55.5–125.4, Very Unhealthy 125.5–225.4, Hazardous 225.5 and above. The WHO 24-hour guideline is 15 µg/m³.

Known weaknesses. A model forecast, not a measurement. Local sources (traffic, cooking, nearby fires) can make levels much higher or lower than the 40 km average.

Maps and borders

Country shapes come from Natural Earth (public domain). Borders and names imply no position on any dispute.

Change log

• 15 November 2026: first version.