What El Niño does to every country.
El Niño does not add rain to the planet. It moves it, and it moves heat. This atlas asks one question of every country and every season from the developing summer to the winter after: in how many of the strong El Niños did this place go the same way? Each answer carries its sample size, an interval and a test against chance across all 237 countries at once. Where nothing separates from an ordinary year, the map says so.
The atlas
2026 so far: is it behaving like an El Niño?
Deeper
The pattern inside countries and over the ocean
GPCP, 1979 onA country average hides the pattern inside large countries (northern and southern Brazil go opposite ways) and says nothing about the ocean, where the rain belt actually moves. These gridded composites from precip.earth fill that in: satellite-and-gauge rainfall (GPCP 2.3) since 1979, per event, solid where it survives the false-discovery test. Shorter record, different events (MEI at ±0.75σ), same story.
How the atlas counts
ENSO years. A year is El Niño when its Nov–Jan Niño-3.4 value (NOAA's ONI from 1951, HadISST before) is +0.5 °C or more, strong at +1.5 °C, La Niña at −0.5 °C or less. The modern record (1951–2026) has 27 El Niños, 8 of them strong (1957–58, 1965–66, 1972–73, 1982–83, 1997–98, 2009–10, 2015–16, 2023–24), and 25 La Niñas.
The baseline is an ordinary year, not an average. For each country and season, a trend is fitted to the neutral years only and removed, so a warming or drying trend is not mistaken for El Niño. "Typical" means the typical neutral year.
Hit counts. An event counts toward "drier" when it lands below the median neutral year. "Drier in 7 of 8" comes with a 95% Wilson interval. With 237 countries × 19 seasons, many 7-of-8s are luck, so every map runs the Benjamini–Hochberg false-discovery test (q = 0.10) on a Mann–Whitney comparison of event years against neutral years: outlined countries survive it, hatched ones do not.
Where people live. Country values are weighted by 2025 population (what people experience); each country page also gives the area-weighted value. Rain is GPCC gauge analysis from 1891 (ERA5 where a country has no usable gauge cell); temperature is ERA5 from 1940. Other drivers are rank correlations over 1951–2025, with ENSO's share removed where marked.