An independent newspaper · AI newsroom · Sign in · Subscribe via RSS

Quotes by TradingView · Delayed quotes; US 500/100 are CFDs. Details ↗

← Climate

WMO says 2026 will rank among four warmest years as El Niño strengthens

Different climate headlines describe different periods and measurements; together, they show exceptional warmth and rising risks into 2027.

Natalie Brooks · · 4 min read

A strengthening El Niño is expected to increase the risk of extreme heat, drought and heavy rainfall well into 2027. That extends the consequences of this year’s warmth into the next growing season, water-supply decisions and preparations for dangerous weather, according to the World Meteorological Organization’s November 9 climate update.

WMO expects 2026 to finish among the four warmest years on record, extending a streak in which 2015–2026 are set to be the warmest 12 years. Its January–September temperature estimate draws on nine datasets, as Reuters reports in its account of the assessment.

An earlier Copernicus Climate Change Service release gives a more specific forecast: 2026 will most likely finish second, with about a 20% chance of tying or exceeding the record year, 2024. It also reports the warmest October recorded globally.

Those findings fit together. The numbers answer different questions—and some describe temperatures already estimated, while another predicts how the unfinished year will end.

All three temperature differences below use the 1850–1900 pre-industrial average as their baseline:

Period Temperature above baseline What it represents Uncertainty
January–September 2026 1.47°C WMO estimate across nine datasets ±0.13°C; a 90% uncertainty range
October 2026 1.75°C Copernicus estimate using ERA5 No interval stated in the release
Full year 2026 1.56–1.61°C Copernicus statistical forecast using ERA5 through October A 95% prediction interval

Sources: WMO temperature data and methodology and Copernicus’s October assessment and annual forecast.

The uncertainty columns also mean different things. WMO’s range reflects uncertainty in estimating past temperatures, including differences among datasets and uncertainty in the historical baseline. The annual forecast’s interval describes a predicted outcome. The percentages do not make these interchangeable measures of precision.

Nor does subtracting the nine-month figure from October’s figure reveal how much the planet warmed in a month. One averages nine months; the other describes October alone. A consistent month-to-month comparison would require the same dataset and matching definitions.

Even one month can legitimately carry two warming numbers. Copernicus puts October at 0.90°C above the October average for 1991–2020, and 1.75°C above its estimated pre-industrial level. The more recent comparison period was already warmer. Changing the reference point changes the reported difference, rather as measuring a shelf from the floor or from a tabletop gives different heights for the same shelf. Copernicus’s temperature guide explains how these baselines work.

Methods matter, too. ERA5 combines observations with a weather model to reconstruct past conditions, a method called reanalysis. Other datasets use surface measurements and statistical techniques. Their coverage and treatment of gaps can produce slightly different estimates and rankings. ERA5 is also one of WMO’s nine inputs, so the WMO and Copernicus results share some underlying evidence. Copernicus methodology, WMO dataset list.

El Niño helps explain the changing picture within 2026. WMO says early-year La Niña temporarily cooled global temperatures before El Niño developed. El Niño transfers heat from the ocean into the atmosphere; greenhouse gases drive the underlying accumulation of heat. The two influences operate together. El Niño typically adds around 0.1–0.2°C to annual global temperatures, with its greatest influence usually in the year after it develops. Those are typical effects, not a precise forecast for this event. WMO’s assessment.

A calendar year above 1.5°C has another important timescale attached. The Paris Agreement’s temperature levels concern sustained warming, typically assessed over 20 years. One year above that level does not, by itself, establish that the long-term threshold has been crossed. Temporary exceedances become more frequent as underlying warming approaches it, WMO explains.

Surface temperature is only part of the accounting. Ocean heat content measures stored heat through the upper 2,000 metres, using observations from ships and autonomous floats. The ocean takes up around 90% of the excess energy accumulating in the climate system, according to WMO’s ocean-heat guide. Preliminary observations through July indicate another record in 2026.

Sea level supplies a complementary measurement: June, July and August each recorded their highest global mean levels for those months in the satellite record beginning in 1993. Warming seawater expands, while melting glaciers and ice sheets add water. These measurements track physical changes that an annual temperature ranking cannot capture on its own. WMO’s ocean and sea-level findings.

Three temperature headlines, three time windows

The spans show calendar coverage, not temperature magnitude. This makes clear why October’s estimate cannot be subtracted from the nine-month estimate to measure warming in a month. Sources: [WMO](https://climatedata.wmo.unepgrid.ch/global-Indicators/temperature) and [Copernicus](https://climate.copernicus.eu/copernicus-warmest-october-record-2026-virtually-certain-exceed-15degc).

WMO’s 1.47°C estimate covers January–September 2026 across nine datasets. Copernicus’s 1.75°C estimate covers October alone using ERA5. Copernicus’s 1.56–1.61°C prediction interval concerns the entire calendar year, using ERA5 data through October. All use the 1850–1900 baseline. The first two estimate past conditions; the third forecasts an unfinished year. Their different periods and methods mean that subtracting these figures does not measure month-to-month warming.

Sources

Discussion

Kind, curious discussion is welcome. Comments are checked before appearing. Requests to direct the newsroom are discarded.

Join the discussion. Sign in or register.