Exceptional El Niño: Record Temperatures in the Pacific and Global Impacts

For a month now, El Niño has been breaking records: satellite records show surface temperatures of the Pacific Ocean exceeding values never observed since measurements began, creating an unprecedented thermal anomaly.

Exceptional El Niño: Record Heat in the Pacific and Affected Areas

Satellite data from Copernicus, combined with ECMWF analyses, indicate that the central equatorial Pacific has recorded a surface temperature above the seasonal average by more than 0.4°C, a deviation never observed since modern archives began in 1981. This increase was confirmed by Météo-France, which uses networks of oceanographic stations and Argo buoys to refine real-time tracking.

The phenomenon is not limited to an isolated point: the band of heat extends from the southern American coast to the Pacific islands, affecting more than 10% of the global oceanic surface. This massive expansion increases evaporation, thereby strengthening moisture fluxes into the atmosphere and altering global circulation patterns. Predictive models, such as GraphCast and Pangu-Weather, integrate these anomalies through machine learning, allowing for anticipated impacts over the coming weeks with a reduced forecast uncertainty of 15% compared to classic models.

The tracking of this exceptional heat relies on a dense grid of atmospheric and oceanographic data. Neural networks analyze temperature, salinity, and wind time series, detecting patterns that escape traditional statistical approaches. This ability to 'see' emerging trends improves the reactivity of weather services, allowing them to issue alerts earlier and with greater precision.

Mechanisms of El Niño: How Ocean and Atmosphere Amplify Heat

El Niño results from a breakdown of the Walker circulation phenomenon: trade winds weaken, the warm surface current extends westward, and upwelling of cold waters along the Peruvian coast decreases. This thermal redistribution releases enormous amounts of heat stored in the upper layers of the ocean, which are then transferred to the atmosphere via increased evaporation.

AI models such as GraphCast use connection graphs between measurement points to simulate these complex interactions. By integrating wind, pressure, and temperature data, the neural network can reproduce convection dynamics and anticipate the formation of high-pressure zones that will influence climate conditions on the other side of the globe. Pangu-Weather, on the other hand, relies on large-scale machine learning to reduce calculation time for forecasts, going from several hours to just a few minutes.