Retour au blog
TechniqueNouveau 10 min

AI and Weather Forecasting: Models, Accuracy and Decisions in 2026

For fifty years, forecasting the weather required supercomputers solving the equations of the atmosphere for hours. In two years, machine learning models have reshuffled the deck: GenCast beats Europe's reference ensemble on 97.2% of targets and produces a global 15-day forecast in 8 minutes, ECMWF has been running its own AI model operationally since February 2025 with roughly 1,000 times less energy, and the National Hurricane Center ranked an AI model first for cyclone track and intensity in 2025. For companies whose demand, production or risks depend on the sky, this is a concrete change. Here is what it means, seen from Belgium.

Article generated by AI. Content written with the help of an artificial intelligence model and reviewed by a human before publication. The figures cited point to their sources, listed at the end of the article.

The breakthrough in numbers

97.2%
of 1,320 verification targets on which GenCast beats ECMWF's reference ensemble (Google DeepMind · Nature, December 2024)
×1,000
approximate reduction in energy used per forecast with ECMWF's AIFS, operational since 25 February 2025
92%
of the $224 billion in natural catastrophe losses in 2025 were weather-related (Munich Re, January 2026)

The scientific tipping point came in December 2024: Google DeepMind published in Nature the GenCast model, trained on 40 years of ERA5 reanalysis data. Across 1,320 combinations of variables and lead times, GenCast beats ENS, the reference ensemble of the European Centre for Medium-Range Weather Forecasts (ECMWF), in 97.2% of cases, and in 99.8% of cases beyond 36 hours. Where a supercomputer grinds for hours, GenCast produces an ensemble of 50 or more scenarios over 15 days, at 0.25-degree resolution, in 8 minutes on a single TPU v5 chip. Its predecessor GraphCast, published in Science in late 2023, already beat the deterministic reference model on more than 90% of 1,380 targets.

The institutional shift followed within months. ECMWF, the European intergovernmental organisation of which Belgium is a member, took its own machine learning model into operations: AIFS Single on 25 February 2025, the first fully operational AI model at a major weather centre, with gains of up to 20% on tropical cyclone tracks and energy use per forecast cut by a factor of roughly 1,000; then AIFS ENS on 1 July 2025, an ensemble of 51 forecasts that outperforms physics-based models on many measures, including surface temperature, while running more than 10 times faster. Both systems run side by side with the physics-based IFS model, and their outputs are made available as open data.

The software giants are widening the field. Aurora, Microsoft's foundation model published in Nature in May 2025 and trained on more than one million hours of geophysical data, outperforms operational forecasts on air quality, ocean waves and cyclone tracks, at roughly 5,000 times lower computational cost; it beat the National Hurricane Center (NHC) on 5-day cyclone tracks for the 2022-2023 seasons. Google followed through: for the 2025 season, the NHC's verification report ranked WeatherNext the top-performing individual model for track and intensity; the model anticipated Hurricane Melissa's intensification to Category 5 five days ahead of its historic landfall in Jamaica. Since 18 November 2025, WeatherNext 2, 8 times faster and with hourly resolution, has powered forecasts in Google Search, Gemini and Pixel.

The Belgian context

Belgium knows the price of a forecast taken seriously: the July 2021 floods killed 39 people and generated nearly 74,000 claims for €2.3 billion, the costliest natural disaster in the country's history (Assuralia). The World Meteorological Organization notes that 24 hours' advance warning can cut storm damage by up to 30%. As an ECMWF member state, Belgium benefits directly from AIFS, and its weather-exposed sectors (North Sea offshore wind, agriculture, construction, port logistics) are among the first potential users of forecasts that are more accurate, more frequent and cheaper.

What it changes for business

Weather is not a niche topic: Munich Re puts natural catastrophe losses at $224 billion in 2025, 92% of them weather-related (and 97% of insured losses). Below the catastrophes, ordinary weather variability weighs every week on demand, lead times and costs. Better, cheaper forecasts translate directly into recurring decisions.

Why this is a breakthrough, not a gadget

Three factors combine. Speed: an ensemble forecast that required hours of supercomputer time now comes out of a single chip in minutes, which allows scenarios to be multiplied (WeatherNext 2 generates hundreds) and refreshed more often. Cost: roughly 1,000 times less energy per forecast for AIFS; at constant budget, a national weather service or a company can produce far more. Access: AIFS outputs are open data, Aurora is published as open source, GenCast's code and weights have been released; state-of-the-art forecasting is no longer reserved for those who own a supercomputer. The humanitarian stake follows the same slope: only half of the world's countries have multi-hazard early warning systems, and the UN aims for universal coverage by the end of 2027 with its Early Warnings for All initiative; the World Bank estimates universal access to early warning would prevent at least $13 billion in asset losses every year.

The limits to know

AI models do not start from nothing: they depend on the data assimilation of physics-based centres for their initial conditions, and therefore on the satellites, radars and stations that feed it. Their resolution remains coarser (31 km for AIFS ENS versus 9 km for ECMWF's physics-based ensemble), which limits the very local: isolated thunderstorms, valley gusts, hail. And unprecedented events remain the hard spot for systems trained on the past. Weather centres speak of complementarity and hybrid systems, not replacement.

Using AI weather, step by step

1

Quantify your weather exposure

Which decisions depend on the weather: demand, production, sites, transport, safety? What did a bad forecast cost over 12 months? This quantification frames the project.

2

Access the forecasts

ECMWF open data (AIFS), services from the national weather institute, commercial APIs that integrate AI models: start with existing feeds before considering an in-house model.

3

Plug into a recurring decision

Demand forecasting, staff planning, maintenance windows, internal alerts: one case, before/after indicators, one test season.

4

Govern and extend

Use ensembles rather than a single scenario, compare AI and physics-based output, keep human validation for safety decisions, then extend to the other weather-sensitive decisions.

Frequently asked questions

Do AI weather models replace physics-based models?

No, they complement them. AI models such as ECMWF's AIFS depend on physics-based data assimilation for their initial conditions, and their resolution (31 km for the AIFS ensemble) remains below that of the physics-based model (9 km). Weather centres run both side by side and are exploring hybrid systems that combine the strengths of both approaches.

How accurate have AI weather models become?

GenCast (Google DeepMind, Nature, December 2024) beats ECMWF's reference ensemble on 97.2% of 1,320 verification targets, and on 99.8% beyond 36 hours. ECMWF's AIFS improves tropical cyclone tracks by up to 20%. For the 2025 cyclone season, the National Hurricane Center's verification report ranked Google's WeatherNext the top individual model for track and intensity.

How can a company use AI weather forecasting?

In three steps: identify the weather exposure of the business (demand, production, logistics, sites, risks), access the forecasts (ECMWF's AIFS outputs are open data, Aurora is open source, commercial providers integrate them), then plug this data into recurring decisions: demand forecasting, staff and site planning, energy hedging, internal alerts. Measure the before/after effect on a limited perimeter, then extend.

Are AI forecasts reliable for extreme events?

They are progressing fast: WeatherNext anticipated Hurricane Melissa's intensification to Category 5 five days ahead in October 2025, and GenCast outperforms the European reference on extremes. But rare events remain the hard spot: models learn from the past, and an unprecedented phenomenon is by definition under-represented in it. Good practice remains the use of ensembles (50 or more scenarios), comparison with physics-based models, and human validation for safety decisions.

Sources

  1. Price, I. et al. (Google DeepMind), Probabilistic weather forecasting with machine learning, Nature, December 2024 (GenCast beats ECMWF's ENS on 97.2% of 1,320 targets, 99.8% beyond 36 hours; ensemble of 50+ scenarios; 15 days at 0.25° in 8 minutes on one TPU v5; trained on 40 years of ERA5). nature.com · DeepMind blog: deepmind.google
  2. ECMWF, ECMWF's AI forecasts become operational, 25 February 2025 (AIFS Single operational; gains of up to 20% on tropical cyclone tracks; energy per forecast cut roughly 1,000 times). ecmwf.int
  3. ECMWF, ECMWF's ensemble AI forecasts become operational, 1 July 2025 (AIFS ENS: 51 forecasts, gains of up to 20% including surface temperature, 31 km resolution versus 9 km for the physics-based ensemble, over 10 times faster, roughly 1,000 times less energy). ecmwf.int
  4. Bodnar, C. et al. (Microsoft Research), A foundation model for the Earth system, Nature, May 2025 (Aurora, trained on more than one million hours of geophysical data; air quality, waves, cyclones; roughly 5,000 times lower computational cost; beat the NHC on 5-day tracks for the 2022-2023 seasons). nature.com · feature: news.microsoft.com
  5. Google DeepMind, How WeatherNext helped the National Hurricane Center better predict Hurricane Melissa's historic landfall in Jamaica, 2025-2026 (NHC verification report for the 2025 season: WeatherNext top individual model for track and intensity; Category 5 anticipated 5 days ahead; WeatherNext 2 launched 18 November 2025, 8x faster, hourly resolution, deployed in Search, Gemini and Pixel). deepmind.google
  6. Munich Re, Natural disaster figures 2025, January 2026 ($224 billion in total losses, $108 billion insured; weather disasters account for 92% of total and 97% of insured losses; around 17,200 fatalities). munichre.com
  7. World Meteorological Organization (WMO) and UN, Early Warnings for All initiative (24 hours' warning can cut damage by up to 30%; only half of countries have multi-hazard early warning systems; universal coverage targeted by end 2027; World Bank estimate of $13 billion in asset losses prevented per year). wmo.int · un.org
  8. Assuralia, Actualisation relative aux inondations de juillet 2021 (39 fatalities, nearly 74,000 claims, €2.3 billion in losses, the costliest natural disaster in Belgian history; 98.3% of claims closed five years on). press.assuralia.be
  9. Lam, R. et al. (Google DeepMind), Learning skillful medium-range global weather forecasting (GraphCast), Science, 2023 (beat the deterministic HRES model on more than 90% of 1,380 targets). science.org

Does your business depend on the sky?

Molderez Consult helps Belgian companies quantify their weather exposure, plug AI forecasts (ECMWF open data, specialised APIs) into their demand, planning and risk decisions, and measure the gain, from the first indicator to the internal alert system.

Discuss my project
Article generated by AI. Content written with the help of an artificial intelligence model and reviewed by a human before publication. The figures cited point to their sources, listed at the end of the article.
Partager