Blog

New potential for wind and solar power generation forecast improvements - "Using ECMWF's forecasts" Workshop 2024

Illustrative image.

In July 2024, enercast shared observations from the June 2024 Using ECMWF's Forecasts workshop. Senior meteorologist Huan Zang's report covered data-driven weather prediction, experimental information tools, higher-resolution modelling and atmospheric composition data. These observations describe that event and should not be read as the current status of every initiative discussed.

Data-driven forecasting in its 2024 context

The original report discussed ECMWF's AIFS work and the evaluation of its weather predictions. It described both promising behaviour and areas requiring further testing. That discussion did not establish a universal power-forecast advantage or confirm that a particular enercast product had adopted the input.

A subsequent milestone

ECMWF announced that AIFS became operational on 25 February 2025, alongside its physics-based IFS. This later milestone updates the historical expectation that an operational AI system was still ahead. Renewable forecasting still needs appropriate variables, timely delivery and asset-specific validation; operational status alone does not determine the resulting power accuracy.

Other themes from the workshop

The report also discussed Destination Earth and the potential relevance of atmospheric composition information to events such as Saharan dust. Those are distinct data sources and research directions. Their usefulness for a PV forecast needs evaluation against generation observations rather than inference from the size or novelty of a dataset. The archived workshop report remains a record of ideas being assessed in 2024, with the later AIFS milestone clearly separated from the original account.

Get in touch

Put your forecasts to the test

Describe your assets, decision horizon and available measurements. Discuss the scope of a forecast evaluation with our team.

Discuss a backtest