Photo: Depositphotos

The European Center for Medium-Range Weather Forecasts (ECMWF) has introduced a new weather forecasting model based on artificial intelligence. About it writes Bloomberg.

The European language model AIFS ENS has been trained on long-term climate and weather data and is able to accurately predict atmospheric phenomena with significantly less computational resources than other modern methods.

Disable ads
Disable ads

In tests, the model also performed better than traditional models, particularly in predicting temperature, rain, and wind.

The emergence of such a tool opens up new opportunities for the energy market. Traders working with the renewable energy market, such as wind and solar, will especially benefit.

The new tool can provide them with quick calculations and predictions to use for their own artificial intelligence models and data algorithms without additional computing costs.

ECMWF also released the so-called "weights" of the model – a map of patterns and relationships that the AI found in the data. This means that users have access not only to the results but also to the internal logic of the model.

According to industry analysts, this could be the impetus for the creation of new weather systems aimed at both public forecasting services and businesses.

The technology has already attracted interest from Europe's largest energy companies. This includes Denmark, where InCommodities said it sees the new model as a "turning point" in forecasting.