In recent years, Weather forecasting has undergone a true revolution thanks to the development of new scientific models that incorporate artificial intelligence.The European Centre for Medium-Range Weather Forecasts (ECMWF) has taken a firm step with the arrival of AIFS ENS, an innovative probabilistic system that redefines the way weather forecasts are generated and managed.
What is the new probabilistic model?
AIFS ENS v1 is an ensemble model that uses machine learning techniques to simulate atmospheric behavior and generate weather forecasts with a broader view of possible future situations. This system performs multiple simulations from the same initial situation, sampling a learned distribution, which allows to capture the uncertainty inherent in weather predictions.
Thanks to this approach, the forecasts are achieved more precise and realisticThe model employs the CRPS loss function, which helps calibrate the results, taking into account the limitations associated with working with a finite number of ensemble members. As a result, AIFS ENS has outperformed traditional physical ensemble models in medium-range forecasting and is very competitive in subseasonal forecasts..
Main differences with respect to traditional models
One of the most relevant characteristics of the AIFS ENS It's the way it incorporates the control member. While in traditional physics-based models, this member acts as a deterministic, unperturbed reference, in the AI-based model, this role is different. The AIFS ENS control member is a product of the internal sampling of the distribution learned by the system., which means that uncertainty cannot be turned off to run a simulation exactly identical to the classical scheme.
This innovation represents an advance in the capacity of anticipate complex weather phenomena and assess associated risks by considering the natural variability of the atmosphere in predictions. If you want to delve deeper into how weather models work, you can consult other weather models and its importance in weather prediction.
Evolution and chronology of implementation
The model went through an experimental phase in which different methodologies were tested, such as the diffusion technique, although the operational version focuses exclusively on optimization with the CRPS loss function. The incorporation of AIFS ENS into ECMWF's forecasting systems is scheduled for 1 July 2025 at 06 UTC., following a testing phase that began on June 23.
For now, users of other models such as IFS and AIFS Single will not experience any changes, as the operational versions of these systems remain intact.
Impact and recommendations for users
The arrival of AIFS ENS marks a before and after in the managing meteorological uncertainty and forecast accuracy. However, those who intend to use these data, particularly for operational purposes, should thoroughly review available information on known and pending issues. ECMWF also encourages the scientific and technical community to provide feedback for further refinement of the system.
AIFS ENS is not intended to replace traditional models immediately, but rather complements the range of tools available for weather forecasting with more advanced approaches adapted to the era of machine learning. To better understand the evolution of these models, it may be interesting to review .
The development and application of models such as AIFS ENS opens a new stage in meteorological forecasting, improving anticipation and risk management capabilities In a global context where extreme events are gaining prominence, the continued improvement of these tools promises more useful forecasts for both professional users and the general public.
