Researchers have developed AIFS-SUBS, a new machine-learning model designed to improve sub-seasonal weather forecasting. This model adapts ECMWF's AIFS-CRPS system, employing a 24-hour autoregressive time step to mitigate error accumulation and incorporating additional atmospheric data. AIFS-SUBS demonstrates comparable probabilistic skill to the operational Integrated Forecasting System (IFS) for weeks 2-6, while significantly extending skillful forecasts for the Madden-Julian Oscillation and improving predictions for sudden stratospheric warming events. Notably, AIFS-SUBS achieves this with substantially lower energy consumption compared to traditional numerical models. AI
IMPACT This development could lead to more accurate and energy-efficient weather predictions, impacting fields reliant on sub-seasonal forecasts.
RANK_REASON The cluster describes a new research paper detailing a novel machine learning model for weather forecasting.
AI-generated summary · Google Gemini · from 3 sources. How we write summaries →
