AIFS-SUBS: Extending Data-Driven Forecasting to Sub-Seasonal Timescales
2026-07-06 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors developed a machine-learning weather model called AIFS-SUBS aimed at forecasting weather from 2 to 6 weeks ahead. They improved it by using a daily step to reduce error buildup and adding new atmospheric data, such as stratospheric levels. Their model performs similarly or better than the traditional operational weather system in predicting key patterns like the Madden-Julian Oscillation and stratospheric events, while using much less energy. This work represents the first ECMWF machine-learning model focused on sub-seasonal (weeks-long) forecasts.
sub-seasonal forecastingautoregressive modelingMadden-Julian OscillationIntegrated Forecasting System (IFS)ERA5 reanalysisstratospheric warmingprobabilistic skill scoretop-of-atmosphere radiation
Authors
Jakob Schloer, Steffen Tietsche, Christopher D. Roberts, Lorenzo Zampieri, Simon Lang, Gert Mertes, Gareth Jones, Matthew Chantry, Frederic Vitart
Abstract
Data-driven models now rival numerical weather prediction in the medium range, but extending them to sub-seasonal lead times raises challenges absent at shorter horizons. Errors accumulate over long autoregressive rollouts, systematic biases grow with lead time, and several years of data must be held out for independent verification, even though machine-learning models otherwise benefit from longer training records. To address these challenges, we adapt ECMWF's AIFS-CRPS medium-range model. AIFS-SUBS adopts a 24h autoregressive time step to reduce error accumulation, adds stratospheric levels and top-of-atmosphere thermal radiation as predictors, and reserves 2007--2011 as an independent verification window. We evaluate two config-durations: AIFS-SUBS, fine-tuned on operational analyses, and AIFS-SUBS-ERA5, trained on ERA5 alone. Across weeks 2--6, AIFS-SUBS matches the operational Integrated Forecasting System (IFS) in probabilistic skill while reducing systematic biases. For the convective (OLR) component of the Madden--Julian Oscillation (MJO), AIFS-SUBS extends skilful forecasts (correlation > 0.5) by eight days relative to the IFS, while matching or exceeding the IFS for the full multivariate RMM index. AIFS-SUBS also reproduces the observed MJO modulation of tropical cyclone activity comparably. Stratospheric skill is particularly strong with AIFS-SUBS reproducing sudden stratospheric warming (SSW) frequency and surface impact. In the AI Weather Quest, AIFS-SUBS-ERA5 attains a variable-averaged ranked probability skill score slightly ahead of the IFS at weeks 3 and 4. At inference, AIFS-SUBS uses about 200 times less energy than the IFS, opening the door to much larger real-time ensembles. AIFS-SUBS is ECMWF's first machine-learning model targeted at sub-seasonal time-scales.