AIFS-SUBS: Extending Data-Driven Forecasting to Sub-Seasonal Timescales

2026-07-06

AI Summary

This study addresses key challenges in subseasonal (2–6 week) weather forecasting—namely, error accumulation from autoregressive modeling, growing systematic biases, and limited verification data—by extending deep learning to this timescale for the first time. Building upon ECMWF’s AIFS-CRPS model, the approach employs a 24-hour autoregressive step, incorporates stratospheric variables and top-of-atmosphere thermal radiation as predictors, and uses 2007–2011 as an independent validation period. The resulting model matches the probabilistic skill of ECMWF’s operational IFS system while significantly extending the predictability of Madden–Julian Oscillation (MJO) convection—increasing lead time by eight days for forecasts maintaining correlation above 0.5—and accurately captures MJO modulation of tropical cyclones and sudden stratospheric warming events, all at just 1/200th the inference energy cost of IFS.

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.

Problem

• sub-seasonal forecasting
• data-driven models
• error accumulation
• systematic bias
• verification

Innovation

• sub-seasonal forecasting
• machine learning
• autoregressive modeling
• stratospheric predictors
• energy-efficient inference