Enhancing AI and Dynamical Subseasonal Forecasts with Probabilistic Bias Correction

Hannah Guan, Soukayna Mouatadid, Paulo Orenstein, Judah Cohen, Haiyu Dong, Zekun Ni, Jeremy Berman, Genevieve Flaspohler, Alex Lu, Jakob Schloer, Joshua Talib, Jonathan A. Weyn, Lester Mackey · 2026

Read Enhancing AI and Dynamical Subseasonal Forecasts with Probabilistic Bias Correction by Hannah Guan, Soukayna Mouatadid, Paulo Orenstein, Judah Cohen, Haiyu Dong, Zekun Ni, Jeremy Berman, Genevieve Flaspohler, Alex Lu, Jakob Schloer, Joshua Talib, Jonathan A. Weyn, Lester Mackey (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-16238.txt.

Introduction

Guan and 12 colleagues introduce probabilistic bias correction, a machine-learning framework to improve subseasonal weather forecasts two to six weeks ahead. Skill drops sharply at that range as errors compound and biases build in the chaotic atmosphere. Applied to leading European Centre models, their method doubles the AI Forecasting System's modest subseasonal skill. It also improves the debiased dynamical model for 91 percent of pressure targets, 92 percent of temperature targets, and 98 percent of precipitation targets.

Catalog record

CREATORHannah Guan, Soukayna Mouatadid, Paulo Orenstein, Judah Cohen, Haiyu Dong, Zekun Ni, Jeremy Berman, Genevieve Flaspohler, Alex Lu, Jakob Schloer, Joshua Talib, Jonathan A. Weyn, Lester Mackey
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE54703
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.16238

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Enhancing AI and Dynamical Subseasonal Forecasts with Probabilistic Bias Correction Hannah Guan, Soukayna Mouatadid, Paulo Orenstein, Judah Cohen, Haiyu Dong, Zekun Ni, Jeremy Berman, Genevieve Flaspohler, Alex Lu, Jakob Schloer, Joshua Talib, Jonathan A. Weyn, Lester Mackey Abstract Decision-makers rely on weather forecasts to plant crops, manage wildfires, allocate water and energy, and prepare for weather extremes ( White et al. 2017 ; Vitart and Robertson 2018 ; Merryfield et al. 2020 ) . Today, such forecasts enjoy unprecedented accuracy out to two weeks thanks to steady advances in physics-based dynamical models and data-driven artificial intelligence (AI) models ( Bauer et al. 2015 ; Ben Bouallègue et al. 2024 ; Bi et al. 2023 ; Lam et al. 2023 ; Bodnar et al. 2025 ; Chen et al. 2023 ; Rasp et al. 2024 ) . However, model skill drops precipitously at subseasonal timescales (2 – 6 weeks ahead), due to compounding errors and persistent biases ( Chen et al. 2024 ; Nathaniel et al. 2024 ; Vitart et al. 2012 ; Richter and Joseph 2025 ) . To counter this degradation, we introduce probabilistic bias correction (PBC) , a machine learning framework that substantially reduces systematic error by learning to correct historical probabilistic forecasts. When applied to the leading dynamical and AI models from the European Centre for Medium-Range Weather Forecasts (ECMWF), PBC doubles the subseasonal skill of the AI Forecasting System ( Schloer et al. 2026 ) and improves the skill of the operationally-debiased dynamical model ( Tuppi et al. 2023 ) for 91% of pressure, 92% of temperature, and 98% of precipitation targets. We designed PBC for operational deployment, and, in ECMWF’s 2025 real-time forecasting competition ( Loegel et al. 2025 ) , its global forecasts placed first for all weather variables and lead times, outperforming the dynamical models from six operational forecasting centers, an international dynamical multi-model ensemble, ECMWF’s AI Forecasting System, and the forecasting systems of 34 teams worldwide. These probabilistic skill gains translate into more accurate prediction of extreme events and have the potential to improve agricultural planning, energy management, and disaster preparedness in vulnerable communities ( Shrader et al. 2026 ) . † † equal-contributors: These authors contributed equally to this work. † † equal-contributors: These authors contributed equally to this work. 1 Introduction Subseasonal forecasts—weather predictions 2 to 6 weeks ahead—lie at the heart of agricultural planning, water allocation, disaster preparedness, and energy management ( White et al. 2017 ; Vitart and Robertson 2018 ; Merryfield et al. 2020 ) . Yet accurate prediction at this timescale remains stubbornly difficult. Today, our best estimates of future weather come from dynamical models that numerically simulate how the atmosphere and oceans evolve over time ( Bauer et al. 2015 ; Ben Bouallègue et al. 2024 ) and artificial intelligence (AI) models that learn to predict the weather from historical data ( Bi et al. 2023 ; Lam et al. 2023 ; Bodnar et al. 2025 ) . Thanks to steady improvements in modeling, computation, and data collection over the past half-century, deterministic forecasting systems can now provide accurate point estimates of the weather up to two weeks in advance ( Chen et al. 2023 ; Rasp et al. 2024 ) . However, the two-week horizon also marks a transition in predictability: beyond this point, even small differences in initial weather conditions can lead to dramatic differences in outcomes, and a single deterministic forecast cannot capture this uncertainty ( Lorenz 1963 ; Lorenz 1969 ) . To surmount this predictability barrier, one often turns to probabilistic forecasting, that is, predicting a distribution over future weather rather than a single point estimate ( Gneiting and Raftery 2005 ; Palmer 2002 ; Mylne 2026 ) . Probabilistic forecasting explicitly encodes our uncertainty about the future and provides actionable estimates of the likelihood of high-impact extreme events. For both dynamical and AI models, probabilistic forecasts are typically generated by perturbing the inputs, parameters, or internal state of a model to produce an ensemble of possible future weather states ( Palmer 2019 ; Chen et al. 2024 ; Price et al. 2025 ; Mylne 2026 ) . While these ensembles routinely enjoy high skill at shorter lead times (out to 15 days), operational dynamical ensembles and even state-of-the-art probabilistic AI routinely underperform the climatological average at longer subseasonal lead times, especially outside of the tropics ( Chen et al. 2024 ) . A core difficulty is that leading subseasonal ensembles are generated iteratively over many time steps, and even small per-step modeling errors compound into large systematic errors over several weeks’ time ( Nathaniel et al. 2024 ) . This effect is especially pronounced for precipitation, which is sensitive to cascading inaccuracies across model components and highly variable in space and time ( Stephens et al. 2010 ; Bauer et al. 2015 ) . Indeed, subseasonal forecasting has long been considered a “predictability desert” due to compounding model errors and complex interdependencies among atmosphere, land, ocean, and sea ice components that evolve with different time scales ( Vitart et al. 2012 ; Richter and Joseph 2025 ) . Here we show that while existing subseasonal ensembles are indeed error-prone, their systematic errors are both predictable and correctable using machine learning, allowing for substantial gains in skill with minimal cost. To achieve this, we introduce a probabilistic bias correction (PBC) framework that adaptively learns to correct a model’s forecast distribution using its own historical predictions and observations of past weather. PBC is compatible with any input model, and we use it to advance the state of the art in both data-driven and hybrid (AI + dynamical) subseasonal forecasting. The resulting forecasts are designed for operational deployment and, in real-time competition Loegel et al. 2025 , outperform the best alternative dynamical, AI, and hybrid forecasting systems from forecasting centers and competitors around the globe. 2 Results 2.1 Probabilistic bias correction Figure 1 : Probabilistic bias correction (PBC) schematic. PBC takes as input lagged observations, climatological quantiles (e.g., quintiles, terciles, or deciles), and the deterministic forecasts of any dynamical or AI model and outputs a learning-enhanced probabilistic forecast designed to correct systematic prediction errors. PBC is a new machine learning framework for improving subseasonal forecasts from dynamical, AI, or hybrid prediction systems. Given a raw ensemble of deterministic forecasts, PBC outputs learning-enhanced probabilities designed to correct systematic errors while preserving the predictive signal in the underlying model. Our approach is inspired by the deterministic bias correction framework of Mouatadid et al. 2023 but operates directly on the space of probability distributions rather than at the level of raw observations. The added flexibility allows us to directly optimize not only the center of a distribution but also its spread and shape. Fig. 1 illustrates a sample application of PBC. First, an input ensemble is converted into an initial probabilistic forecast over climatological quantile bins derived from historical observations or from the model’s own hindcasts. Next, two computationally efficient machine learning models—Debias ++ and Persistence ++ —are applied in parallel to produce complementary corrections of the initial forecast. Debias ++ perturbs the forecast using location-, date-, and quantile-specific corrections learned by minimizing probabilistic forecasting error over adaptively-selected training periods. Meanwhile, Persistence ++ blends the

Read the complete text