Relaxation experiments in ML-based weather prediction models to study subseasonal predictability

This study explores the use of relaxation experiments in machine learning-based weather prediction (MLWP) models to identify sources of subseasonal forecast errors and compares their effectiveness to a traditional numerical weather prediction (NWP) system. Relaxation involves nudging specific regions of a model toward reanalysis data to isolate their influence on forecast skill. We apply this technique to two MLWP models, Pangu-Weather (fully data-driven) and NeuralGCM (hybrid) and compare them to the Unified Forecast System (UFS) during Week 3–4 forecasts of two major precipitation events in western North America in winter 2022/2023, both linked to Madden-Julian Oscillation (MJO) activity. MLWP models outperform UFS at subseasonal lead times. While tropical relaxation improves all forecasts, gains are greater for UFS, reflecting MLWP models’ stronger baseline performance. These results highlight relaxation experiment as a low-cost, effective diagnostic for understanding and improving subseasonal forecasts, especially in emerging MLWP systems.