Domain-Adaptive Data Assimilation for Global AI Weather Forecasting

2026-10-02

AI Summary

This study addresses the degradation in forecast performance of AI-based weather models caused by distribution shifts between training and real-time initial conditions. We propose a Domain Adaptation Data Assimilation (DADA) framework that optimizes initial state perturbations and incorporates observational constraints, transforming real-time observations into a universal interface for independent analysis and forecasting systems. Crucially, this approach enables frozen-parameter pretrained models to adapt effectively to external analysis systems without requiring model retraining or ERA5 data reconstruction. Experiments conducted across five global AI weather models demonstrate that DADA significantly mitigates the deterioration in short-range forecast skill induced by switching initial condition sources, yielding comprehensive improvements in both deterministic and probabilistic forecast accuracy.

Abstract

AI weather forecasting models are commonly trained on the ERA5 reanalysis, which is unavailable in real time. Operational deployment therefore relies on initial conditions produced by numerical or AI analysis systems that differ from those encountered during training. This mismatch can degrade forecast skill, while retraining for every analysis system is costly. Here, we present Domain-Adaptive Data Assimilation (DADA), an observation-guided framework that adapts external analyses to pretrained AI weather models. Starting from a background state, DADA optimizes only an initial-state perturbation while keeping the forecast model frozen. The perturbed state is propagated through the model, and its short-range trajectory is constrained by real-world observations through a learned observation operator. The resulting initial condition is shaped jointly by observational constraints and the dynamics learned by the target model. We evaluate DADA across five global AI weather models using backgrounds from the Global Forecast System and the AI-based HealDA. Across deterministic and probabilistic forecasts, DADA substantially reduces short-range skill loss caused by changes in the initial-condition source. More broadly, DADA turns observations into a common interface between independently developed analysis and forecasting systems, enabling pretrained AI weather models to accommodate evolving operational initial conditions without reconstructing ERA5 or retraining the forecast model.

Problem

• AI weather forecasting
• domain adaptation
• data assimilation
• initial condition mismatch
• ERA5 reanalysis

Innovation

• Domain-Adaptive Data Assimilation
• AI Weather Forecasting
• Learned Observation Operator
• Initial-State Perturbation
• Data Assimilation