FlowDA: Accurate, Low-Latency Weather Data Assimilation via Flow Matching

TL;DR

FlowDA introduces a flow matching-based generative data assimilation framework that significantly improves weather prediction accuracy and efficiency, especially under low observation rates and noisy conditions, while maintaining robustness over long-term forecasts.

Contribution

FlowDA is a novel, low-latency generative data assimilation method using flow matching, fine-tuned on a foundation model for scalable and robust weather analysis.

Findings

Outperforms strong baselines at low observation rates

Robust to observational noise

Maintains stable long-horizon auto-regressive performance

Abstract

Data assimilation (DA) is a fundamental component of modern weather prediction, yet it remains a major computational bottleneck in machine learning (ML)-based forecasting pipelines due to reliance on traditional variational methods. Recent generative ML-based DA methods offer a promising alternative but typically require many sampling steps and suffer from error accumulation under long-horizon auto-regressive rollouts with cycling assimilation. We propose FlowDA, a low-latency weather-scale generative DA framework based on flow matching. FlowDA conditions on observations through a SetConv-based embedding and fine-tunes the Aurora foundation model to deliver accurate, efficient, and robust analyses. Experiments across observation rates decreasing from $3.9\%$ to $0.1\%$ demonstrate superior performance of FlowDA over strong baselines with similar tunable-parameter size. FlowDA further shows robustness to observational noise and stable performance in long-horizon auto-regressive cycling DA. Overall, FlowDA points to an efficient and scalable direction for data-driven DA.