Generative Lagrangian data assimilation for ocean dynamics under extreme sparsity

TL;DR

This paper introduces a deep learning framework combining neural operators and diffusion models to accurately reconstruct high-resolution ocean states from extremely sparse Lagrangian observations, addressing a key challenge in ocean data assimilation.

Contribution

It presents a novel generative data assimilation method that effectively captures small-scale ocean dynamics under extreme sparsity, outperforming existing deep learning approaches.

Findings

Accurately reconstructs ocean states at 99-99.9% sparsity levels.

Demonstrates robustness on synthetic and real satellite data.

Outperforms baseline models in capturing mesoscale turbulence.

Abstract

Reconstructing ocean dynamics from observational data is fundamentally limited by the sparse, irregular, and Lagrangian nature of spatial sampling, particularly in subsurface and remote regions. This sparsity poses significant challenges for forecasting key phenomena such as eddy shedding and rogue waves. Traditional data assimilation methods and deep learning models often struggle to recover mesoscale turbulence under such constraints. We leverage a deep learning framework that combines neural operators with denoising diffusion probabilistic models (DDPMs) to reconstruct high-resolution ocean states from extremely sparse Lagrangian observations. By conditioning the generative model on neural operator outputs, the framework accurately captures small-scale, high-wavenumber dynamics even at $99\%$ sparsity (for synthetic data) and $99.9\%$ sparsity (for real satellite observations). We validate our method on benchmark systems, synthetic float observations, and real satellite data, demonstrating robust performance under severe spatial sampling limitations as compared to other deep learning baselines.