OASIS: Harnessing Diffusion Adversarial Network for Ocean Salinity Imputation using Sparse Drifter Trajectories

2025-08-29

AI Summary

Ocean salinity observations from drifting buoys are extremely sparse, spatiotemporally irregular, and highly noisy. Conventional interpolation methods suffer from restrictive linear or stationarity assumptions and remote-sensing limitations (e.g., cloud contamination, low revisit frequency), while existing deep learning models exhibit poor generalization under extreme sparsity and lack physically interpretable mechanisms for integrating auxiliary covariates. Method: We propose the Ocean Salinity Imputation System (OASIS), the first framework coupling diffusion modeling with adversarial training, incorporating spatiotemporal embeddings and multi-task learning to jointly leverage physical covariates—such as temperature and geographic position—without requiring dedicated sensors. A physics-informed consistency constraint is further introduced to enhance interpretability and robustness. Contribution/Results: Evaluated on real-world buoy data, OASIS significantly outperforms classical interpolation and state-of-the-art deep learning baselines, enabling high-fidelity, physically consistent global salinity field reconstruction.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Ocean salinity plays a vital role in circulation, climate, and marine ecosystems, yet its measurement is often sparse, irregular, and noisy, especially in drifter-based datasets. Traditional approaches, such as remote sensing and optimal interpolation, rely on linearity and stationarity, and are limited by cloud cover, sensor drift, and low satellite revisit rates. While machine learning models offer flexibility, they often fail under severe sparsity and lack principled ways to incorporate physical covariates without specialized sensors. In this paper, we introduce the OceAn Salinity Imputation System (OASIS), a novel diffusion adversarial framework designed to address these challenges.

Problem

• Imputing sparse ocean salinity data from drifters
• Overcoming limitations of traditional linear interpolation methods
• Incorporating physical covariates without specialized sensors

Innovation

• Diffusion adversarial network for salinity imputation
• Handles sparse irregular drifter trajectory data
• Incorporates physical covariates without specialized sensors