Accurate Mediterranean Sea forecasting via graph-based deep learning

2025/06/30 by Holmberg, Daniel, Clementi, Emanuela, Epicoco, Italo +1 · 1 citation
#Atmospheric and Oceanic Physics (physics.ao-ph) #FOS: Physical sciences

paper · doi:10.48550/arxiv.2506.23900

Abstract

Accurate ocean forecasting systems are essential for understanding marine dynamics, which play a crucial role in sectors such as shipping, aquaculture, environmental monitoring, and coastal risk management. Traditional numerical solvers, while effective, are computationally expensive and time-consuming. Recent advancements in machine learning have revolutionized weather forecasting, offering fast and energy-efficient alternatives. Building on these advancements, we introduce SeaCast, a neural network designed for high-resolution regional ocean forecasting. SeaCast employs a graph-based framework to effectively handle the complex geometry of ocean grids and integrates external forcing data tailored to the regional ocean context. Our approach is validated through experiments at a high horizontal resolution using the operational numerical forecasting system of the Mediterranean Sea, along with both numerical and data-driven atmospheric forcings. Results demonstrate that SeaCast consistently outperforms the operational model in forecast skill, marking a significant advancement in regional ocean prediction.

Citations

• Building Machine Learning Limited Area Models: Kilometer-Scale Weather Forecasting in Realistic Settings
• Diffusion-LAM: Probabilistic Limited Area Weather Forecasting with Diffusion
• Samudra: An AI Global Ocean Emulator for Climate
• Probabilistic weather forecasting with machine learning
• Continuous Ensemble Weather Forecasting with Diffusion models
• Regional data-driven weather modeling with a global stretched-grid
• Coupled Ocean-Atmosphere Dynamics in a Machine Learning Earth System Model
• Probabilistic Weather Forecasting with Hierarchical Graph Neural Networks
• AIFS -- ECMWF's data-driven forecasting system
• Data-driven Global Ocean Modeling for Seasonal to Decadal Prediction
• Building Ocean Climate Emulators
• XiHe: A Data-Driven Model for Global Ocean Eddy-Resolving Forecasting
• Scaling transformer neural networks for skillful and reliable medium-range weather forecasting
• Learning skillful medium-range global weather forecasting
• OceanNet: A principled neural operator-based digital twin for regional oceans
• Accurate medium-range global weather forecasting with 3D neural networks
• ClimaX: A foundation model for weather and climate
• FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operators
• Forecasting Global Weather with Graph Neural Networks
• Learning to Simulate Complex Physics with Graph Networks
• Interaction Networks for Learning about Objects, Relations and Physics
• Layer Normalization
• A hierarchical approach to defining marine heatwaves
• The ECMWF Ensemble Prediction System: Methodology and validation

Cited by

• MedFormer: a data-driven model for forecasting the Mediterranean Sea

Related

• Regional Ocean Forecasting with Hierarchical Graph Neural Networks
• MedFormer: a data-driven model for forecasting the Mediterranean Sea
• Deep Learning Weather Models for Subregional Ocean Forecasting: A Case Study on the Canary Current Upwelling System
• Njord: A Probabilistic Graph Neural Network for Ensemble Ocean Forecasting
• Graph-Based Deep Learning for Sea Surface Temperature Forecasts