arXiv Machine Learning

HaiRuo: Forecasting Ocean Eddies as Fields and Objects with a Satellite-Altimetry Foundation Model

HaiRuo is a new satellite‑altimetry foundation model that forecasts ocean mesoscale eddies both as continuous fields and as discrete objects. Using 30 days of sea‑level maps, a single forward pass produces 15 days of sea‑level and geostrophic velocity fields, daily eddy masks, and the trajectories of all eddies present today. The system combines a 1.05‑billion‑parameter pretrained component (HaiRuo‑1B) that delivers high‑skill physical fields and masks, with a Lagrangian object decoder (XuanJi) that refines eddy paths, achieving significant reductions in centre‑error and outperforming fourteen existing methods across global ocean evaluations.

arXiv Machine Learning
Sep 25

HClimRep-Ocean: A Global Ocean Emulator on an Unstructured Mesh

HClimRep‑Ocean is a machine‑learning emulator that operates directly on the native unstructured mesh of the FESOM2 ocean model, trained on a 209‑year AWI‑CM3 control run and run without atmospheric forcing except at initialization. It shows strong skill for current forecasts at 30‑day lead times, outperforming all references, while temperature and salinity forecasts are best served by a damped‑anomaly persistence approach. In independent OceanBench testing, a reanalysis‑trained variant achieves the lowest RMSE against GLORYS reanalysis, demonstrating the competitiveness of the native‑mesh approach.

By Kacper Nowak, Aleksei Koldunov, Nikolay Koldunov, Savvas Melidonis, Ankit Patnala, Simon Grasse, Julius Polz, Christian Lessig, Martin Schultz, Thomas Jung
Hugging Face Trending Papers
Jul 6

AIFS-SUBS: Extending Data-Driven Forecasting to Sub-Seasonal Timescales

Data-driven models now rival numerical weather prediction in the medium range, but extending them to sub-seasonal lead times raises challenges absent at shorter horizons. Errors accumulate over long autoregressive rollouts, systematic biases grow with lead time, and several years of data must be held out for independent verification, even though machine-learning models otherwise benefit from longer training records.

arXiv AI
Jun 3

Samudra 2: Scaling Ocean Emulators across Resolutions

arXiv:2606. 02610v1 Announce Type: cross Abstract: Ocean general circulation models (OGCMs) are essential to climate science but computationally expensive, limiting ensemble size and forcing scenarios.

By Yuan Yuan, Jesse Rusak, Alexander Merose, Adam Subel, Pavel Perezhogin, Alistair Adcroft, Carlos Fernandez-Granda, Laure Zanna
arXiv AI
Oct 19

Rescene: stable century-long climate emulation from a frozen weather model

Rescene is a 0.4 M‑parameter wrapper that transforms a frozen 1.5° weather model into a stochastic climate emulator. By anchoring the model to a calendar climatology with a deterministic slow clock and adding a band‑limited AR(1) perturbation tuned to the observed anomaly spectrum, Rescene produces century‑long runs with no detectable drift and realistic variability. A 100‑year, 8‑member experiment shows stable temperature trends, calibrated ensemble spread/skill, and accurate reproduction of day‑to‑day variability and blocking frequency.

By Minjong Cheon
arXiv AI
Sep 10

Neptune: An AI model for Global Ocean Subseasonal Prediction

Neptune is an end‑to‑end data‑driven framework that emulates global ocean and sea‑ice states for subseasonal‑to‑seasonal (S2S) forecasting up to 60 days. It combines Convolutional Neural Networks and Spherical Fourier Neural Operators to capture both local features and global cross‑scale interactions, producing daily outputs for temperature, salinity, currents, sea‑surface height, and sea‑ice metrics at 1° and 0.25° resolution. Evaluated against metrics such as RMSE, CRPS, ACC, and climate indices (ENSO, IOD), Neptune reproduces the spatio‑temporal evolution of oceanic fields and remains stable over long timescales.

By Davide Donno, Italo Epicoco, Massimo Cafaro, Gabriele Accarino, Mohammad M. Amirian, Viviana Acquaviva, Paola Nassisi, Doroteaciro Iovino, Annalisa Bracco, Simona Masina, Pierre Gentine