CORTEXA
← Browse
crossrefJournal of Geophysical Research: Machine Learning and Computation2026-06-27Cited by 0

Data‐Driven Emulation of Numerically Simulated Baltic Sea Surface Currents With a Deep Convolutional U‐Net: Explainability and Potential Forecast Skill

Amirhossein Barzandeh, Christoph Manss, Frederic Stahl, Ilja Maljutenko, Sander Rikka, Urmas Raudsepp

Abstract Ocean models can represent surface circulation at kilometer scales, but their computational cost limits broad experimentation. We present DeepCUN, a deep convolutional encoder–decoder (U‐Net) that emulates daily mean Baltic Sea surface current components on a 1‐nautical‐mile grid. DeepCUN is trained on 2015–2023 reanalysis currents with atmospheric reanalysis forcing. An occlusion‐sensitivity channel ablation indicates that wind provides the dominant predictive information when combined with the antecedent surface‐current state, while additional atmospheric variables contribute negligibly. Accordingly, DeepCUN is configured to use only the prior surface‐current state and the subsequent‐day wind components to predict next‐day surface current fields. On an independent 2024 test year, DeepCUN captures the dominant spatial patterns and temporal variability, with the highest errors concentrated in the southwestern boundary exchange corridor and narrow straits and lower errors across the basin interior, where correlations exceed 0.9 in most locations. For post hoc explainability of this data‐driven mapping, we apply layer‐wise relevance propagation to characterize attributed support and a diagonal Jacobian elasticity metric to quantify local responsiveness, revealing spatially varying reliance on state memory in the southwestern boundary‐influenced region and wind‐modulated adjustment across the basin interior. In addition, to assess the potential forecast skill of this approach, we perform recursive rollouts, which show domain‐mean absolute error increasing smoothly from at 1 day to at 21 days while correlation decreases from to , indicating stable multi‐week behavior under idealized forcing conditions. The framework provides an efficient and interpretable template for configuring, pruning, and stress‐testing reduced‐input emulators of regional surface circulation.

View free PDFSource page

Related papers

crossrefJournal of Geophysical Research: Machine Learning and Computation2026-07-18

Surrogate‐Assisted Bayesian Inference of Fracture Network Parameters From Elastic Waves: A Sensitivity‐Guided Approach

Le Zhang, Qinghua Lei, Longjun Dong, Chuanyin Jiang, Thomas Hermans

Abstract We develop a sensitivity‐guided, surrogate‐assisted Bayesian framework to infer fracture network parameters from elastic waves. Synthetic fracture networks characterized by power‐law length exponent , fracture density , and percolation parameter are constructed. Elastic…

View free PDFSource page
crossrefJournal of Geophysical Research: Machine Learning and Computation2026-07-11

Spatially Aware Calibration of NWP and AI Precipitation Forecasts

Belinda Trotta, Esteban Abellan

Abstract Rainfall is often highly localized and its location is difficult to predict exactly with a numerical weather prediction (NWP) model. Previous research has shown that this problem can be mitigated by spatially aware calibration methods which incorporate forecast informati…

View free PDFSource page
crossrefJournal of Geophysical Research: Machine Learning and Computation2026-07-09

Glacier Mass Balance Modeling Using a Long Short‐Term Memory Network

Marijn van der Meer, Harry Zekollari, Alban Gossard, Kamilla Hauknes Sjursen, Jordi Bolibar, Matthias Huss, et al.

Abstract Glacier mass balance (MB) is a key indicator of climate change and a central driver of glacier evolution, yet most glaciers worldwide lack long‐term in situ measurements. For estimating glacier MB, data‐driven models provide a complementary alternative to traditional num…

View free PDFSource page
crossrefJournal of Geophysical Research: Machine Learning and Computation2026-07-04

Toward Generative Machine Learning for Boosting Ensembles of Climate Simulations

Parsa Gooya, Reinel Sospedra‐Alfonso, Johannes Exenberger

Abstract Accurately quantifying uncertainty in predictions and projections arising from irreducible internal climate variability is critical for decision‐making. Such uncertainty is typically assessed using ensembles produced with climate models. However, computational constraint…

View free PDFSource page
crossrefJournal of Geophysical Research: Machine Learning and Computation2026-07-01

Joint Satellite SST and Dynamic SSS as Key Constraints on the Thermodynamics of Tropical Instability Wave Variability in the Eastern Equatorial Pacific

Yinfei Zhou, Haoyu Wang, Xiaofeng Li

Abstract Tropical instability waves (TIWs) generate mesoscale sea surface temperature (SST) fluctuations in the eastern equatorial Pacific and influence the evolution of the El Niño‐Southern Oscillation (ENSO). Yet satellite‐based prediction of TIW‐related SST anomalies remains l…

View free PDFSource page
crossrefJournal of Geophysical Research: Machine Learning and Computation2026-06-29

Super‐Resolution of Planetary Images Based on Generative Adversarial Network

Xiaoran Zhang, Yiran Wang, Miao Zhuo

Abstract Currently, satellite imagery serves as the primary means of observing terrestrial planets such as the Mars, the Moon, and Mercury. Enhancing the resolution and quality of these images can provide more detailed insights into planetary surfaces. However, improvements in im…

View free PDFSource page