Calibration of a neural network ocean closure for improved mean state and variability

2026-04-07

AI Summary

This study addresses the significant biases in mean state and variability exhibited by coarse-resolution global ocean models due to their inability to resolve mesoscale eddies, a limitation exacerbated by the reliance of traditional parameterizations on manual tuning. The authors propose a novel approach that, for the first time, applies ensemble Kalman inversion (EKI) to systematically calibrate neural-network-based parameterizations. Their method introduces an efficient optimization protocol that avoids the need to integrate the model to statistical equilibrium, thereby effectively mitigating statistical noise arising from chaotic dynamics. Using an idealized model and time-averaged statistics, the approach reduces simulation errors in both the mean state and variability of the fluid interface by approximately 50%, substantially outperforming simulations without parameterization or those employing offline-trained neural networks.

Abstract

Global ocean models exhibit biases in the mean state and variability, particularly at coarse resolution, where mesoscale eddies are unresolved. To address these biases, parameterization coefficients are typically tuned ad hoc. Here, we formulate parameter tuning as a calibration problem using Ensemble Kalman Inversion (EKI). We optimize parameters of a neural network parameterization of mesoscale eddies in two idealized ocean models at coarse resolution. The calibrated parameterization reduces errors in the time-averaged fluid interfaces and their variability by approximately a factor of two compared to the unparameterized model or the offline-trained parameterization. The EKI method is robust to noise in time-averaged statistics arising from chaotic ocean dynamics. Furthermore, we propose an efficient calibration protocol that bypasses integration to statistical equilibrium by carefully choosing an initial condition. These results demonstrate that systematic calibration can substantially improve coarse-resolution ocean simulations and provide a practical pathway for reducing biases in global ocean models.

Problem

• ocean modeling
• model bias
• mesoscale eddies
• parameterization
• calibration

Innovation

• Ensemble Kalman Inversion
• neural network parameterization
• mesoscale eddies
• model calibration
• coarse-resolution ocean modeling