Physics-informed deep-learning parameterization of ocean vertical mixing improves climate simulations

TLDR

Uncertainties in ocean‑mixing parameterizations are primary sources of ocean and climate modeling biases, especially in the tropics where physics‑driven approaches perform poorly due to limited process understanding. The study aims to develop a data‑driven parameterization of ocean vertical mixing using deep learning and long‑term turbulence observations. An artificial neural network was trained on a decadal‑long record of tropical Pacific hydrographic and turbulence data, incorporating physical constraints to guide learning. The resulting parameterization outperforms existing schemes, generalizes well, and improves ocean‑only and coupled climate simulations when integrated into an ocean model.

Abstract

Uncertainties in ocean-mixing parameterizations are primary sources for ocean and climate modeling biases. Due to lack of process understanding, traditional physics-driven parameterizations perform unsatisfactorily in the tropics. Recent advances in the deep-learning method and the new availability of long-term turbulence measurements provide an opportunity to explore data-driven approaches to parameterizing oceanic vertical-mixing processes. Here, we describe a novel parameterization based on an artificial neural network trained using a decadal-long time record of hydrographic and turbulence observations in the tropical Pacific. This data-driven parameterization achieves higher accuracy than current parameterizations, demonstrating good generalization ability under physical constraints. When integrated into an ocean model, our parameterization facilitates improved simulations in both ocean-only and coupled modeling. As a novel application of machine learning to the geophysical fluid, these results show the feasibility of using limited observations and well-understood physical constraints to construct a physics-informed deep-learning parameterization for improved climate simulations.