Deep Learning of Systematic Sea Ice Model Errors From Data Assimilation Increments

Abstract

Data assimilation is often viewed as a framework for correcting short‐term error growth in dynamical climate model forecasts. When viewed on the time scales of climate however, these short‐term corrections, or analysis increments, can closely mirror the systematic bias patterns of the dynamical model. In this study, we use convolutional neural networks (CNNs) to learn a mapping from model state variables to analysis increments, in order to showcase the feasibility of a data‐driven model parameterization which can predict state‐dependent model errors. We undertake this problem using an ice‐ocean data assimilation system within the Seamless system for Prediction and EArth system Research (SPEAR) model, developed at the Geophysical Fluid Dynamics Laboratory, which assimilates satellite observations of sea ice concentration every 5 days between 1982 and 2017. The CNN then takes inputs of data assimilation forecast states and tendencies, and makes predictions of the corresponding sea ice concentration increments. Specifically, the inputs are states and tendencies of sea ice concentration, sea‐surface temperature, ice velocities, ice thickness, net shortwave radiation, ice‐surface skin temperature, sea‐surface salinity, as well as a land‐sea mask. We find the CNN is able to make skillful predictions of the increments in both the Arctic and Antarctic and across all seasons, with skill that consistently exceeds that of a climatological increment prediction. This suggests that the CNN could be used to reduce sea ice biases in free‐running SPEAR simulations, either as a sea ice parameterization or an online bias correction tool for numerical sea ice forecasts.

Parameterizing Vertical Mixing Coefficients in the Ocean Surface Boundary Layer Using Neural Networks

“…Convolutional neural networks (CNNs) have been used to predict parameterizations of ocean momentum backscatter in a variety of models (Bolton & Zanna, 2019;Guillaumin & Zanna, 2021;Zanna & Bolton, 2020) and have been implemented in an ocean primitive equation model (Zhang et al, 2023). Gregory et al (2023) recently employed CNNs to learn data assimilation increments for sea-ice and showed that networks could be used to reduce biases in sea-ice.…”

“…Gregory et al. (2023) recently employed CNNs to learn data assimilation increments for sea‐ice and showed that networks could be used to reduce biases in sea‐ice.…”

Machine Learning for Online Sea Ice Bias Correction Within Global Ice‐Ocean Simulations

“…Readers are also referred to Bocquet et al (2020) and Brajard et al (2020) for seminal works in this field. In a recent sea ice study by Gregory et al (2023), hereafter G23, the authors presented a DA-based ML framework in which convolutional neural networks (CNNs) were used to predict state-dependent sea ice errors within an ice-ocean configuration of the Geophysical Fluid Dynamics Laboratory (GFDL) Seamless system for Prediction and EArth System Research (SPEAR) model, as a way to highlight the feasibility of a data-driven sea ice model parameterization within SPEAR. They approached this by first showing that the climatological sea ice concentration analysis increments from an ice-ocean DA experiment map closely onto the systematic bias patterns of the equivalent free-running model.…”

“…

In this study, we perform online sea ice bias correction within a Geophysical Fluid Dynamics Laboratory global ice-ocean model. For this, we use a convolutional neural network (CNN) which was developed in a previous study (Gregory et al, 2023, https://doi.org/10.1029/2023ms003757) for the purpose of predicting sea ice concentration (SIC) data assimilation (DA) increments. An initial implementation of the CNN shows systematic improvements in SIC biases relative to the free-running model, however large summertime errors remain.

…”

Interpretable Structural Model Error Discovery From Sparse Assimilation Increments Using Spectral Bias‐Reduced Neural Networks: A Quasi‐Geostrophic Turbulence Test Case

“…A number of recent studies have followed this approach and used ML to learn the model errors, showing promising results in various systems, from simple toy models to NWP models (Arcomano et al, 2023;Bora et al, 2023;Bretherton et al, 2022;Carrassi & Vannitsem, 2011;N. Chen & Zhang, 2023;T.-C. Chen et al, 2022;Clark et al, 2022;Farchi et al, 2021;Gregory et al, 2023;Mojgani et al, 2022;Mitchell & Carrassi, 2015;Lang et al, 2016;Pawar et al, 2020;Pathak et al, 2020;Watson, 2019;Watt-Meyer et al, 2021;Yuval et al, 2021). However, most of these studies have directly learned the model error using deep artificial neural networks (ANNs): ANNs are trained using many pairs of state and analysis increments from a training set.…”