Assessment of High‐Resolution Dynamical and Machine Learning Models for Prediction of Sea Ice Concentration in a Regional Application. Issue 11 (28th October 2020)

• Record Type: Journal Article
• Authors: Fritzner, Sindre; Graversen, Rune; Christensen, Kai H.
• Is Part Of: Journal of geophysical research. Volume 125:Issue 11(2020)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-10-28
• Subjects: Data Assimilation -- Sea ice -- Machine learning -- Dynamical model -- Forecast; Oceanography -- Periodicals; 551.4605
• DOI: 10.1029/2020JC016277
• Languages: English
• ISSNs: 2169-9275

Abstract

Abstract: In this study, the potential for sea ice concentration prediction using machine‐learning methods is investigated. Three different sea ice prediction models are compared: one high‐resolution dynamical assimilative model and two statistical machine‐learning models. The properties of all three models are explored, and the quality of their forecasts is compared. The dynamical model is a state‐of‐the‐art coupled ocean and sea ice ensemble‐prediction system with assimilation. The observations assimilated are high‐resolution sea ice concentration from synthetic aperture radar (SAR) and sea surface temperature from infrared instruments. The machine‐learning prediction models are a fully convolutional network and a k ‐nearest neighbors method. These methods use several variables as input for the prediction: sea ice concentration, sea surface temperature, and 2‐m air temperature. Earlier studies have applied machine‐learning approaches primarily for seasonal ice forecast. Here we focus on short‐term predictions with a length of 1–4 weeks, which are of high interest for marine operations. The goal is to predict the future state of the sea ice using the same categories as traditional ice charts. The machine‐learning forecasts were compared to persistence, which is the assumption that the sea ice does not change over the forecasting period. The machine‐learning forecasts were found to improve upon persistence in periods of substantial change. In addition, compared to the dynamical model, the k ‐nearest neighbor algorithm was found to improve upon the 7‐day forecast during a period of small sea ice variations. The fully convolutional network provided similar quality as the dynamical forecast. The study shows that there is a potential for sea ice predictions using machine‐learning methods. Plain Language Summary: This study investigates the use of statistically based models and compares them to a physically based model for sea ice prediction. The physical model uses assimilation of observations to improve the forecast. When substantial changes in the sea ice are observed, the machine‐learning models show skilful forecasts compared to assuming that the sea ice does not change during the forecasting period (persistence). A comparison between the dynamical and statistical forecast shows that the statistical model may be a simple alternative to the physical model during periods of small variations in the sea ice extent. Key Points: Both dynamical and machine‐learning methods are applied for sea ice modeling We demonstrate the potential of machine learning in sea ice forecasting The dynamical model utilizes data assimilation of high‐resolution sea‐ice concentration and sea surface temperature satellite observations