13 Nov 2023
 | 13 Nov 2023
Status: this preprint is open for discussion.

Calibration of short-term sea ice concentration forecasts using deep learning

Cyril Palerme, Thomas Lavergne, Jozef Rusin, Arne Melsom, Julien Brajard, Are Frode Kvanum, Atle Macdonald Sørensen, Laurent Bertino, and Malte Müller

Abstract. Reliable short-term sea ice forecasts are needed to support maritime operations in polar regions. While sea ice forecasts produced by physical-based models still have limited accuracy, statistical post-processing techniques (often called calibration) can be applied to reduce forecast errors. In this study, post-processing methods based on supervised machine learning have been developed for improving the skill of sea ice concentration forecasts from the TOPAZ4 prediction system for lead times from 1 to 10 days. The deep learning models use predictors from TOPAZ4 sea ice forecasts, weather forecasts, and sea ice concentration observations. On average, the forecasts from the deep learning models have a root mean square error 41 % lower than TOPAZ4 forecasts, and 29 % lower than forecasts based on persistence of the sea ice concentration observations. They also significantly improve the forecasts for the location of the ice edges, with similar improvements as for the root mean square error. Furthermore, the impact of different type of predictors (observations, sea ice and weather forecasts) on the predictions has been evaluated. Sea ice observations are the most important type of predictors, and the weather forecasts have a much stronger impact on the predictions than sea ice forecasts.

Cyril Palerme et al.

Status: open (until 29 Dec 2023)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse

Cyril Palerme et al.


Total article views: 23 (including HTML, PDF, and XML)
HTML PDF XML Total Supplement BibTeX EndNote
20 1 2 23 2 0 0
  • HTML: 20
  • PDF: 1
  • XML: 2
  • Total: 23
  • Supplement: 2
  • BibTeX: 0
  • EndNote: 0
Views and downloads (calculated since 13 Nov 2023)
Cumulative views and downloads (calculated since 13 Nov 2023)

Viewed (geographical distribution)

Total article views: 20 (including HTML, PDF, and XML) Thereof 20 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
Latest update: 29 Nov 2023
Short summary

Sea ice forecasts are operationally produced using physical-based models, but these forecasts are often not accurate enough for maritime operations. In this study, we developed a statistical correction technique (also called calibration) using machine learning in order to improve the skill of short-term (up to 10 days) sea ice concentration forecasts produced by the TOPAZ4 model. This technique allows to reduce the errors from the TOPAZ4 sea ice concentration forecasts by 41 % on average.