Automatic detection of Arctic polynyas using hybrid supervised-unsupervised deep learning

Heuzé, Céline; Wong, Carmen Hau Man

doi:10.5194/tc-19-6043-2025

Articles | Volume 19, issue 11

https://doi.org/10.5194/tc-19-6043-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/tc-19-6043-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 19, issue 11

Research article

|

21 Nov 2025

Research article |

| 21 Nov 2025

Automatic detection of Arctic polynyas using hybrid supervised-unsupervised deep learning

Céline Heuzé and Carmen Hau Man Wong

Data sets

Sea Ice Concentrations from Nimbus-7 SMMR and DMSP SSM/I-SSMIS Passive Microwave Data (NSIDC-0051, Version 2) N. DiGirolamo et al. https://doi.org/10.5067/MPYG15WAA4WX

MIROC MIROC6 model output prepared for CMIP6 CMIP historical H. Tatebe and M. Watanabe https://doi.org/10.22033/ESGF/CMIP6.5603

Arctic and Antarctic Regional Masks for Sea Ice and Related Data Products (NSIDC-0780, Version 1) W. N. Meier and J. S. Stewart https://doi.org/10.5067/CYW3O8ZUNIWC

Short summary

Polynyas are areas with no- or thin-ice within the ice pack. They play a crucial role for the Earth system, yet their monitoring in the Arctic is challenging because polynya detection is non-trivial. We here demonstrate that polynyas can successfully be detected with a novel, machine-learning based method. In fact, we argue that they are better detected than with traditional methods, which seem to fail as sea ice decreases because of climate change.