Hernández-Macià et al. (2025) BEC SMOS Sea Ice Thickness [Dataset]

Identification

• Journal: DIGITAL.CSIC (Spanish National Research Council (CSIC))
• Year: 2025
• Date: 2025-11-27
• Authors: Hernández-Macià, Ferran, Gabarró, Carolina
• DOI: 10.20350/digitalcsic/17749

Research Groups

• Instituto de Ciencias del Mar (ICM), CSIC (Consejo Superior de Investigaciones Científicas)

Short Summary

This paper presents the BEC SMOS Sea Ice Thickness dataset, which provides sea ice thickness derived from SMOS L-band radiometry using a hybrid retrieval approach combining a physical emission model with a machine learning algorithm, primarily for thin ice up to 1 meter.

Objective

• To generate and provide a dataset of sea ice thickness retrieved from SMOS L-band measurements using a hybrid physical-machine learning approach.

Study Configuration

• Spatial Scale: Global (polar regions where sea ice is present), derived from satellite observations.
• Temporal Scale: The dataset is issued on 1 December 2025; the temporal coverage of the underlying SMOS data is not specified but implies ongoing or historical satellite observations.

Methodology and Data

• Models used: Burke emission model (Burke et al., 1979) combined with a Random Forest machine learning algorithm.
• Data sources: SMOS (Soil Moisture and Ocean Salinity) L-band radiometer measurements, specifically ESA SMOS v724 L1C brightness temperature.

Main Results

• The primary result is the "BEC SMOS Sea Ice Thickness [Dataset]", providing sea ice thickness values.
• The retrieval method is effective for thin sea ice, with a maximum retrievable thickness of 1 to 1.5 meters due to the limited sensitivity of L-band radiometry.
• For safety and trust, the current product provides reliable thickness values up to 1 meter.

Contributions

• Provides a novel dataset of sea ice thickness derived from SMOS L-band radiometry.
• Introduces a hybrid retrieval methodology that combines a physical emission model with a machine learning algorithm (Random Forest) to enhance sea ice thickness estimation.
• Addresses the challenge of accurately retrieving thin sea ice thickness, which is crucial for polar climate studies.

Funding

• Agencia Estatal de Investigación (España)
• Ministerio de Ciencia e Innovación (España)

Citation

@article{HernándezMacià2025BEC,
  author = {Hernández-Macià, Ferran and Gabarró, Carolina},
  title = {BEC SMOS Sea Ice Thickness [Dataset]},
  journal = {DIGITAL.CSIC (Spanish National Research Council (CSIC))},
  year = {2025},
  doi = {10.20350/digitalcsic/17749},
  url = {https://doi.org/10.20350/digitalcsic/17749}
}