MMSeaIce: a collection of techniques for improving sea ice mapping with a multi-task model
Open Access
- 8 April 2024
- journal article
- research article
- Published by Copernicus GmbH in The Cryosphere
- Vol. 18 (4) , 1621-1632
Abstract
The AutoICE challenge, organized by multiple national and international agencies, seeks to advance the development of near-real-time sea ice products with improved spatial resolution, broader spatial and temporal coverage, and enhanced consistency. In this paper, we present a detailed description of our solutions and experimental results for the challenge. We have implemented an automated sea ice mapping pipeline based on a multi-task U-Net architecture, capable of predicting sea ice concentration (SIC), stage of development (SOD), and floe size (FLOE). The AI4Arctic dataset, which includes synthetic aperture radar (SAR) imagery, ancillary data, and ice-chart-derived label maps, is utilized for model training and evaluation. Among the submissions from over 30 teams worldwide, our team achieved the highest combined score of 86.3 %, as well as the highest scores on SIC (92.0 %) and SOD (88.6 %). Notably, the result analysis and ablation studies demonstrate that instead of model architecture design, a collection of strategies/techniques we employed led to substantial enhancement in accuracy, efficiency, and robustness within the realm of deep-learning-based sea ice mapping. Those techniques include input SAR variable downscaling, input feature selection, spatial–temporal encoding, and the choice of loss functions. By highlighting the various techniques employed and their impacts, we aim to underscore the scientific advancements achieved in our methodology.Keywords
All Related Versions
Funding Information
- Environment and Climate Change Canada (n/a)
- Natural Sciences and Engineering Research Council of Canada (n/a)
This publication has 40 references indexed in Scilit:
- AI4SeaIce: selecting loss functions for automated SAR sea ice concentration chartingScientific Reports, 2023
- Light from space illuminating the polar silk roadInternational Journal of Digital Earth, 2022
- C- and L-band SAR signatures of Arctic sea ice during freeze-upRemote Sensing of Environment, 2022
- Spatio-temporal distribution of sea-ice thickness using a machine learning approach with Google Earth Engine and Sentinel-1 GRD dataRemote Sensing of Environment, 2022
- Swin Transformer: Hierarchical Vision Transformer using Shifted Windows2021
- Evaluation of a Neural Network on Sea Ice Concentration Estimation in MIZ Using Passive Microwave Data2021
- Half a century of satellite remote sensing of sea-surface temperatureRemote Sensing of Environment, 2019
- Deep learning based retrieval algorithm for Arctic sea ice concentration from AMSR2 passive microwave and MODIS optical dataRemote Sensing of Environment, 2019
- Focal Loss for Dense Object Detection2017
- U-Net: Convolutional Networks for Biomedical Image Segmentation2015