Fang et al. (2025) IMPMCT: a dataset of Integrated Multi-source Polar Mesoscale Cyclone Tracks in the Nordic Seas

Identification

• Journal: Earth system science data
• Year: 2025
• Date: 2025-11-12
• Authors: Runzhuo Fang, Jinfeng Ding, Wenjuan Gao, Xi Liang, Zhuoqi Chen, Chuanfeng Zhao, Haijin Dai, Liu L
• DOI: 10.5194/essd-17-6049-2025

Research Groups

• College of Meteorology and Oceanography, National University of Defense Technology, Changsha, China
• Key Laboratory of High Impact Weather (special), China Meteorological Administration, Changsha, China
• Key Laboratory of Marine Hazards Forecasting, National Marine Environmental Forecasting Center, Ministry of Natural Resources, Beijing, China
• School of Geospatial Engineering and Science, Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Sun Yat-sen University, Zhuhai, China
• Department of Atmospheric and Oceanic Sciences, School of Physics, and China Meteorological Administration Tornado Key Laboratory, Peking University, Beijing, China

Short Summary

The Integrated Multi-source Polar Mesoscale Cyclone Tracks (IMPMCT) dataset provides a comprehensive 24-year (2001–2024) record of wintertime Polar Mesoscale Cyclone (PMC) tracks in the Nordic Seas, integrating ERA5 reanalysis, AVHRR infrared imagery, and scatterometer wind data. This dataset, containing 1110 tracks, 16 001 cloud features, and 4472 wind records, aims to advance the understanding of PMC genesis and intensification mechanisms.

Objective

• To establish a comprehensive, long-term (2001–2024) dataset of Polar Mesoscale Cyclone (PMC) tracks in the Nordic Seas by integrating reanalysis data, satellite infrared imagery, and scatterometer wind measurements, providing multi-dimensional attributes for atmospheric and oceanic research.

Study Configuration

• Spatial Scale: Nordic Seas (Greenland, Norwegian, and Barents Seas), covering 50–85° N latitude and 40° W–80° E longitude. Vortex-Centered Infrared (VCI) images cover a 1600 km × 1600 km domain.
• Temporal Scale: 24 years (2001–2024), specifically the extended winter seasons (November–April). ERA5 data is hourly, while satellite data has irregular temporal intervals.

Methodology and Data

• Models used:
  • Objective mesoscale vortices-tracking algorithm (adapted from Shimizu and Uyeda, 2012) applied to 850 hPa relative vorticity fields.
  • Deep learning-based method: YOLOv8-obb-pose model for automated detection and extraction of cyclonic cloud features (type, center coordinates, oriented bounding box).
  • Gaussian filtering for smoothing sea-level pressure (SLP) fields.
• Data sources:
  • Reanalysis: ERA5 (European Centre for Medium-Range Weather Forecasts Reanalysis) for 850 hPa relative vorticity, sea-level pressure, and wind fields (0.25° × 0.25° spatial grid, hourly).
  • Satellite observation:
    • Advanced Very High-Resolution Radiometer (AVHRR) (from NOAA and MetOp satellites) Level 1B GAC (approximately 4 km resolution) and LAC (1.1 km resolution) forth-band infrared imagery.
    • Advanced Scatterometer (ASCAT) (from MetOp satellites) Level 2 near-surface wind vector retrieval products (12.5 km resolution).
    • Quick Scatterometer (QuikSCAT) (NASA) Level 2 near-surface wind vector retrieval products (12.5 km resolution).

Main Results

• The IMPMCT dataset contains 1110 validated cyclone-related vortex tracks, 16 001 distinct cyclonic cloud features, and 4472 instances of measurable near-surface wind patterns, with 794 tracks exhibiting maximum wind speeds exceeding 15 m/s.
• Validation against the Stoll (2022) dataset showed that 90% of matched vortex points were within 50 km, with mean absolute differences of 1.11 × 10⁻⁵ s⁻¹ for relative vorticity, 0.43 hPa for sea-level pressure, and 22.79 km for vortex equivalent diameter.
• Comparison with manual Polar Low (PL) lists (Noer et al., 2011; Rojo et al., 2015, 2019) demonstrated improved matching rates using more lenient vortex identification thresholds, suggesting the capture of weaker PLs.
• Approximately 95% of matched cyclone-vortex track pairs showed average matching distances below 100 km, indicating strong consistency between cloud features and reanalysis vortices.
• The dataset's tracks largely satisfy established PL criteria: 88.4% meet the polar-front criterion (track-averaged tropopause potential temperature < 300.8 K), 90% meet the mesoscale criterion (maximum vortex diameter < 430 km), and 84% fulfill the cyclonic intensity criterion (maximum pressure anomaly > 0.4 hPa).

Contributions

• Provides the first comprehensive, long-term (24-year) multi-source dataset for all Polar Mesoscale Cyclones (PMCs) in the Nordic Seas, expanding beyond the previous focus on Polar Lows (PLs).
• Integrates reanalysis data with remote sensing observations to overcome limitations of single-source datasets, offering more complete cyclone life cycle trajectories, intuitive cloud imagery visualization, and a richer set of parameters.
• Enhances detection sensitivity for weaker PMCs and captures their full lifecycle evolution, from genesis to dissipation.
• Serves as a critical benchmark for evaluating high-latitude numerical weather prediction models and a unique case library for comparative studies of PMC/PL formation mechanisms, intensity thresholds, and sea-ice interaction dynamics.
• Facilitates automated identification of model-undetected systems through advanced deep learning frameworks, enabling systematic evaluation of model representation fidelity for PMCs/PLs.

Funding

• National Key R&D Program of China (grant no. 2021YFC2802501)
• National Science Foundation of China (grant no. 42476205)

Citation

@article{Fang2025IMPMCT,
  author = {Fang, Runzhuo and Ding, Jinfeng and Gao, Wenjuan and Liang, Xi and Chen, Zhuoqi and Zhao, Chuanfeng and Dai, Haijin and L, Liu},
  title = {IMPMCT: a dataset of Integrated Multi-source Polar Mesoscale Cyclone Tracks in the Nordic Seas},
  journal = {Earth system science data},
  year = {2025},
  doi = {10.5194/essd-17-6049-2025},
  url = {https://doi.org/10.5194/essd-17-6049-2025}
}