Skip to main content

ORIGINAL RESEARCH article

Front. Mar. Sci.
Sec. Ocean Observation
Volume 11 - 2024 | doi: 10.3389/fmars.2024.1536363
This article is part of the Research Topic Remote Sensing Applications in Oceanography with Deep Learning View all 6 articles

Machine Learning-Based Analysis of Sea Fog's Spatial and Temporal Impact on Near-Miss Ship Collisions Using Remote Sensing and AIS Data

Provisionally accepted
Zhe Zeng Zhe Zeng 1*Dan Liu Dan Liu 1Ling Ke Ling Ke 2*Shuo Zhang Shuo Zhang 1*Shanwei Liu Shanwei Liu 1
  • 1 China University of Petroleum, Qingdao, China
  • 2 National Satellite Meteorological Center (NSMC), Beijing, China

The final, formatted version of the article will be published soon.

    Sea fog is a severe marine environmental disaster that significantly threatens the safety of maritime transportation. It is a major environmental factor contributing to ship collisions. The Himawari-8 satellite's remote sensing capabilities effectively bridge the spatial and temporal gaps in data from traditional meteorological stations for sea fog detection. Therefore, the study of the influence of sea fog on ship collisions becomes feasible and is highly significant. To investigate the spatial and temporal effects of sea fog on vessel near-miss collisions, this paper proposes a general-purpose framework for analyzing the spatial and temporal correlations between satellite-derived large-scale sea fog using a machine learning model and the near-miss collisions detected by the automatic identification system through the Vessel Conflict Ranking Operator. First, sea fog-sensitive bands from the Himawari-8 satellite, combined with the Normalized Difference Snow Index (NDSI), are chosen as features, and an SVM model is employed for sea fog detection. Second, the geographically weighted regression model investigates spatial variations in the correlation between sea fog and near-miss collisions. Third, we perform the analysis for monthly time series data to investigate the within-year seasonal dynamics and fluctuations. The proposed framework is implemented in a case study using the Bohai Sea as an example. It shows that in large harbor areas with high ship density (such as Tangshan Port and Tianjin Port), sea fog contributes significantly to near-miss collisions, with local regression coefficients greater than 0.4. While its impact is less severe in the central Bohai Sea due to the open waters. Temporally, the contribution of sea fog to near-miss collisions is more pronounced in fall and winter, while it is lowest in summer. This study sheds light on how the spatial and temporal patterns of sea fog, derived from satellite remote sensing data, contribute to the risk of near-miss collisions, which may help in navigational decisions to reduce the risk of ship collisions.

    Keywords: Himawari-8 satellite data, Sea fog, Near miss, Ship collision, spatio-temporal pattern

    Received: 28 Nov 2024; Accepted: 30 Dec 2024.

    Copyright: © 2024 Zeng, Liu, Ke, Zhang and Liu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

    * Correspondence:
    Zhe Zeng, China University of Petroleum, Qingdao, China
    Ling Ke, National Satellite Meteorological Center (NSMC), Beijing, 100081, China
    Shuo Zhang, China University of Petroleum, Qingdao, China

    Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.