Citation

Chen J, Bian W, Wan Z, Wang S, Zheng H, Cheng C. Int. J. Disaster Risk Reduct. 2020; 47: e101560.

Copyright

(Copyright © 2020, Elsevier Publishing)

DOI

10.1016/j.ijdrr.2020.101560

PMID

unavailable

Abstract

Total-loss marine accident, the most serious type of ship mishap, not only incurs huge economic losses to shipping companies, but is also usually accompanied by casualties and ecological damage. It is therefore crucial to identify the relationship between the main factors and casualties of total-loss marine accidents for the purpose of reducing such incidents and the casualty rate to maintain maritime safety. This paper collates the data of total-loss marine accidents of ships of 100 gross tonnage or above between 1998-2018 based on the Lloyd's List Intelligence Casualty Statistics. The casualties have been divided into three grades for research purpose, establishing a partial proportional odds logistic model. The model shows that foundering and collision of vessels play a significant part in causing casualties in total-loss marine accidents. On the other hand, in hull damage and stranding accidents, crew members have a high chance of survival. While setting up an ordered logistic model, a problem arose - independent variables wouldn't pass the parallel line test. This paper uses the proportional odds model in an innovative way to improve the logistic model application in the field of marine accidents.

Language: en

Keywords

Casualties; Maritime safety; Partial proportional odds model; Total-loss marine accidents