A Comprehensive Model for Quantifying, Predicting, and Evaluating Ship Emissions in Port Areas Using Novel Metrics and Machine Learning Methods

Seaports, as major transportation hubs, generate significant air pollution due to intensive ship traffic, directly affecting local air quality. While emission inventories are commonly used to manage ship-based air pollution, they reflect only the emission-related aspect of a specified period and are...

Full description

Saved in:
Bibliographic Details
Main Authors: Filip Bojić, Anita Gudelj, Rino Bošnjak
Format: Article
Language:English
Published: MDPI AG 2025-06-01
Series:Journal of Marine Science and Engineering
Subjects:
Online Access:https://www.mdpi.com/2077-1312/13/6/1162
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850168047990472704
author Filip Bojić
Anita Gudelj
Rino Bošnjak
author_facet Filip Bojić
Anita Gudelj
Rino Bošnjak
author_sort Filip Bojić
collection DOAJ
description Seaports, as major transportation hubs, generate significant air pollution due to intensive ship traffic, directly affecting local air quality. While emission inventories are commonly used to manage ship-based air pollution, they reflect only the emission-related aspect of a specified period and area, limiting the broader interpretability and comparability of the results. To overcome the mentioned challenges, this research presents the PrE-PARE model, which enables the prediction, analysis, and risk evaluation of ship-sourced air pollution in port areas. The model comprises three interconnected modules. The first is applied for quantifying emissions using detailed technical and movement datasets, which are combined into individual voyage trajectories to enable a high-resolution analysis of ship-based air pollutants. In the second module, the Multivariate Adaptive Regression Splines (MARS) machine learning method is adapted to predict emissions in varying operational scenarios. In the third module, novel metric methods are introduced, enabling a standardised efficiency comparison between ships. These methods are supported by a unique classification system to determine the emission risk in different periods, evaluate the intensity of various ship types, and rank individual ships based on their operational efficiency and emission optimisation potential. By combining new methods with technical and operational shipping data, the model provides a transparent, comparable, and adaptable system for emissions monitoring. The results demonstrate that the PrE-PARE model has the potential to improve strategic planning and air quality management in ports while remaining flexible enough to be applied in different contexts and future scenarios.
format Article
id doaj-art-d8637f472ea14bc284c2999261acadfb
institution OA Journals
issn 2077-1312
language English
publishDate 2025-06-01
publisher MDPI AG
record_format Article
series Journal of Marine Science and Engineering
spelling doaj-art-d8637f472ea14bc284c2999261acadfb2025-08-20T02:21:03ZengMDPI AGJournal of Marine Science and Engineering2077-13122025-06-01136116210.3390/jmse13061162A Comprehensive Model for Quantifying, Predicting, and Evaluating Ship Emissions in Port Areas Using Novel Metrics and Machine Learning MethodsFilip Bojić0Anita Gudelj1Rino Bošnjak2Faculty of Maritime Studies, University of Split, 21000 Split, CroatiaFaculty of Maritime Studies, University of Split, 21000 Split, CroatiaFaculty of Maritime Studies, University of Split, 21000 Split, CroatiaSeaports, as major transportation hubs, generate significant air pollution due to intensive ship traffic, directly affecting local air quality. While emission inventories are commonly used to manage ship-based air pollution, they reflect only the emission-related aspect of a specified period and area, limiting the broader interpretability and comparability of the results. To overcome the mentioned challenges, this research presents the PrE-PARE model, which enables the prediction, analysis, and risk evaluation of ship-sourced air pollution in port areas. The model comprises three interconnected modules. The first is applied for quantifying emissions using detailed technical and movement datasets, which are combined into individual voyage trajectories to enable a high-resolution analysis of ship-based air pollutants. In the second module, the Multivariate Adaptive Regression Splines (MARS) machine learning method is adapted to predict emissions in varying operational scenarios. In the third module, novel metric methods are introduced, enabling a standardised efficiency comparison between ships. These methods are supported by a unique classification system to determine the emission risk in different periods, evaluate the intensity of various ship types, and rank individual ships based on their operational efficiency and emission optimisation potential. By combining new methods with technical and operational shipping data, the model provides a transparent, comparable, and adaptable system for emissions monitoring. The results demonstrate that the PrE-PARE model has the potential to improve strategic planning and air quality management in ports while remaining flexible enough to be applied in different contexts and future scenarios.https://www.mdpi.com/2077-1312/13/6/1162sustainable shippingair pollutionmetric systemmachine learningrisk assessment
spellingShingle Filip Bojić
Anita Gudelj
Rino Bošnjak
A Comprehensive Model for Quantifying, Predicting, and Evaluating Ship Emissions in Port Areas Using Novel Metrics and Machine Learning Methods
Journal of Marine Science and Engineering
sustainable shipping
air pollution
metric system
machine learning
risk assessment
title A Comprehensive Model for Quantifying, Predicting, and Evaluating Ship Emissions in Port Areas Using Novel Metrics and Machine Learning Methods
title_full A Comprehensive Model for Quantifying, Predicting, and Evaluating Ship Emissions in Port Areas Using Novel Metrics and Machine Learning Methods
title_fullStr A Comprehensive Model for Quantifying, Predicting, and Evaluating Ship Emissions in Port Areas Using Novel Metrics and Machine Learning Methods
title_full_unstemmed A Comprehensive Model for Quantifying, Predicting, and Evaluating Ship Emissions in Port Areas Using Novel Metrics and Machine Learning Methods
title_short A Comprehensive Model for Quantifying, Predicting, and Evaluating Ship Emissions in Port Areas Using Novel Metrics and Machine Learning Methods
title_sort comprehensive model for quantifying predicting and evaluating ship emissions in port areas using novel metrics and machine learning methods
topic sustainable shipping
air pollution
metric system
machine learning
risk assessment
url https://www.mdpi.com/2077-1312/13/6/1162
work_keys_str_mv AT filipbojic acomprehensivemodelforquantifyingpredictingandevaluatingshipemissionsinportareasusingnovelmetricsandmachinelearningmethods
AT anitagudelj acomprehensivemodelforquantifyingpredictingandevaluatingshipemissionsinportareasusingnovelmetricsandmachinelearningmethods
AT rinobosnjak acomprehensivemodelforquantifyingpredictingandevaluatingshipemissionsinportareasusingnovelmetricsandmachinelearningmethods
AT filipbojic comprehensivemodelforquantifyingpredictingandevaluatingshipemissionsinportareasusingnovelmetricsandmachinelearningmethods
AT anitagudelj comprehensivemodelforquantifyingpredictingandevaluatingshipemissionsinportareasusingnovelmetricsandmachinelearningmethods
AT rinobosnjak comprehensivemodelforquantifyingpredictingandevaluatingshipemissionsinportareasusingnovelmetricsandmachinelearningmethods