Particulate Matter (PM1, 2.5, 10) Concentration Prediction in Ship Exhaust Gas Plume through an Artificial Neural Network

Abstract

In the last decade the reduction of carbon dioxide emissions in the transport sector, including the marine sector, has become the direction of its strategic development. Increased air pollution in the air is one of the main reasons for premature deaths around the globe. It was determined that while many methods provide adequate information about pollution levels, improvements could be made to avoid major errors. The traditional methods are either expensive or require a lot of data and human resources to correctly evaluate those data arrays. To avoid these problems, artificial neural networks (ANN) and other machine learning methods are widely used nowadays. Many ANN models for ship pollution evaluation in ports either included the whole port area or went even further and included cities near port areas. These studies show that ANNs can be effectively used to evaluate air pollution in a wide area. However, there is a lack of research on ANN usage for individual ship pollution or ship plume evaluation. This study attempts to fill this gap by developing an ANN model to evaluate an individual ship’s plumes by combining several data sources such as AIS data, meteorological data, and measured the ship’s plume pollutants concentration. Results show good correlation; however, additional limitations have to be overcome regarding data filtering and the overall accuracy of the model.

Keywords

• artificial neural networks
• air quality
• particulate matter
• shipping
• port emissions