Measurements and modelling of the three-dimensional near-field dispersion of particulate matter emitted from passenger ships in a port environment

Abstract

The maritime sector poses significant challenges in controlling the emission of harmful pollutants, such as particulate matter, and reducing their impact on coastal areas and the atmospheric environment. Efforts to regulate the sector necessitate new knowledge and methods to characterise the evolution and physicochemical transformation of maritime particle emissions dispersing in port areas. An experimental campaign at the Port of Rafina, Greece, was conducted with this in mind. In this paper, we report on the first multi-characteristic particle measurements, including particle number (PN), lung deposited surface area (LDSA), and black carbon (BC), performed at both land and sea using a novel instrument set-up mounted on an unmanned aerial vehicle (UAV). Land-based measurements showed that LDSA averages, which are influenced by particle number and size, increase up to 20 times above background levels as an emission plume progresses downwind, whereas BC concentrations, which are dominated by mass, are ∼12 times higher than the background concentration. Ground and UAV-based particle comparisons showed that PN and LDSA measurements exhibit greater differences than BC relative to the plume's location. Ground-based sensors had ∼50% lower LDSA and PN concentrations, whereas BC was about equal. The experimental observations are further substantiated by coupling a Gaussian plume dispersion model and a new computationally attractive approach, known as the Incompletely Stirred Reactor Network (ISRN) method, to predict the three-dimensional evolution of particle characteristics considering the effects of dilution, segregation, and physicochemical transformations, such as coagulation. Based on simplifying assumptions for the particle size distribution sources within the port area, the ISRN estimates suggest that the discrepancy between various metrics can be partly explained by coagulation, responsible for a non-linear increase in particle size, depending on the local level of dilution and mixing intensity, leading up to a ∼25% decrease in PN and LDSA compared to BC. Combined, measurements and modelling highlight the effect of the sampling location and the importance of monitoring more than one particle metric to characterise particle evolution.