Connecting inland ship emissions from on-board sampling, onshore measurements and large-eddy simulations

We present the results of a comparison between inland ship exhaust emissions derived from onshore and on-board measurements, and numerical simulations. Our aim is to discuss the applicability and potential limitations of commonly used approaches to determine ship-related air pollutant emissions. For this purpose, we made a push boat pass a measurement station on an inland canal repeatedly under similar driving conditions. We observed a large variability in the detected pollutant concentration peaks with regard to their amplitude, shape and time of occurrence, both in measurement and model. We trace this back to randomly distributed turbulence, which strongly affects the plume dispersion close to the emission source. Consequently, it is hardly possible to reliably derive an emission rate based on a single ship plume measurement onshore. In contrast, an emission factor can be reliably deduced from individual measurements, as revealed by the good agreement between onshore and on-board measurement data. • Measurements and simulations reveal a large variability in ship-related air pollutant concentrations observed onshore. • Plume dispersion close to the emission source is strongly affected by randomly distributed turbulence. • Single ship plume measurements can be used to calculate emission factors, but this method is not reliable for calculating emission rates.