资源环境科技发展态势分析平台

Abstract. Shipping has a significant contribution to the emissions of air pollutants such as NO_x and particulate matter (PM), and the global maritime transport volumes are projected to increase further in the future. The Mediterranean Sea contains the major route for short sea shipping within Europe and contains the main shipping route between Europe and East Asia. Thus, it is a highly frequented shipping area, and high levels of air pollutants with significant contributions from shipping emissions are observed at monitoring stations in many cities along the Mediterranean coast.

The present study is part of the EU H2020 project SCIPPER (Shipping contribution to Inland Pollution Push for the Enforcement of Regulations). Five different regional chemistry transport models (CAMx, CHIMERE, CMAQ, EMEP, LOTOS-EUROS) were used to simulate the transport, chemical transformation and fate of atmospheric pollutants in the Mediterranean Sea for 2015. Shipping emissions were calculated with STEAM version 3.3.0, and land-based emissions were taken from the CAMS-REG v2.2.1 dataset for a domain covering the Mediterranean Sea on a resolution of 12x12 km² (or 0.1° x 0.1°). All models used their standard setup for further input. Ship contribution was calculated with the zero-out method. One run using the tagging method was performed with LOTOS-EUROS. The model outputs were compared against each other and to measured background data at monitoring stations.

The results showed differing outputs regarding the time series and pattern of model outputs but similar results with regard to the overall underestimation of NO₂ and overestimation of O₃. The contribution from ships to the total NO₂ concentration was especially high at the main shipping routes and coastal regions (25 % to 85 %). The contribution from ships to the total O₃ concentration was lowest in regions with the highest NO₂ contribution (down to -20 %). A comparison of the zero-out and tagging methods has shown that the annual mean ship contribution to the total NO₂ concentration is smaller (up to 75 %) and has a lower range when the tagging method is used. CAMx and CHIMERE simulated the highest ship contributions to the NO₂ and O₃ air concentrations. Additionally, the strongest correlation was found between CAMx and CHIMERE, which can be traced back to the usage of the same meteorological input data. The CMAQ, EMEP and LOTOS-EUROS simulated values were within one range for the NO₂ and O₃ air concentrations. Regarding deposition output, larger differences between the models were found when compared to air concentration. These uncertainties and deviations between models are caused by deposition mechanisms, which are unique within each model. A reliable output from models simulating ship contributions can be expected for air concentrations of NO₂ and O₃.