Local and long-range influences on European mercury deposition

To evaluate and identify source-receptor relationships at local scales, and their temporal trends, a modified version of the WRF/Chem model has been developed within the GMOS project, in order to include the emissions, interactions, transport and deposition of atmospheric mercury with high spatial resolution. Anthropogenic and "natural" (marine and from wildfires) mercury emissions have been included in this model version, and gaseous and aqueous phase oxidation were implemented, while chemical initial and boundary conditions were obtained from ECHMERIT global on-line chemical transport model. Using the WRF/Chem with Hg model simulations of the atmospheric mercury cycle over Europe and the Mediterranean Basin for the year 2009 have been performed using the most recent AMAP/UNEP emissions (for the year 2010). The Mediterranean troposphere is heavily influenced by emissions originating within Europe, however, at certain periods of the year, it is also impacted by emissions from North America and even Asia. Thus model simulations were performed to seek to distinguish the local versus long-range transport contributions to mercury deposition fields. The impact of the different emission sources on deposition fluxes vary over the year and over the modelling domain: considering total mercury deposition, long-range transport of mercury is the dominant influence for half of the year from October to March, particularly over the Eastern Mediterranean basin, North Africa and Southern Spain. Local emissions make a greater contribution to mercury deposition in the modelling domain principally in hot months (from April to September), particularly over Northern and Northeastern Europe, central Europe and mountainous regions. Simulations were also performed using the previous AMAP/UNEP emissions (reference year 2005), to see whether the changes in European emissions have influenced the balance of local to non-local contributions to the mercury deposition flux. The comparison of these simulations show that although there is a large difference in anthropogenic emissions between the two reference inventories (2010 is around 30% less than 2005, over Europe), the total simulated deposition in the regions diminishes only slightly (around 10%). Moreover simulations using the 2010 inventory reproduce observations somewhat better than those using the 2005 inventory for 2009.