Global/regional numerical transport and transformation of mercury in the atmosphere

During the last decade, a number of large-scale numerical regional models with varying degree of complexity have been developed for the transport and transformation of mercury in the atmosphere. However mercury unlike other heavy metals has been identified to have a long residence time of the order of one year, which makes it a global pollutant. This is due to the fact that the most significant form of mercury in the atmosphere, namely elemental mercury exists in gaseous form, it is chemically least reactive, has low solubility in water and takes part in volatilization process at the earth surface. Therefore, although very useful in analyzing episodic situations, the regional scale models developed thus far are limited in their capability in providing insights into mercury budgets, long term trends, trans-boundary exchanges and polar mercury pollution because they have to depend on prescribed background concentrations and lateral boundary fluxes of mercury. For the reasons mentioned above, global atmospheric mercury model is a more appropriate tool to address the question of mercury cycle in the atmosphere. The primary purpose of this paper is to describe development of a global mercury model with a variable-resolution grid structure which could be used to construct a regional high-resolution window for a desired region on the global domain.

The mercury transport model being developed is an Eulerian model embedded in the Canadian meteorological centre's 'global environmental multiscale model', which is the operational weather forecasting model. This has an advantage of availability of advanced parameterizations of physical processes such as cloud related processes in the model for the modeling of physio-chemical processes of mercury.

The model will be integrated in a data assimilation mode where the meteorological data is reinitialized to observed fields once a day. Multi-year simulations of mercury transport in the atmosphere will be performed. Description of the model and preliminary evaluation of results will be presented at the conference.