Intercomparison Study of Numerical Models for Long-Range Atmospheric Transport of Mercury. Stage III. Comparison of Modelling Results with Long-term Observations and Comparison of Calculated Items of Regional Balances.

EMEP/MSC-E Technical Report 1/2005

A.Ryaboshapko, R.Artz, R.Bullock, J.Christensen, M.Cohen, R.Draxler, I.Ilyin, J.Munthe, J.Pacyna, G.Petersen, D.Syrakov, O.Travnikov

ABSTRACT

Recently it has been widely recognized that mercury is a hazardous heavy metal, which possesses unique characteristics in the environment. It enters the atmosphere from both anthropogenic and natural sources. It migrates through the environment in different physical-chemical states. It is chemically and biochemically active in soils, water bodies and in the atmosphere. Finally, mercury in the forms of its organic compounds can be accumulated it trophic chains. In these forms it is the most dangerous for human and animals whose diet is connected with fish and other fresh water or oceanic food.

Taking into account high jeopardy of mercury and its high mobility in the environment the 1979 Convention on Long-Range Transboundary Air Pollution included this metal into its consideration. The Protocol on Heavy Metals, which came into forth on December 2004, determines mercury as on of three metals of the highest priority. A number of investigations have shown that anthropogenic activity of recent centuries led to long-term accumulation of mercury in all main geospheres - the atmosphere, the pedosphere (soils), and the Ocean. To assess global biogeochemical aspects of mercury environmental cycle and pollution UNEP launched the "Global Mercury Assessment" project.

Complicated mercury behaviour in the environment and diversity of its form make study of this metal especially difficult. In such cases numerical modelling becomes a powerful tool of investigations. One should also bear in mind that routine monitoring of mercury is very expensive, and therefore density of monitoring networks cannot be high. All these facts were recognized by a number of researchers and decision-makers. The Protocol on Heavy Metals considers the modelling approach as a basis for assessment of mercury transboundary pollution in Europe. It determined the main task for the European Monitoring and Evaluation Program (EMEP) in this field as follows: "EMEP shall, using appropriate models …., provide to the Executive Body for the Convention calculations of transboundary fluxes and depositions of heavy metals within the geographical scope of EMEP". Meteorological Synthesizing Centre "East" (MSC-E) of EMEP has a responsibility to perform model calculations of transboundary transport and deposition of heavy metals over Europe.

Many numerical models of different types have been developed to evaluate mercury atmospheric transport and deposition on local, regional and global levels. They are widely used as purely scientific instruments or as applied methods to solve problems of local or national levels. It is understandable that the models can differ from each other depending on their complexities and their tasks. Natural questions arise: Can the models produce reliable results? How far are the modelling results from the available observations? To what extent do different models agree with each other?

Recognizing the importance of these questions the Steering Body of EMEP at its 18th session decided (EB.AIR/GE.1/24, 1994) that MSC-E had to organize an intercomparison study of atmospheric long-range transport models for heavy metals. Such a study is considered by the Steering Body to be one of the essential prerequisites for development and application of EMEP operational models. Intercomparison studies have already been completed for lead [Sofiev et al., 1996] and for cadmium [Gusev et al., 2000]. Besides, model intercomparison exercises were initiated by World Meteorological Organization for sulphur compounds on regional and global scale [Rasch et al., 2000; Barrie et al., 2001].

The mercury model intercomparison study was launched in 1999. The mercury intercomparison study is focused on:
- an evaluation of parameterisations of the main physical-chemical processes of mercury transformations in the gaseous and the liquid phase;
- a comparison of modelling results with measurements obtained from both short-term campaigns and from the EMEP monitoring network and other international and national programs;
- a comparison of the main features of long-range transport of different mercury forms.

Taking into account significant complexity of mercury models, necessity to consider main modelling processes separately it was decided to divide the program of the mercury model intercomparison study into three stages:

Stage I. Comparison of modules for physico-chemical transformations of mercury species in a cloud/fog environment with prescribed initial mercury concentrations in ambient air and other physical and chemical parameters relevant for atmospheric mercury transformations.

Stage II. Comparison of model results with observations during 1-2 weeks episodes. Hourly and daily averages and event based averages of mercury concentrations in air, obtained from the joint Swedish/Canadian/German field campaign TRANSECT 1995 and from the European Union Environment & Climate project Mercury Species Over Europe (MOE-1999) have been used.

Stage III. Comparison of model results with observed monthly and annual means of mercury concentrations in air and precipitation and deposition fluxes available from European monitoring stations in 1999. Comparison of model predicted atmospheric budgets of mercury species in the entire EMEP domain and for selected European countries (UK, Poland and Italy), including dry and wet deposition from sources within and outside the area of the countries.

The first stage of the intercomparison study was started in 1999 and finished in 2001. Five scientific groups from Germany, Sweden, the USA and MSC-East took part in the study. The results were presented in a MSC-East technical report [Ryaboshapko et al., 2001] and published in scientific literature [Ryaboshapko et al., 2002a].

The second stage (started in 2001) is focused on the comparison of modelled and observational results. The observations were performed at five measurement sites in Europe during two short-term campaigns in 1995 and 1999. In the first case main attention was paid to mercury elemental form. In 1999 reactive gaseous and aerosol mercury were measured in addition to the elemental form. Seven scientific groups involved in atmospheric mercury modelling participated in the second stage. They represent the most advanced scientific and operational mercury models of regional and global types. The results of the second stage were analysed and discussed in MSC-East reports [Ryaboshapko et al., 2002b; 2003]. A corresponding scientific article is under preparation.

The last stage of the mercury model comparison combined performances of seven models of atmospheric mercury transport and deposition of regional and global levels. They are:

- GKSS-Forschungszentrum Geesthacht GmbH (Germany), the European mercury version of the Acid Deposition and Oxidants Model (ADOM).
- U.S. Environmental Protection Agency (USA), the Community Multi-Scale Air Quality (CMAQ) model.
- National Ocean and Atmosphere Administration (USA), Hybrid Single Particle Lagrangian Integrated Trajectory model, version 4 for mercury (HYSPLIT)
- National Institute of Meteorology and Hydrology (Bulgaria), Eulerian Model for Air Pollution (EMAP)
- National Environmental Research Institute (Denmark), Danish Eulerian Hemispheric Model (DEHM)
- EMEP Meteorological Synthesizing Centre-East, MSCE Heavy Metal model, Regional version (MSCE-HM)
- EMEP Meteorological Synthesizing Centre-East, MSCE Heavy Metal model, Global version (MSCE-HM-Hem)

At this stage the EMEP / MSC-East was presented by two model versions: regional one - for EMEP domain, and global one - for the Northern Hemisphere.

The program of the third stage of the study met to the greatest degree the requirements of EMEP. It was focused on answering two important questions:
- To what extent can the current atmospheric mercury models of regional and global levels reproduce annual and monthly mean values of mercury concentrations and depositions observed at monitoring stations in Central Europe?
- What is accuracy of model assessment of items of mercury atmospheric balances for individual European countries?

To discuss the program of the third stage, the volume of calculations, methods of their statistical processing, and formats of reporting data two workshops was organised in Moscow in April 2003 and in April 2004 (MSC-East). All details were agreed and accepted as a working plan. Some first results of the last stage and the study as a whole were presented by the participants at the International Conference "Mercury as a Global Pollutant" (Slovenia, Ljubljana, 2004).