Institute for Coastal Research / System Analysis and Modelling 6th Meeting of the EMEP Task Force on Measurements and Modelling Zagreb, Croatia, April 4-7 2005 Approaches to Modelling HMs and POPs Distribution in the Environment Gerhard Petersen GKSS Research Centre Institute for Coastal Research / System Analysis and Modelling Max-Planck-Strasse 1 D-21502 Geesthacht GERMANY

UN-ECE Convention on Long Range Transboudary Air Pollution Protocols on Heavy Metals and Persistent Organic Pollutants The Aarhus Protocol on Persistent Organic Pollutants (POPs). Adopted in 1998, the Protocol on Persistent Organic Pollutants (POPs) entered into force in October 2003. It bans the production and use of some substances outright, while scheduling others for elimination or severe restriction at a later stage (aldrin, chlordane, dieldrin, dioxins, endrin, hexabromophenyl, furans, mirex, PAH, HCH (or lindane) toxaphenes, DDT, heptachlor, hexachlorobenzene, PCBs). 2. The Aarhus Protocol on Heavy Metals (HMs). Adopted in 1998, the Protocol on Heavy Metals entered into force in December 2003. It focuses on cadmium, lead and mercury – heavy metals particularly harmful to human health and the environment.

priority metals of concern and their typical concentrations in ambient air over Europe lead 5 - 50 ng m-3 cadmium 0.05-0.5 ng m-3 mercury Hg0 1 - 4 ng m-3 HgCl2 0.005 - 0.05 ng m-3 MeHg 0.0005 - 0.01 ng m-3 Hg(part.) 0.01 - 0.1 ng m-3

Observed Ranges of POP Concentrations in Ambient Air 1990 - 1999 in Europe and in Remote Areas (North Atlantic and Arctic) (adopted from Malanichev et al., 2002)

Main Properties of POPs affecting their Long-Range Transport Potential (adpoted from Malanichev et al., 2002) * at average scavenging and gas exchange fluxes ** at average scavenging and gas exchange fluxes and 4 m s-1 transport velocity

The Approaches to Modelling POPs (1) Multi-Compartmental Mass Balance Models for POPs The first approach to describing the distribution of POPs in the environment are relatively simple mathematical descriptions of the natural environment designed to simulate the behavior of chemicals, which are likely to be found in more than one environmental medium or phase. Such models subdivide the environment into a number of compartments – well mixed “boxes“, which are assumed to have homogeneous environmental characteristics and chemical concentrations. These models tend to involve multiple environmental compartments. have a low spatial resolution. involve simplifying assumptions of equilibrium and steady-state.

The Approaches to Modelling POPs (2) Adapting Atmospheric Dispersion Models to POPs The second approach to describing the distribution of POPs in the environment is the adaption of classical air pollution dispersion models to POPS. These models tend to be atmospheric models where the Earth‘s surface is a more or less complex boundary. have a high spatial resolution with often several thousands grid cells. be sometimes limited to simulating relatively short time periods on the scale of weeks and months, although with the increase in computing power even multi-year calculations have become possible.

The Approaches to Modelling POPs (3) Simulating low volatility POPs with atmospheric dispersion models Specifically, the group of low volatility POPs, which include the more highly chlorinated dibenzo-p-dioxins and dibenzofurans (PCDDs/Fs) are as well as the polycyclic aromatic hydrocarbons (PAHs) with more than four fused rings, are distinct, in that: they are present in the atmosphere essentially completely in the particle phase and their atmospheric deposition is dominated by particle associated deposition. their potential for volatilization from soils and vegetation is negligible. at least for the combustion derived PAHs the emission situation more closely resembles that of classical air pollutants (no secondary sources, emissions quantifiable by measurements of various combustion sources etc.). Atmospheric models, that are capable of simulating the dispersion behaviour of such particles, should be well suited for the simulation of low volatility POPs. Box models with limited spatial resolution, on the other hand, are poorly equipped to describe aerosol transport and deposition in the atmosphere.

Evolution of a global distribution model from a simple box model to a zonally averaged, two-dimensional multimedia model Wania et al., 1999 McKay et al., 1992 atmosphere terrestrial environment aquatic cold warm hot Wania and McKay , 1992 N-Polar N-Boreal N-Temperate N-Subtropic N-Tropic S-Tropic S-Subtropic S-Temperate S-Polar Wania and McKay, 1993 N-Polar N-Boreal N-Temperate N-Subtropic N-Tropic S-Tropic S-Subtropic S-Temperate S-Subpolar S-Polar Wania et al., 1999

The POPCYCLING Baltic Model aiming to quantify the pathways of POPs from the terrestrial environment to the marine environment via atmosphere and rivers (adopted from Wania: Differences, Similarities, and Complementarity of Various Approaches to Modelling Persistent Organic Pollutant Distribution in the Environment, WMO/EMEP/UNEP Workshop, Geneva, 1999 )

A B C Compartmentalisation of the terrestrial (A), marine (B) and atmospheric (C) environment of the Baltic Sea Drainage Basin in the POPCYCLING Baltic model (adopted from Wania: Differences, Similarities, and Complementarity of Various Approaches to Modelling Persistent Organic Pollutant Distribution in the Environment, WMO/EMEP/UNEP Workshop, Geneva, 1999 ) A B C

Spatial scales of atmospheric processes of heavy metals and POPs European scale. Motions of whole weather systems, on scales of hundreds to thousands of kilometers: Heavy metals (except mercury), PAHs (e.g B[a]P), PCCDs/Fs Hemispheric and Global scale. Phenomena occurring on scales exceeding 5 x 103 kilometers: Mercury, high volatility POPs (e.g. some PCB congeners, HCB)

Domain of an Eulerian chemistry and transport model used within the UNECE CLRTAP (source: EMEP Meteorological Synthesizing Centre West, Oslo, Norway)

ADOM–EUROPE for Heavy Metals model domain 3-d Eulerian grid model with a time step of 1 hour rotated spherical coordinate system of the HIgh Resolution Limited Area Weather Prediction Model (HIRLAM) with 76 by 76 grid cells ~55x55 km2 grid cell size 12 vertical layers between 1 and 10 000 m with 8 layers below 2000 m

Spatial distribution of PCB-153 deposition in theNorthern Hemisphere and in the EMEP domain units: g km-2 a-1 (adopted from EMEP MSC-E Information Note 5/2004) 1990

Spatial distribution of PCB-153 deposition in the Northern Hemisphere and in the EMEP domain units: g km-2 a-1 (adopted from EMEP MSC-E Information Note 5/2004) 2000

Outflow of POPS from the EMEP model domain (adopted from EMEP MSC-E Information Note 5/2004) 20 - 80% POP emissions COMPOUND OUTFLOW [% of emissions in the EMEP domain] B[a]P 20 PCDD/Fs 40 PCBs 50 γ-HCH 75 HCB 80

Modelling of heavy metals transport on a hemispheric scale with the GRAHM model mercury concentration in surface air January 16, 1997 (Dastoor and Laroque: Global circulation of atmospheric mercury…., Atmos. Env. 38, 147-161, 2004 ) Units: ng m-3

The Community Multi-scale Air Quality Model (CMAQ): Emissions and Meteorological Modeling Systems and the CMAQ Chemical Transport Model and Interface Processor MM5 Meteorological Modelling System SMOKE Emission Modelling System MCIP Meteorology – Chemistry Interface Processor ECIP Emissions – Chemistry Interface Processor LUPROC Land use Processor Process Analysis CMAQ Chemical Transport Model (CTM) ICON and BCON Initial and Boundary Conditions Processor Aerosol Chemistry and Dynamics Plume-in-grid Treatment Cloud Governing Equations Gas Phase Chemistry Transport Algorithms Aggregation JPROC Photolysis Rate Processor

C The Community Multi-scale Air Quality Model (CMAQ) for Europe CMAQ Europe 81x90 grid cells 54x54 km2 grid cell size CMAQ North Sea 55x58 grid cells,18x18 km2 grid cell size PM10 concentration in air, Sept 21, 2001

The Complementarity of the two Approaches Box models Box models aim to describe the big picture of a POP‘s fate in the environment with a focus on large-scale spatial patterns and long-term trends. The use of average conditions is possible and common and it is conceivable to simulate the entire lifetime of a POP in a region from initial release to final degradation, even over several centuries. Box models deliberately ignore the fine detail and have no capacity to resolve concentration differences on a small spatial or temporal scale. Conceivable = denkbar, vorstellbar

The Complementarity of the two Approaches Dispersion models Dispersion models aim to describe primarily the atmospheric transport taking into account the highly variable nature of meteorological and transport conditions. short term fluctuations on the scale of davs and even hours are resolved. Rather than aiming to describe long-term averaged conditions dispersion models are often describing events, such as particular long-range transport episodes.

Summary and Conclusions Box models are appropriate tools to simulate the entire lifetime of a POP from initial release to final degradation in more than one environmental medium and phase, even over several centuries. Current atmospheric dispersion models are well suited for the simulation of heavy metals and low volatility POPs (e.g. some of the PAHs and PCDDs/Fs) on time scales of weeks to several years. In the mid-term future, advanced atmospheric dispersion models like CMAQ incorporating an extensive treatment of aerosol chemistry and dynamics and emission processing are expected to play an increasingly important role, provided that a more detailed knowledge on HMs and POPs emissions and atmospheric processes including air/surface exchange becomes available.