SIXTH FRAMEWORK PROGRAMME
Project contract no. 502527
ESPREME
Estimation of willingness-to-pay to reduce risks of exposure to heavy metals
and cost-benefit analysis for reducing heavy metals occurrence in Europe
Specific Targeted Research Project
Research priority 1.6. Assessment of environmental technologies for support of policy decisions, in
particular concerning effective but low-cost technologies in the context of fulfilling environmental
legislation
Workpackage 04
Atmospheric modelling of HM concentrations
Start date of project: 1st of January 2004
Duration: 36 months
Lead authors for this deliverable: Oleg Travnikov and Ilia Ilyin (MSC-E)
Project co-funded by the European Commission within the Sixth Framework Programme (20022006)
PU
PP
RE
CO
Dissemination Level
Public
Restricted to other programme participants (including the Commission Services)
Restricted to a group specified by the consortium (including the Commission Services)
Confidential, only for members of the consortium (including the Commission Services)
MODELLING OF HEAVY METALS ATMOSPHERIC
DISPERSION IN EUROPE
Oleg Travnikov and Ilia Ilyin
Meteorological Synthesizing Centre – East of EMEP
May 2007
CONTENTS
1. INTRODUCTION
3
2. HEAVY METAL CHEMICAL TRANSPORT MODEL MSCE-HM
3
2.1. Brief model description
4
2.2. Emissions data
7
3. HEAVY METAL AIRBORNE POLLUTION LEVELS IN EUROPE
IN THE BASE YEAR 2000
17
3.1. Air concentration and deposition fields and model evaluation
17
3.2. Uncertainties
27
4. REDUCTION OF HEAVY METAL ATMOSPHERIC DEPOSITIONS
IN EUROPE BY 2010
28
5. SENSITIVITY OF HEAVY METAL DEPOSITION IN EUROPE
TO EMISSION REDUCTION IN EUROPEAN COUNTRIES
31
6. CONCLUSIONS
35
7. REFERENCES
36
2
1.
INTRODUCTION
The atmospheric transport can be considered as the most dynamic pathway of heavy metal
dispersion in the environment. Such heavy metals and metalloids as Cd, Cr, Hg, Ni, Pb, and
As are emitted to the atmosphere from a variety of industrial and mobile sources, dispersed
over hundreds of kilometres, and deposited to the ground contaminating remote ecosystems
or contributing to adverse exposure of human health.
The work package WP04 of the project was aimed at evaluation of the atmospheric
dispersion of the selected heavy metals (As, Cd, Cr, Hg, Ni, Pb) in Europe and assessment of
air concentration and deposition fields to support other activities within the whole chain of
the project. Particularly, the main objectives of the work package included:



Review and improvement of heavy metal chemical transport model (MSCE-HM);
Calculation of concentration and deposition fields of As, Cd, Cr, Hg, Ni, and Pb for
the base year 2000 and 2010 scenarios (BAU, MFTR);
Calculation of source-receptor reduction matrixes of the selected heavy metals for
2010 scenarios.
This report contains description of activities undertaken by MSC-E within WP04 in
accordance with the work-plan and overview of the major results obtained.
In particular, Chapter 2 contains short description of the chemical transport model MSCEHM applied for the atmospheric dispersion modelling along with review of major
modifications and improvements performed within the framework of the project. In Chapter 3
analysis of the modelling results for the base year 2000 is presented together with evaluation
of the modelling results against measurements. Chapter 4 presents calculations of deposition
reduction of the selected heavy metals by 2010 according to emission scenarios (BAU,
MFTR) prepared in WP02. Chapter 5 illustrates calculations of the source-receptor reduction
matrixes indicating sensitivity of heavy metal pollution levels in different parts of Europe
with regard to emission reduction in European countries.
2.
HEAVY METAL CHEMICAL TRANSPORT MODEL MSCE-HM
The European-scale atmospheric transport model MSCE-HM is actively used for operational
calculations of heavy metal transboundary pollution within the European region, in
connection with the EMEP programme and other activities relating to the LTRAP
Convention. Detailed description of the model is available in (Travnikov and Ilyin, 2005).
The model formulation and performance was thoroughly evaluated within the EMEP/TFMM
Workshop on the model review (ECE/EB.AIR/GE.1/2006/4). A brief model description is
presented below. Besides, the main model improvements performed within the framework of
the project included:





Refinement and evaluation of input data (meteorological pre-processor, landuse etc.);
Improvement of dry and wet deposition schemes;
Refinement of chemical scheme of Hg transformations in the atmosphere;
Development of model parameterization for As, Cr, Ni;
Development of wind re-suspension scheme for heavy metals.
The last point (wind re-suspension scheme) is considered in more details because of
particular importance of this process for evaluation of heavy metal pollution levels.
3
2.1
Brief model description
The EMEP/MSC-E regional model of heavy metals airborne pollution (MSCE-HM) is a
three-dimensional Eulerian type chemical transport model driven by off-line meteorological
data. The model considers heavy metal emissions from anthropogenic and natural sources,
transport in the atmosphere, chemical transformations (of mercury only) both in gaseous and
aqueous phases, and deposition to the ground. The computation domain of the model is
defined on the polar stereographic projection and covers European region along with adjacent
territories. The model operates in the regular EMEP grid, which covers the area from
approximately 35°W to 60°E and from the North Pole to about 20°N, and includes Europe,
partly the North Atlantic and the Arctic oceans, Northern Africa, and part of Middle East (see
Figure 2.1). The vertical structure of the model is formulated in the sigma-pressure
coordinate system. The model domain consists of 15 irregular sigma-layers and has a top at
100 hPa. The layers are confined by surfaces of constant  and do not intersect the ground
topography.
Figure 0.1. The EMEP 50×50 km2 grid
The atmospheric advection and the vertical transport are described in the model using mass
conservative and monotone Bott’s advection scheme with fourth-order area-preserving
polynomials (Bott, 1989a; 1989b, 1992). An implicit treatment of the vertical eddy diffusion
is chosen in order to avoid restrictions on the integration time step because of possible sharp
gradients of the pollutant mixing ratio.
Such heavy metals as lead and cadmium and their compounds are characterized by very low
volatility. It is assumed in the model that these metals (as well as some others – nickel,
chromium, zinc etc.) are transported in the atmosphere only in the composition of aerosol
particles. It is believed that their possible chemical transformations do not change properties
of their particles-carriers with regard to removal processes. On the contrary, mercury
transformations in the atmosphere include transitions between the gaseous, aqueous and solid
phases, chemical reactions in the gaseous and aqueous environment. The chemical scheme of
4
mercury transformation in the atmosphere is based on the kinetic mechanism developed by
Petersen et al. (1998). The original scheme was modified neglecting too slow reactions and
replacing fast ones by appropriate equilibriums in order to accelerate the model performance
for operational calculations. The physical and chemical transformations of mercury include
dissolution in cloud droplets, gas-phase and aqueous-phase oxidation by ozone, chlorine and
hydroxyl radical, aqueous-phase reduction via decomposition of sulphite complexes,
formation of chloride complexes, and adsorption by soot particles in cloud water. Summary
of all chemical transformations included to the model is presented in Table 2.1.
Table 2.1. Summary of mercury transformations included into the model
Hg (0gas )  O3(gas )  Hg (II)( part )  products
Reactions and equilibriums
2.110-18exp(-1247/T)
Units
Reference
cm3/(molec∙s) Hall, 1995
Hg(0gas ) OH(gas )  Hg(II )( part )  products
8.710-14
cm3/(molec∙s) Sommar et al., 2001
Hg(0gas )  Cl2(gas )  Hg(II )(gas )
2.610-18
cm3/(molec∙s) Ariya et al., 2002

Hg(0aq )  O3(aq )  Hg(2aq
)  products
4.7∙107
M-1s-1
Hg(0aq ) OH(aq )
2.4∙10

 Hg(2aq
)
 products
Hg(0aq )

 Cl(I )(aq )  Hg(2aq
)

Hg(2aq
)
 2SO32(aq )
Hg(SO3 )22(aq )
k or H
 products
 Hg(SO3 )22(aq )
 Hg(0aq )
9
2∙106
1.1∙10 [SO2(gas)] ·10
-21
4.4∙10
 products
f([Cl ])
HgnClm(dis )  HgnClm(soot )
0.2
Hg(SO3 )22 (dis )
0.2
Hg(0gas )
-
 Hg(SO3 )22 (soot )
 Hg(0aq )
OH(gas ) OH(aq )

1.82  10  7
M-1s-1
Lin and Pehkonen, 1999
-1
Petersen et al., 1998
-1
s
Petersen et al., 1998
1
Lurie, 1971
1
Petersen et al., 1998
s
Petersen et al., 1998
3
3.41·10-21Texp(17.72(T0/T-1)) M·cm3/molec Jacobson, 1999
Cl2(gas )  Cl(I )(aq )

[Hg (2aq
)]
Gårdfeldt et al., 2001
M·cm /molec Andersson et al., 2004
Ryaboshapko et al.,
1.75·10-16Texp(18.75(T0/T-1)) M·cm3/molec
2001
-24
3
1.58·10 Texp(7.8(T0/T-1))
M·cm /molec Sander, 1997

- [SO2(gas)] is in ppbv
- [Hg nCl m(dis ) ]
[Cl (aq ) ]
M s
1
1.76·10 Texp(9.08(T0/T-1))
O3(gas )  O3(aq )
**
4pH *
**
-23
HgCl2(gas )  HgCl2(aq )
*
2
-4

HgnClm(dis )  Hg(2aq
)
Munthe, 1992
-1 -1
4.48·10-16

[Cl (aq ) ]2
6.03  10 14

[Cl (aq ) ]3
8.51  10 15

M·cm3/molec Lin and Pehkonen, 1999
[Cl (aq ) ]4
8.51  10 16
Model description of removal processes includes dry deposition and wet scavenging. The dry
deposition scheme is based on the resistance analogy approach and allows taking into account
deposition to different land cover types (forests, grassland, water surface etc.). Dry deposition
of particles to vegetation is described using the theoretical formulation from Slinn (1982) and
fitted to experimental data (Ruijgrok et al., 1997; Wesely et al., 1985). The parameterization
of dry deposition to water surfaces is based on the approach suggested by Williams (1982)
taking into account the effects of wave breaking and aerosol washout by seawater spray.
Parameters of dry deposition of gaseous oxidised mercury are chosen as for those of nitric
acid (zero surface resistance) because of similar solubility of these substances. Besides, dry
deposition of gaseous elemental mercury to vegetation is considered. The deposition velocity
of this form varies from 0 to 0.03 cm/s depending on surface temperature, solar radiation and
vegetation type. The model distinguishes in-cloud and sub-cloud wet scavenging of
particulate mercury and highly soluble reactive gaseous mercury. Based on empirical data
from literature survey the wet deposition coefficient is taken as a function of precipitation
rate as 3·10-4 R0.8 and 1·10-4 R0.7 1/s for in-cloud and below-cloud scavenging, respectively,
5
where R is precipitation rate in mm/h. Besides, the precipitation rate is scaled for convective
precipitation according to Walton et al. (1988) to take into account fractional coverage of a
grid cell with precipitating clouds.
Regional modelling of heavy metal airborne pollution requires setting appropriate boundary
condition at lateral and upper boundaries of the computation domain. These conditions are
aimed at taking into account influence of emission sources located outside the domain. It is
particularly important for Hg because of long residence time of this pollutant in the
atmosphere. Contribution of the intercontinental transport to deposition of mercury in Europe
is comparable (up to 40%) with contribution of regional sources (Ilyin et al., 2003).
Therefore, setting the proper boundary concentrations of mercury species at the domain
boundaries is principally important for assessment of realistic pollution levels. Available
measurement data are too sparse to supply concentrations of mercury along the whole model
boundary. To set the boundary conditions adequately we use modeling results obtained with
the hemispheric model MSCE-HM-Hem (Travnikov and Ryaboshapko, 2002). Monthly mean
concentrations of three atmospheric mercury forms – GEM, TPM, RGM – calculated at the
EMEP domain boundaries are used as the regional model boundary conditions. Figure 2.2
illustrates distribution of gaseous elemental mercury concentration in the vicinity of the
model boundary.
Figure 2.2. Distribution of gaseous elemental mercury in the Northern Hemisphere. Rectangular
shows the EMEP domain
Boundary conditions are not so critical for As, Cd, Cr, Ni, and Pb modelling as for Hg
because of significantly shorter residence time of these pollutants in the atmosphere. Besides,
hemispheric modelling of these pollutants is not feasible because of lack of up-to-date
emissions data at the hemispheric or global scale. Therefore prescribed boundary
concentrations were used for As, Cd, Cr, Ni, and Pb.
MSCE-HM model is driven by off-line meteorological data pre-processed by MM5 - Fifth
Generation Penn State/NCAR Mesoscale Model (Grell et al., 1995). The pre-processor
utilizes NCAR/NCEP re-analysis as the input information and provides 6-hour weather
prediction data along with estimates of the atmospheric boundary layer parameters with the
same spatial resolution as that of the transport model. Meteorological fields for 2000 were
used for the base year calculations.
6
Calculations of pollution scenarios for future years (2010) require selection of meteorological
data to be used in the model. There are two aspects governing such selection process. Firstly,
the chosen meteorological dataset should as far as possible reflect average “climatological”
conditions to avoid the effect of meteorological variability on the modelling results.
Secondly, it impossible to use just average meteorological fields since the atmospheric
transport modelling requires the fields to be strongly coupled in terms of equations of fluid
mechanics. Therefore, it is more preferable to average modelling results obtained for several
years instead of averaging the meteorological data itself.
To investigate the effect of meteorological variability on the modelling results long-term
calculations of atmospheric dispersion of some heavy metals have been performed for the
period 1990-2003. Identical input parameters have been used for each year of the period
except meteorological data. Heavy metal concentration in air and precipitation and total
deposition were considered as main model output. Obtained results were averaged over the
whole period and compared with annual means of each individual year. Besides, the overall
average was also compared with averages over several years.
Figure 2.3 presents an example of this analysis for Pb. Only annual and two-years average
results with smallest deviation are presented in the figure. Analysis for other metals has very
similar character. As seen from Fig. 2.3a, relative deviation of the model output parameters
for individual years from the average over the period generally exceeds 15%. On the other
hand, averaging even over two years leads to twofold reduction (down to 6-7%) of the
deviation of modelling results due to meteorological variability. Besides, as seen from the
figure there is a number of two-years combinations leading to similar levels of the deviation.
Therefore it was decided to apply two-years averaging of all modelling results (over 1998 and
2001) for 2010 scenarios calculations.
35
15
10
0
a
1999
1994
1992
2002
1990
1991
1993
1998
1995
0
b
1991,1995
5
1997,1999
5
20
1995,1998
10
Total dep
25
1992,2002
15
Conc in prec
1992,1999
20
1997,1998
25
Conc in air
30
1995,2002
Total dep
1992,1998
Conc in prec
1998,2001
Conc in air
30
Relative deviation, %
Relative deviation, %
35
Figure 2.3. Estimated deviation of Pb model output parameters due to the inter-annual meteorological
variability for individual years (a) and average over two years (b). Bars present median deviation over
the model domain; ranges show 90% confidence intervals
2.2
Emissions data
Emissions data are among the most important input parameters for atmospheric dispersion
modeling. Data on anthropogenic and natural emissions of heavy metals utilized in the
modeling are described below. Particular attention is paid to parameterization of wind resuspension of particle-bound heavy metals from soil and seawater developed within the
framework of the project (WP04).
7
2.2.1 Anthropogenic emissions
Spatially and temporally resolved anthropogenic emissions dataset has been prepared by the
WP02 of the project for the dispersion modelling purposes. The dataset consists of daily
gridded emissions of As, Cd, Cr, Hg, Ni, and Pb with spatial resolution 50x50 km for the
base year 2000 and two emission scenarios for 2010 (BAU+Climate and MFTR). Emission
fluxes in each grid cell are disaggregated into three classes (h<50, 50<h<150, h>150 m)
according to height of emission sources. Besides, emissions of Hg include separate data for
three chemical species: gaseous elemental Hg (GEM), total particulate Hg (TPM), and
reactive gaseous Hg (RGM).
Spatial distribution of annual emission flux of all six heavy metals from anthropogenic
sources in 2000 is illustrated in Fig. 2.4. Major areas with elevated emissions coincide for all
the metals: high emission fluxes are in Belgium, Netherlands, Western Germany, Southern
Poland, Northern Italy, Eastern part of Ukraine, and Western part of Russia. However, there
are some differences in spatial distributions of various metals. Particularly, some metals are
characterized by more even distribution of emission fluxes (e.g. Cr, Ni), whereas others have
pronounced “hot spots” of high emissions over relatively low background (As, Cd, Pb).
a
b
c
d
e
f
Figure 2.4. Spatial distribution of As (a), Cd (b), Cr (c), Hg (d), Ni (e), and Pb (f) anthropogenic
emissions in Europe in 2000
8
2.2.2 Wind re-suspension of particle-bound heavy metals
Wind re-suspension of particle-bound heavy metals (such as As, Cd, Cr, Ni, Pb) from soil
and seawater appears to be important process affecting ambient concentration and deposition
of these pollutants, particularly, in areas with low direct anthropogenic emissions. Recent
data demonstrate significant decrease of heavy emissions in Europe during last decade.
However, long-term historic emissions of these metals during previous centuries from
industrial processes, fuel combustion, road transport etc. followed by atmospheric dispersion
and deposition to the ground resulted in their accumulation in the surface soils all over
Europe. Aeolian erosion of bare soils or human disturbed areas leads to suspension of dust
particles containing heavy metals into the atmosphere.
In mineral dust production models the process of wind erosion and suspension of dust aerosol
from the ground is commonly parameterized as combination of two major processes: saltation
and sandblasting (e.g. Gomes et al., 2003; Zender et al., 2003; Gong et al., 2003). The first
process (saltation) presents horizontal movement of large soil aggregates driven by wind
stress. Indeed, in natural soils small particles (below 20 m) never occur in free state, but are
embedded in larger soil aggregates by cohesion forces (up to a few centimeters). These
aggregates are too heavy to be directly suspended by wind in usual conditions. Instead, they
are moved by wind stress close to the surface jumping from one place to another. When the
saltating aggregates impact the ground they can eject much smaller particles (few
micrometers), which can be easily suspended by wind and transported far away from the
source region. This process is called the sandblasting.
The saltation process is characterized by the critical wind stress value, over which movement
of soil particles can be initiated. This critical wind stress can be described by the threshold
wind friction velocity, which depends on the soil particle size, soil wetness, and protection of
the erodible soil by roughness elements (drag partitioning). In order to characterize this
threshold friction velocity (Ut*) in the model we used a simplified empirically based
parameterization proposed by Marticorena and Bergametti (1995):
0.129
 *
,
 Ut 
(1.928 Re 0.092  1)0.5

U t*  0.129 1  0.858 exp 0.0617 Re  10  ,
where  
Ds 
6  10 7 
  s g 
,
a 
Ds5 / 2 
Re  10
(2.1)
Re  10
Re  1.755 Ds1.56  0.38 .
Here Ds is the soil particle size, a and s are air and soil mass densities, respectively; g is the
gravity acceleration.
A parameterization of the threshold friction velocity dependence on soil wetness was
proposed by Fécan et al. (1999) based on empirical data. According to this work the
threshold is a function of volumetric soil moisture and clay content in soil. Taking into
account that soil moisture data produced by the meteorological pre-processor (MM5) can
contain significant uncertainties and require transformation from gravimetric to volumetric
values, we have chosen to use a simplified approach suggested by Grini et al. (2005). It is
based on rainfall events and implements the following assumptions:

The dust production is stopped if precipitation during the last 24 hours exceeds 0.5
mm.
9


The period without the dust production (in days) is equal to precipitation amount (in
mm) during the last 24 hours.
The dust production is resumed if no rain has fallen in the last 5 days.
A phenomenological drag partition scheme was also proposed by Marticorena and
Bergametti (1995). It modifies the threshold friction velocity in order to take into account the
effect of surface roughness elements hampering the transfer of wind momentum to the
erodible surface. The scheme uses the soil roughness length (Z0) as a surrogate for the
presence of these roughness elements. However, the large-scale roughness lengths commonly
used in atmospheric transport models are deduced the standard deviation of the topography or
from the vegetation height and cannot adequately represent the small-scale roughness to
describe a surface process like aeolian erosion (Marticorena and Bergametti, 1997). For this
reason drag partition parameterization was not included to the current version of the resuspension model.
Ones the wind friction velocity exceeds the threshold value, the vertically integrated sizeresolved saltation flux is given in the following form (Gomes et al., 2003):
Fh (Ds ) 
Ka *
U  Ut* U *  Ut* 2 .
g
(2.2)
The constant K in this expression reflects possibility of the limitation of soil aggregates
supply because of depletion of loose material on the surface. Following Gomes et al. (2003)
we adopted K = 1 for deserts and K = 0.02 for other erodible surfaces.
In general the integrated saltation flux strongly depends on size distribution of soil aggregates
occurring in natural conditions. The continuous multi-modal distribution can be presented by
a combination of lognormal functions:
dM
1

dDs Ds 2
 ln D  ln D 

j lnj j exp  s2 ln2  js, j

2

,


(2.3)
where Ds, j is mass median diameter of the jth mode; and j is its geometric standard
deviation.
The sandblasting model for dust suspension was developed by Alfaro et al. (1997; 1998).
Based on the wind tunnel experiments they derived that the aerosol particles released by
sandblasting from the saltating aggregates of different natural soils can be sorted into three
lognormal populations. Characteristics of the dust populations are presented in Table 2.2.
Table 2.2. Characteristics of tree dust aerosol particle populations released by the sandblasting
(Alfaro and Gomes, 2001)
Mode
1
2
3
ei, kg m2/s2
3.61·10-7
3.52·10-7
3.46·10-7
i
di, m
1.5
6.7
14.2
1.7
1.6
1.5
According to the sandblasting model the vertical dust flux can be presented in the following
form (Alfaro and Gomes, 2001):
10
Fv ,i 
dM
 F (D ) (D ) dD
h
s
i
s
Ds
i  1, 3 ,
dDs ,
(2.4)
s
where the efficiency of the sandblasting process i is given by:
 i (Ds ) 

6
s 
pi d i3
.
ei
(2.5)
Here  = 163 m/s2 is an empirical constant; pi is the fraction of kinetic energy of a soil
aggregate required to release aerosol particles of mode i; di is the aerosol mass median
diameter of mode i; and ei is binding energy of aerosol particles for mode i.
The fraction pi of the aerosol modes release depends on the kinetic energy of an individual
soil aggregate
ec 
100
 s Ds3U *2 .
3
(2.6)
A scheme of the dependence is presented in Table 2.3. The binding energies ei correspond to
values presented in Table 2.2.
Table 2.3. Fractions (pi) of the dust aerosol modes release as a function of the kinetic energy ec of an
individual soil aggregate
Mode 1 (p1)
Mode 2 (p2)
Mode 3 (p3)
ec<e3
0
0
0
e3<ec<e2
0
0
1
e2<ec<e1
0
(ec-e2)/(ec-e3)
1-p2
e1<ec
(ec-e1)/(ec-e3)
(1-p1)(ec-e2)/(ec-e3)
1-p1-p2
An example of calculated relative fractions of the dust aerosol modes as a function of soil
aggregates size are shown in Fig. 2.5a for given wind friction velocity. As seen from the
figure sandblasting of larger soil aggregates, which have higher kinetic energy, releases
smaller dust particles. Figure 2.5b presents an example of size distribution of the vertical dust
flux for given wind friction velocity and size distribution of saltating soil aggregates. As seen
the Mode 1 corresponds mostly to fine particles (below 2 m), where as two other modes
present coarse particles (5-20 m).
0.08
Total
mode 3
mode 2
mode 1
-1
-1
0.8
Mode 1
Mode 2
Mode 3
0.4
0.2
180
200
0.06
-2
U* = 0.5 m/s
0.6
0.0
160
a
dFv/dDp, kg m s Dp
Modes fractions (pi )
1.0
0.04
0.02
0.00
220
Ds, m
U* = 0.5 m/s
0.1
b
1
10
100
Dp, m
Figure 2.5. Fractions of dust aerosol modes as functions of soil aggregates size (a) and size
distribution of vertical dust flux (b)
11
In general the size distribution of the dust suspension flux strongly depends on the size
distribution of soil aggregates. As it was mentioned above, at the moment there is no spatially
resolved Europe-wide dataset applicable to characterize the properties of erodible soils in
Europe. However, in some cases one can neglect distinctions between different soil types.
Indeed, Figure 2.6 shows the dust suspension flux as a function of the wind friction velocity
for different dust particle modes and soil populations from Table 2.1. As seen from the figure
the vertical dust flux of the coarse Modes 2 and 3 strongly depends on the soil populations
(Figs.2.6b and c). The difference can reach two orders of magnitude. On the other hand, for
the fine mode, which is the most important from the point of view of the long-range
atmospheric transport, the vertical dust flux only slightly depends on the soil type (Fig. 2.6a).
One can hardly expect that dust particles of the coarse modes can significantly contribute to
the long-range transport of heavy metals in Europe. Therefore, in the first approximation it is
possible to restrict further consideration only by the fine particles mode and neglect
distinctions between different soil populations.
10
0.01
0.1
0.01
0.001
0.4
0.6
0.8
1.0
0.001
0.2
0.4
U*, m/s
a
0.1
0.01
0.001
0.2
Mode 3
Sa
CS
FS
ASS
1
2
2
0.1
Mode 2
Sa
CS
FS
ASS
1
Fv, kg/m /s
2
Fv, mg/m /s
1
10
Mode 1
Sa
CS
FS
ASS
Fv, kg/m /s
10
0.6
0.8
U*, m/s
b
1.0
0.2
0.4
0.6
0.8
1.0
U*, m/s
c
Figure 2.6. Vertical dust flux as a function of the wind friction velocity for different soil populations
and dust particle modes: (a) – mode 1; (b) – mode 2; (c) – mode 3
Implementing the discussed above assumptions we performed calculations of the dust
suspension flux in Europe and adjacent territories in 2000. The dust suspension was estimated
for the following types of land cover:



deserts and bare soils;
agricultural soils (during the cultivation period);
urban areas.
To estimate particle-bound heavy metal emission with dust suspension from soil it is
necessary to know content of these metals in erodible soils. For this purpose we used detailed
measurement data on heavy metals concentration in topsoil from the Geochemical Atlas of
Europe developed under the auspices of the Forum of European Geological Surveys
(FOREGS) [www.gtk.fi/publ/foregsatlas/]. The data cover most parts of Europe (excluding
Eastern European countries) with more than 2000 measurement sites. The kriging
interpolation was applied to obtain spatial distribution of heavy metal concentration in soil.
For Eastern Europe as well as for the rest of the model domain (Africa, Asia) we used default
concentration values based on the literature data (Table 2.4).
Table 2.4. Default concentrations of heavy metals in soil
Metal
As
Cd
Cr
Ni
Pb
Soil concentration, mg/kg
5
0.2
50
15
15
Reference
Beyer & Cromartie, 1987
Nriagu, 1980a
Shacklette et al., 1970
Nriagu, 1980b
Reimann and Cariat, 1998
12
The resulting spatial distributions of Pb, Cd, As, Cr, and Ni concentration in topsoil of
Europe and adjusted territories are presented in Figure 2.7. These concentrations reflect both
natural content of these metals in the Earth’s crust and accumulation of anthropogenic
depositions during long-term period of human industrial activity.
a
b
c
d
e
Figure 2.7. Spatial distribution of heavy metal concentration in topsoil of Europe and adjacent
territories: (a) – Pb; (b) – Cd; (c) – As; (d) – Cr; (e) - Ni
Description of sea-salt generation and wind suspension from the sea surface has also been
included into the model. For this purpose we applied the empirical Gong-Monahan
parameterization of the vertical number flux density (Gong, 2003):
2
dFn
 1.373U103.41R p A (1  0.057 R p3.45 ) 10 1.6 exp(  B ) ,
dR p
where A  4.7(1  Rp )0.017Rp
1.44
,
B  1
log R p
0.433
(2.7)
.
Here Rp is the sea-salt aerosol radius; U10 is wind speed at 10 m height; and  = 30 is an
adjustable parameter that controls the shape of the sub-micron size distribution.
The size distribution of the vertical sea-salt aerosol mass flux based on this parameterization
is shown in Fig. 2.7a for different wind speeds. Figure 2.7b illustrates dependence of the
integral sea-salt aerosol flux on wind speed at 10 m height for different cut-off aerosol
diameters. In the following calculations we used the cut-off value of aerosol diameter equal
to 10 m, since larger particles can hardly be transported far from the ocean coastal areas.
13
1.E-06
U = 5 m/s
U = 10 m/s
U = 15 m/s
1.E-07
2
1.E-09
Fm, kg/m /s
-2 -1
dFm/dDp, kg m s m
-1
1.E-08
1.E-10
1.E-11
1.E-09
Dmax = 10 um
Dmax = 20 um
Dmax = 30 um
1.E-10
1.E-12
1.E-11
1.E-13
0.1
a
1.E-08
1
0
10
Dp, m
b
10
20
30
40
50
U10, m/s
Figure 2.7. Sea-salt aerosol mass flux as a function of particle size (a) and dependence of the sea-salt
flux on wind speed at 10 m height for different cut-off aerosol diameters (b)
In order to estimate suspension with sea-salt aerosol we used the emission factors derived
from the literature (Table 2.5). These emission factors depend on measured heavy metal bulk
concentration in seawater and on the estimated enrichment factor of the heavy metal in seasalt particles comparing to its bulk concentration in seawater. The sea-salt enrichment is
commonly connected with elevated concentrations of heavy metals in the sea surface microlayer.
Table 2.5. Emission factors of heavy metals for suspension with sea-salt aerosol
Metal
As
Cd
Cr
Ni
Pb
Emission factor (g/kg)
300
40
80
180
4000
Reference
Nriagu, 1989
Richardson et al., 2001
Nriagu, 1989
Nriagu, 1989
Richardson et al., 2001
Estimates of re-suspension of particle-bound heavy metals from soil and seawater were
performed for Europe and adjacent territories in 2000. Spatial distributions of the mean
annual re-suspension flux of Pb, Cd, As, Cr, and Ni are presented in Fig. 2.9. In general, the
re-suspension fluxes from soil are significantly higher than those from seawater for all the
metals. High re-suspension fluxes were obtained from desert areas of Africa and Central Asia
because of significant dust production in these regions. Elevated fluxes are also
characteristics of some countries of Western, Central, and Southeastern Europe, which are
conditioned by combination of relatively high concentration in soil and significant dust
suspension from urban and agricultural areas.
Aggregated values of lead re-suspension from soil in different European countries are
presented in Figure 2.10a along with total anthropogenic emissions based on official data. As
seen the estimated contribution of lead re-suspension is comparable or even higher than
anthropogenic emissions in such countries as Italy, France, Germany, Greece, Spain, the
United Kingdom etc., where observed concentration of this metal in soil considerably
exceeds its natural content in the Earth’s crust (Fig. 2.10b). The most probable reason for this
is long-term accumulation of historical depositions.
14
Poland
Top soil
0
15
Kazakhstan
Romania
United Kingdom
Poland
Turkey
Spain
Greece
Germany
Slovakia
Bosnia&Herzegovina
Macedonia
Slovakia
Macedonia
0
Bosnia&Herzegovina
Hungary
Netherlands
Hungary
2500
Netherlands
Croatia
Switzerland
Croatia
d
Switzerland
Belgium
10
Czech Rep.
20
Czech Rep.
30
Bulgaria
40
Belgium
Crust
Bulgaria
60
Serbia&Montenegro
a
Serbia&Montenegro
b
Kazakhstan
Romania
United Kingdom
50
Turkey
Spain
Greece
Portugal
Ukraine
France
Italy
Russia
Pb total emissions, t/y
c
Germany
Portugal
Ukraine
France
Italy
Russia
Pb soil concentrations, mg/kg
a
b
e
Figure 2.9. Spatial distribution of annual re-suspension flux of heavy metals in Europe in 2000: (a) –
Pb; (b) – Cd; (c) – As; (d) – Cr; (e) – Ni
3000
Re-suspension
2000
Anthoropogenic emissions
1500
1000
500
Figure 2.10. Lead total anthropogenic emissions and re-suspension from soil (a) and average topsoil
concentration (b) in some European countries
Contrary to lead, cadmium re-suspension from soil insignificantly contributes to total
emission of this metal in most European countries (Fig. 2.11a). The exceptions are France,
Italy and Greece. The reason for this is in relatively low cadmium concentrations measured in
European soils. Only in a few countries of Europe (France, Italy, Greece, Belgium etc.) mean
topsoil concentration noticeably exceeds cadmium natural content in the crust (Fig. 2.11b).
60
Cd total emissions, t/y
Re-suspension
Anthoropogenic emissions
40
Switzerland
Bosnia&Herzegovina
Portugal
Netherlands
Hungary
Azerbaijan
Czech Rep.
Greece
Belgium
0.8
Top soil
Crust
0.6
0.4
Switzerland
Netherlands
Portugal
Azerbaijan
Hungary
Belgium
Czech Rep.
Greece
Kazakhstan
Slovakia
Bosnia&Herzegovina
b
Serbia&Montenegro
Romania
Bulgaria
United Kingdom
Italy
Ukraine
Turkey
France
Spain
Germany
Poland
0
Moldova
0.2
Russia
Cd soil concentrations, mg/kg
a
Kazakhstan
Moldova
Serbia&Montenegro
Romania
United Kingdom
Italy
Bulgaria
Turkey
Ukraine
Spain
France
Poland
Germany
Russia
0
Slovakia
20
Figure 2.11. Cadmium total anthropogenic emissions and re-suspension from soil (a) and average
topsoil concentration (b) in some European countries
2.2.3 Natural emission of Hg
Available data on the natural emission of Hg are rather uncertain. To take into account
natural emission of Hg we used global estimates by Lamborg et al. (2002). Spatially resolved
natural emission fluxes ware obtained by distribution of the total values of natural emission
over the globe depending on the surface temperature for emissions from land and
proportional to the primary production of organic carbon for emissions from the oceans. The
temperature dependence was described by an Arrhenius type equation with empirically
derived activation energy about 20 kcal/mole (Kim et al., 1995; Carpi and Lindberg, 1998;
Poissant and Casimir, 1998; Zhang et al., 2001). Evasion of Hg from geochemical
mercuriferrous belts (Gustin et al., 1999) was assumed to be 10 times higher then that from
background soils. Monthly mean spatially resolved data on the ocean primary production of
carbon (Behrenfeld and Falkowski, 1997) were utilised to distribute the natural Hg emission
flux over the ocean. Based on this methodology, it was estimated that approximately 10% of
the overall global natural emissions (177 tonnes/year) were emitted from the EMEP domain.
Besides, re-emission of previously deposited Hg was taken into account using approach
suggested by Ryaboshapko and Ilyin (2001). In this work the current anthropogenic reemission flux for Europe is estimated to be on the order of 50 tonnes per year. This value was
16
distributed over Europe based on estimates of relative cumulative deposition. For this
purpose Hg deposition in Europe accumulated during last century was estimated using the
regional transport model, and a simple box model considered soil as a reservoir with two
output fluxes – re-emission and hydrological leaching – was applied.
Natural emission and re-emission of Hg from the Mediterranean Sea has been corrected in
accordance with estimates performed in CNR-IIA using the MECAWEx model (see
Appendix ##).
3.
HEAVY METAL AIRBORNE POLLUTION LEVELS IN EUROPE IN THE
BASE YEAR 2000
The described above MSCE-HM model has been used for calculation of ambient air
concentration and deposition fields of As, Cd, Cr, Hg, Ni, and Pb in Europe in the base year
2000. The main aim of these calculations was evaluation of a reference pollution levels for
estimates of future reduction scenarios and evaluation of the model performance against
measurements from the EMEP monitoring network. Results of these modelling activities are
presented below.
3.1
Air concentration and deposition fields and model evaluation
Arsenic
Spatial distribution of calculated As concentration in the ambient air and annual deposition in
Europe in 2000 are presented in Fig. 3.1. Spatial patterns of both concentrations and
deposition fluxes to significant extend reflect the emission field (see Fig. 2.4a). Elevated
concentrations and deposition levels were obtained for Western and Central Europe and in
some areas of Eastern and Southern Europe. Besides, high concentrations of As in air are
characteristics of Africa deserts areas due to significant natural emission and re-suspension of
As with mineral dust. On the other hand, depositions of As in these areas are low because of
insignificant annual precipitation amount and low roughness of the ground surface that
restrict wet and dry depositions, respectively.
a
b
Figure 3.1. Spatial distribution of As mean annual concentration in ambient air (a) and annual
deposition (b) in Europe in 2000
17
Evaluation of modelling results against available measurements is presented in Figs. 3.2 and
3.3. Figure 3.2 shows scatter plots of modelled vs. observed concentrations of As in air and
precipitation. As seen the model demonstrates good correlation with measurements for air
concentrations and somewhat worse correlation for concentration in precipitation.
Nevertheless, most modelling results agree with observations within a factor of two.
1
Dmod = 0.72 Dobs
Rcorr = 0.37
1
Model, g/L
3
Model, ng/m
Cmod = 1.24 Cobs
Rcorr = 0.85
0.1
0.1
As
As
0.01
0.01
0.1
1
3
Observed, ng/m
a
0.1
Observed, g/L
b
1
Figure 3.2. Scatter plots of modeled vs. observed As concentrations in air (a) and precipitation (b) in
2000. Red solid line delineates 1:1 ratio, red and green dashed lines show deviation by a factor of two
and three, respectively
Figure 3.3 demonstrates comparison of modelled and measured values at individual sites of
the EMEP network. It also shows contribution of different source types (such as
anthropogenic sources, natural emission and re-suspension, non non-European sources
evaluated via boundary conditions) to modelled concentrations. As seen contribution of
natural sources and re-suspension is quite significant for As background levels in Europe. At
most sites this contribution is comparable or even exceeds that of anthropogenic sources.
Air concentration, ng/m
3
1.5
Bound
Natural
Anthrop
Obs
1.0
0.5
Concentratio in prec,  g/L
NO99
NL9
a
NO42
IS91
GB91
GB90
FI96
GB14
DK31
DK8
DK10
DK5
DK3
0.0
0.4
Bound
Natural
Anthrop
Obs
0.3
0.2
0.1
SE5
SE51
NO99
NO95
NL91
NO94
NL9
GB91
GB90
FR90
GB14
FI96
FI93
FI92
FI53
FI17
FI9
DK31
DK8
DE9
DK20
b
DE4
DE1
0.0
Figure 3.3. Comparison of modelled vs. measured As concentrations in air (a) and precipitation (b) in
2000 at monitoring sites of the EMEP network. Circles show observed values; bars present model
predictions. Contribution of European anthropogenic, natural and non-European sources to modelled
values are presented by green, blue and orange bars, respectively
18
Cadmium
Modelled concentration and deposition fields of Cd in 2000 are shown in Fig. 3.4. According
to the current estimates contribution of wind re-suspension to ambient levels of Cd is
insignificant as it was mentioned in Section 2.2. Therefore concentration and deposition
patterns are mostly defined by anthropogenic emissions. Contrary to the above mentioned
case of As no elevated concentrations were obtained in Africa. The most significant Cd
depositions are characteristics of some Western, Central and Southern European countries,
such as Germany, Poland, Italy, etc.
a
b
Figure 3.4. Spatial distribution of Cd mean annual concentration in ambient air (a) and annual
deposition (b) in Europe in 2000
Figures 3.5 and 3.6 illustrate comparison of modelled and measured concentrations in air and
precipitation at sites of the EMEP monitoring network. Correlation of predicted and observed
values is significant in both cases (correlation coefficients are 0.56 and 0.53) and majority of
the model-measurement pairs agree within a factor of two. The model tends to underestimate
observations by 15-30%, possibly, because of underprediction of wind re-suspension and
emission from natural sources. As seen from Figs 3.5b and 3.6b the underestimation of
measured concentrations in precipitation is more pronounced for Northern European sites
located in Scandinavia. It can be also connected with peculiarities of scavenging of aerosolbound substances by frozen precipitation under winter conditions.
Cmod = 0.88 Cobs
Rcorr = 0.56
Dmod = 0.71 Dobs
Rcorr = 0.53
0.1
Model, g/L
3
Model, ng/m
1
0.1
0.01
a
0.01
Cd
0.01
0.1
3
Observed, ng/m
Cd
0.01
1
b
0.1
Observed, g/L
Figure 3.5. Scatter plots of modelled vs. observed Cd concentrations in air (a) and precipitation (b) in
2000. Red solid line delineates 1:1 ratio, red and green dashed lines show deviation by a factor of two
and three, respectively
19
Air concentration, ng/m
3
0.8
Bound
Natural
Anthrop
Obs
0.6
0.4
0.2
SK7
SK6
SK5
SK4
SK2
NO99
NL9
NO42
LV16
LV10
IS91
LT15
GB91
GB90
GB14
DK8
DK31
DK5
CZ3
DK3
0.20
Bound
Natural
Anthrop
Obs
0.15
0.10
0.05
0.00
CZ1
CZ3
DE1
DE4
DE9
DK8
DK20
DK31
EE9
EE11
FI8
FI9
FI17
FI22
FI53
FI92
FI93
FI96
FR90
GB14
GB90
GB91
IS2
IS90
LV10
LV16
NL9
NL91
NO1
NO39
NO41
NO55
NO56
NO92
NO93
NO94
NO95
NO99
SE5
SE51
SE97
Concentration in prec, g/L
a
CZ1
AT2
0.0
b
Figure 3.6. Comparison of modelled vs. measured Cd concentrations in air (a) and precipitation (b) in
2000 at monitoring sites of the EMEP network. Circles show observed values; bars present model
predictions. Contribution of European anthropogenic, natural and non-European sources to modelled
values are presented by green, blue and orange bars, respectively
Chromium
Chromium is an intrinsic component of the Earth’s crust widely appearing natural soils.
Therefore contribution of natural sources to ambient levels of Cr is quite significant.
Particularly, high air concentrations of Cr were obtained for desert areas of Africa and
Central Asia due to elevated Cr suspension with mineral dust (Figure 3.7a). Atmospheric
transport of Cr from African natural emission sources significantly affects pollution levels in
Southern Europe where elevated Cr depositions were predicted (Figure 3.7b).
a
b
Figure 3.7. Spatial distribution of Cr mean annual concentration in ambient air (a) and annual
deposition (b) in Europe in 2000
20
Modelled concentrations of Cr in air and precipitation are in good agreement with available
measurements (Fig. 3.8). The correlation coefficients between modelled and observed values
are high and exceed 0.85 and 0.6 for concentrations in air and precipitation, respectively.
Only few modelled values deviate from appropriate measurements more than by a factor of
two. Contribution of natural sources to modelled concentrations is not very large at the
considered monitoring sites (Fig. 3.9) since most of them are located in Central, Western and
Northern Europe far form regions with major mineral dust emission.
1
10
Dmod = 1.06 Dobs
Rcorr = 0.63
Model, g/L
3
Model, ng/m
Cmod = 0.95 Cobs
Rcorr = 0.86
1
0.1
0.1
Cr
0.1
1
3
Observed, ng/m
a
Cr
0.01
0.01
10
0.1
Observed, g/L
b
1
Figure 3.8. Scatter plots of modelled vs. observed Cr concentrations in air (a) and precipitation (b) in
2000. Red solid line delineates 1:1 ratio, red and green dashed lines show deviation by a factor of two
and three, respectively
Air concentration, ng/m
3
3
Bound
Natural
Anthrop
Obs
2
1
SK7
SK6
SK5
SK4
SK2
NO99
GB91
NO42
GB90
GB14
FI96
Bound
Natural
Anthrop
Obs
0.3
0.2
0.1
NO99
NO93
NL9
GB91
GB90
GB14
FR90
FI96
FI93
FI92
FI53
FI9
DK31
DK8
DE9
DE4
0.0
DK20
b
DK31
0.4
DE1
Concentration in prec, g/L
a
DK10
DK8
DK5
DK3
0
Figure 3.9. Comparison of modelled vs. measured Cr concentrations in air (a) and precipitation (b) in
2000 at monitoring sites of the EMEP network. Circles show observed values; bars present model
predictions. Contribution of European anthropogenic, natural and non-European sources to modelled
values are presented by green, blue and orange bars, respectively
21
Mercury
Mercury significantly differs from other heavy metals by the fact that it appears in the
atmosphere in different physical and chemical forms. The bulk Hg form in the atmosphere
(up to 95%) –gaseous elemental mercury (GEM) – is characterized by long residence time
(months to year) and can be transported at the global scale. On the other hand, Hg depositions
are mostly defined by short-lived forms (reactive gaseous and particulate), which originate in
the atmosphere from direct anthropogenic emissions or from oxidation of the elemental form.
Predicted spatial distributions of GEM concentration and annual Hg deposition in Europe in
2000 are presented in Fig. 3.10. Long residence time of GEM leads to significant levelling its
concentration in the atmosphere. It relatively slightly varies from industrial to remote regions
(Fig. 3.10a). Elevated concentrations are in areas with high anthropogenic emissions and over
Mediterranean Sea because of significant natural emission from seawater. Spatial pattern of
Hg depositions considerably differs from that of air concentration because they define by
short-lived species and more strongly connected with anthropogenic emissions (Fig. 3.10b).
As seen from Figs. 3.11 and 3.12 modelling results for Hg fairly well agree with
observations. However, one should mention relatively small number of monitoring sites
measuring Hg in Europe. Contribution of non-European sources to air concentration of total
gaseous Hg is very large because of long residence time as it was mentioned above (Fig.
3.12a). Contrary to air concentrations, Hg depositions are more significantly affected by
anthropogenic sources. Besides, contribution of European natural emissions is very low
because it is expected that Hg is emitted from natural sources in elemental form and mostly
flows out the European region.
a
b
Figure 3.10. Spatial distribution of elemental gaseous Hg concentration in ambient air (a) and annual
deposition (b) in Europe in 2000
22
5
50
Dmod = 1.10 Dobs
Rcorr = 0.93
Model, ng/L
3
Model, ng/m
Cmod = 1.01 Cobs
Rcorr = 0.24
2
10
1
Hg
Hg
0.5
0.5
2
1
2
3
Observed, ng/m
a
5
2
10
Observed, ng/L
b
50
Obs
1.5
1.0
Calabria
20
Anthrop
Natural
Bound
Obs
15
10
5
SE5
SE2
SE11
b
NO99
0
DE1
Sicily
SE2
NO99
NO42
IE31
FI96
DK15
DK10
DE9
Neve Yam
0.5
0.0
a
Bound
NL91
2.0
Natural
DE9
Anthrop
Concentration in prec, ng/L
2.5
Antalya
TGM concentration, ng/m
3
Figure 3.11. Scatter plots of modelled vs. observed Hg concentrations in air (a) and precipitation (b)
in 2000. Red solid line delineates 1:1 ratio, red and green dashed lines show deviation by a factor of
two and three, respectively
Figure 3.12. Comparison of modelled vs. measured Hg concentrations in air (a) and precipitation (b)
in 2000 at monitoring sites of the EMEP network. Circles show observed values; bars present model
predictions. Contribution of European anthropogenic, natural and non-European sources to modelled
values are presented by green, blue and orange bars, respectively
23
Nickel
Spatial distributions of modelled Ni air concentration and deposition in Europe are shown in
Fig. 3.13. High concentration and deposition fluxed were predicted in Western and Southern
Europe in such countries as Germany, the UK, Belgium, the Netherlands, Italy etc. Besides,
elevated concentrations are also characteristics of desert areas in Africa.
a
b
Figure 3.13. Spatial distribution of Ni mean annual concentration in ambient air (a) and annual
deposition (b) in Europe in 2000
Comparison of modelled Ni concentration in air and precipitation with measurements is
presented in Figs. 3.14 and 3.15. Correlation between predicted and observed values is very
high for air concentration (0.89) and lower for concentration in precipitation (0.57). Besides,
the model tends to overestimate measurements by 15-30%. Particularly, it is characteristic for
some Danish, British and Dutch sites. Since contribution of natural emission and resuspension is not significant this overprediction can be connected with some overestimation
of anthropogenic emissions in these countries.
10
Dmod = 1.16 Dobs
Rcorr = 0.57
1
Model, g/L
3
Model, ng/m
Cmod = 1.29 Cobs
Rcorr = 0.89
1
0.1
0.1
Ni
a
0.1
1
3
Observed, ng/m
Ni
10
b
0.1
1
Observed, g/L
Figure 3.14. Scatter plots of comparison of modeled and observed Ni concentrations in air (a) and
precipitation (b) in 2000. Red solid line delineates 1:1 ratio, red and green dashed lines show
deviation by a factor of two and three, respectively
24
Air concentration, ng/m
3
4
Bound
Natural
Anthrop
Obs
3
2
1
SK7
SK6
SK5
SK4
SK2
NO99
GB91
NO42
GB14
Bound
Natural
Anthrop
Obs
1.0
0.8
0.6
0.4
0.2
SE5
SE51
NO99
NO95
NO94
NO93
NO92
IS2
NL9
GB91
GB90
GB14
FI96
FR90
FI93
FI92
FI53
FI17
FI9
DK31
DK8
DE9
DE4
0.0
DK20
b
FI96
1.2
DE1
Concentration in conc, mg/L
a
DK31
DK8
DK10
DK5
DK3
AT2
0
Figure 3.15. Comparison of modeled vs. measured Ni concentrations in air (a) and precipitation (b) in
2000 at monitoring sites of the EMEP network. Circles show observed values; bars present model
predictions. Contribution of European anthropogenic, natural and non-European sources to modeled
values are presented by green, blue and orange bars, respectively
Lead
Contribution of natural emissions and re-suspension to Pb pollution levels in Europe is
significant (see Fig. 2.10), particularly, in industrial and urban areas where accumulation of
Pb in topsoil has taken place during long time period. Therefore spatial patterns of Pb air
concentration and deposition in Europe reflect both influence of direct anthropogenic sources
and re-suspension. High deposition levels are characteristics of areas with significant
anthropogenic emissions in Western Germany, Southern Poland, Northern Italy, Eastern
Ukraine etc.
a
b
Figure 3.16. Spatial distribution of Pb mean annual concentration in ambient air (a) and annual
deposition (b) in Europe in 2000
25
Figures 3.16 and 3.18 demonstrate evaluation of modelling results for Pb against
measurements. As seen from the figures modelled concentrations both in air and precipitation
underestimates observed ones by 20-25%. Nevertheless, correlation is significant in both
cases and in most cases discrepancy between predicted and measured values does not exceed
a factor of two. As for cadmium, the model tends to underpredict Pb concentrations in
precipitation measured at Scandinavian sites (Figs. 3.17b and 3.18b).
10
Dmod = 0.79 Dobs
Rcorr = 0.62
3
Model, ng/m
Cmod = 0.75 Cobs
Rcorr = 0.72
Model, g/L
10
1
1
Pb
Pb
0.1
0.1
1
10
3
Observed, ng/m
a
1
Observed, g/L
b
10
Figure 3.17. Scatter plots of comparison of modeled and observed Pb concentrations in air (a) and
precipitation (b) in 2000. Red solid line delineates 1:1 ratio, red and green dashed lines show
deviation by a factor of two and three, respectively
Air concentration, ng/m
3
20
Bound
Natural
Anthrop
Obs
16
12
8
4
b
SK7
SK6
SK5
SK4
SK2
NO99
NL9
NO42
LV16
LV10
IS91
LT15
GB91
GB90
FI96
GB14
DK31
DK8
DK10
DK5
DK3
DE9
DE8
DE7
DE5
DE4
DE3
CZ3
DE1
AT5
CZ1
4
Bound
Natural
Anthrop
Obs
3
2
1
0
CZ1
CZ3
DE1
DE4
DE9
DK8
DK20
DK31
EE9
EE11
FI8
FI9
FI17
FI22
FI53
FI92
FI93
FI96
FR90
GB14
GB90
GB91
IS2
IS90
LV10
LV16
NL9
NL91
NO1
NO39
NO41
NO55
NO56
NO92
NO93
NO94
NO95
NO99
SE5
SE51
SE97
SK4
SK5
SK6
SK7
Concentration in prec, g/L
a
AT4
AT2
0
Figure 3.18. Comparison of modeled vs. measured Pb concentrations in air (a) and precipitation (b) in
2000 at monitoring sites of the EMEP network. Circles show observed values; bars present model
predictions. Contribution of European anthropogenic, natural and non-European sources to modeled
values are presented by green, blue and orange bars, respectively
26
3.2
Uncertainties
Detailed analysis of the MSCE-HM model uncertainty has been performed based on
sensitivity study (Travnikov and Ilyin, 2005). As an example of this study, Figure 3.19
illustrates estimated uncertainties of the main output variables for Pb caused by inaccuracies
of input parameters. The most significant uncertainties are introduced by anthropogenic and
natural emissions along with re-emission and exceed 30% on average. High uncertainty of
natural emission and re-emission leads to their contribution to the overall uncertainty at least
comparable with the anthropogenic one. Meteorological parameters and characteristics of
removal processes also cause considerable model uncertainty.
Eant
Eant
Eant
Enat
Enat
Enat
Meteo
Meteo
Meteo
Vd
Lwet
Lwet
Lwet
Vd
Vd
Kz
Cbound
Cbound
Cbound
Kz
Kz
Pb in air
LWC
0
a
20
40
Uncertainty, %
Pb in precipitation
LWC
60
0
10
20
30
Uncertainty, %
40
Total Pb deposition
LWC
50
b
0
10
20
30
Uncertainty, %
40
50
c
Figure 3.19. Uncertainty of Pb concentration in air (a), in precipitation (b) and of total Pb deposition
flux (c) due to inaccuracy of main input parameters. The error bars show 90% confidence interval
The estimated overall uncertainties of the main model parameters are presented in Table 3.1.
It was expected that the model uncertainty includes contributions of all model parameters
except anthropogenic, natural emissions and re-emission. The range indicates 90%
confidence interval of the uncertainty variation over the model domain. The intrinsic model
uncertainty of particle-bound heavy metals (As, Cd, Cr, Ni, Pb) varies from 20% to 65% over
the domain with average values 43%, 40% and 33% for concentration in air, concentration in
precipitation and total deposition, respectively. The intrinsic model uncertainty of Hg differs
for different outputs. It does not exceed 20% on average for TGM concentration (the range
16-22%) but reaches 40% for total deposition and 50% for concentration in precipitation (the
ranges 20-57% and 29-74% respectively).
Table 3.1. Model intrinsic uncertainty of the main model output parameters
Uncertainty, %
Output parameter
As, Cd, Cr, Ni, Pb
Hg
Air concentration
43 (22-64)
19 (16-22)
Concentration in precipitation
40 (20-57)
53 (29-74)
Total deposition
33 (19-49)
39 (20-57)
27
4.
REDUCTION OF HEAVY METAL ATMOSPHERIC DEPOSITIONS IN
EUROPE BY 2010
Airborne pollution levels with As, Cd, Cr, Hg, Ni, and Pb in Europe have been also evaluated
for 2010 based on BAU and MFTR emission scenarios. Emissions data for these scenarios
were prepared in WP02 of the project. Calculations of heavy metal atmospheric dispersion
according to these scenarios have been performed in WP04 and delivered to other workpackages of the project. Below we present a short description of the modelling results.
To evaluate reduction of heavy metal depositions between 2000 and 2010 and to avoid effect
of meteorological variability calculations of the atmospheric dispersion in 2000 was repeated
using the same meteorological data as for the year 2010 (see Section 2.1). Besides, natural
emissions, wind re-suspension and boundary conditions were assumed unchangeable. Figure
4.1 presents comparison of estimated total deposition of heavy metals to European countries
in 2000 and two emission scenarios for 2010. As seen the most significant deposition
reduction was predicted for Cd, Pb and Ni (about 20% and 30% for BAU and MFTR,
respectively). It can be explained both by considerable expected reduction of anthropogenic
emissions of these metals and low or moderate contribution of natural emission and resuspension. The lowest reduction was estimated for As and Cr (below 8% and 15% for BAU
and MFTR, respectively), which are significantly affected by the wind-re-suspension.
Deposition reduction of Hg was also predicted to be insignificant (8% and 18% for BAU and
MFTR, respectively) because of large contribution of the intercontinental transport from
other continents. This contribution can be expected to be unchangeable or even increasing
during considered period because of considerable growth of Hg emissions in Asia.
0.6
1.5
1.0
0.5
0.0
BAU
MFTR
b
MFTR
0.10
0.05
0.00
BAU
MFTR
6.0
4.0
2.0
0.0
e
2.0
2000
BAU
MFTR
16
Ni
4.0
2.0
0.0
2000
4.0
c
Total deposition, kt/y
0.15
Cr
0.0
BAU
6.0
Hg
Total deposition, kt/y
Total deposition, kt/y
0.2
2000
0.20
d
0.4
0.0
2000
a
6.0
Cd
Total deposition, kt/y
As
Total deposition, kt/y
Total deposition, kt/y
2.0
Pb
12
8
4
0
2000
2000
Anthropogenic
BAU
BAU
Natural
MFTR
f
2000
BAU
MFTR
MFTR
Boundary
Figure 4.1. Estimated total deposition of heavy metals to European countries in 2000 and 2010
according to BAU and MFTR emission scenarios
Decrease of heavy metal depositions considerably varies from one part of Europe to another
depending on emission reduction in particular countries. Spatial distributions of As, Cd, Cr,
Hg, Ni, and Pg deposition reduction in Europe between 2000 and 2010 according to BAU
28
and MFTR emission scenarios are presented in Figs. 4.2-4.7. The most significant deposition
reduction is expected in Western and Central Europe for al the metals. However, magnitude
and spatial pattern of the reduction varies between different metals.
a
b
Figure 4.2. Spatial distribution of As deposition reduction in Europe in 2010 with regard to 2000 in
accordance with BAU (a) and MFTR (b) emission scenarios
a
b
Figure 4.3. Spatial distribution of Cd deposition reduction in Europe in 2010 with regard to 2000 in
accordance with BAU (a) and MFTR (b) emission scenarios
29
a
b
Figure 4.4. Spatial distribution of Cr deposition reduction in Europe in 2010 with regard to 2000 in
accordance with BAU (a) and MFTR (b) emission scenarios
a
b
Figure 4.5. Spatial distribution of Hg deposition reduction in Europe in 2010 with regard to 2000 in
accordance with BAU (a) and MFTR (b) emission scenarios
a
b
Figure 4.6. Spatial distribution of Ni deposition reduction in Europe in 2010 with regard to 2000 in
accordance with BAU (a) and MFTR (b) emission scenarios
30
a
b
Figure 4.7. Spatial distribution of Pb deposition reduction in Europe in 2010 with regard to 2000 in
accordance with BAU (a) and MFTR (b) emission scenarios
5.
SENSITIVITY OF HEAVY METAL DEPOSITION IN EUROPE TO
EMISSION REDUCTION IN EUROPEAN COUNTRIES
Bulk modelling activity within WP04 of the project was connected with calculations of
source-receptor matrices of heavy metal deposition sensitivity to emission reduction in
different European countries. This kind of matrices define a link between emission reduction
in each European country and heavy metal deposition to each grid cell (or even to each land
use type within a grid cell) of the model domain. The main aim of these calculations was
support of the optimization modelling of emission reduction in Europe for policy-oriented
applications.
For this purpose calculation of atmospheric dispersion of the selected heavy metals (As, Cd,
Cr, Hg, Ni, and Pb) have been repeatedly performed reducing each time emission in one of
European countries by 10%. This sort of calculations has been performed for all European
countries and for two emission scenarios of 2010 (BAU and MFTR). Consideration of two
scenarios with distinct emissions allows taking into account possible non-linear dependence
of deposition on anthropogenic emission reduction. Obtained modelling results have been
delivered to other work-packages for further use in the full chain of the project.
Below a short analysis of the obtained results are presented. To characterize response of
heavy metal deposition to emission reduction in each particular country we calculate the
following sensitivity coefficient
S .c. 
Dij / Dij
E k / E k
,
(5.1)
where Dij is deposition of a heavy metal to a grid cell (i,j); Ek is emissions flux in kth country;
and Dij and Ek are changes of appropriate values. For illustration purposes the sensitivity
coefficient was averaged over two emission scenarios.
Figure 5.1 illustrates spatial distribution of the sensitivity coefficient of Pb deposition to
emission reduction in three countries located in different parts of Europe – Spain, Germany,
31
Russia and Ukraine. As seen the highest sensitivity characterizes depositions to own territory
of the countries. Besides, spatial pattern of deposition sensitivity depends on location of a
country. Spain is located aside of major European territory, therefore, emission reduction in
this country mostly affect its own territory and southern part of France. Reduction of Pb
emissions in Germany leads to decrease of depositions in surrounding countries such as
France, Poland, Austria, Czech Republic etc. Emissions reduction in Russia mostly effect its
own territory but also neighbouring countries – Finland, Estonia, Latvia, Lithuania, Belarus
and Ukraine. On the other hand, reduction of Pb emissions in Ukraine significantly effect
depositions in southern part of Russia.
a
b
c
d
Figure 5.1. Spatial distribution of the sensitivity coefficient of Pb deposition in Europe to emission
reduction in some European countries: (a) – Spain; (b) – Germany; (c) – Russia
Frequency distribution functions of the sensitivity coefficient values for Pb, Cd, and Hg
emissions reduction in these countries are presented in Fig. 5.2. As seen the distributions
have a multi-peak shapes for all the countries. A peak of the sensitivity coefficient close to
unity indicate high sensitivity of depositions to the country’s own territory. This peak is
relatively big for Russia because of large territory of the country. Other maximums present
lower sensitivity of deposition to other areas. In general, Cd depositions are characterised by
the highest sensitivity because they more depend on anthropogenic emissions then other
32
metals. Mercury depositions are the least sensitive to anthropogenic emissions reduction in
Europe because of significant contribution of global sources.
1.5
1.5
Cd
Hg
Pb
Frequency
1.2
Germany
0.9
0.6
0.3
0.001
0.01
0.1
0.6
a
0.0
0.0001
1
Sensitivity coefficient
0.001
0.01
0.1
1
Sensitivity coefficient
b
1.5
1.5
Cd
Hg
Pb
1.2
0.9
0.6
0.3
0.0
0.0001
Cd
Hg
Pb
Ukraine
1.2
Frequency
Russia
Frequency
0.9
0.3
0.0
0.0001
c
Cd
Hg
Pb
1.2
Frequency
Spain
0.9
0.6
0.3
0.001
0.01
0.1
Sensitivity coefficient
0.0
0.0001
1
d
0.001
0.01
0.1
1
Sensitivity coefficient
Figure 5.2. Frequency distribution functions of the sensitivity coefficient of heavy metal deposition
over territory of Europe to emission reduction in some countries: (a) – Spain; (b) – Germany;
(c) – Russia; (d) – Ukraine
To characterize effectiveness of heavy metal emissions reduction in European countries from
the viewpoint of decrease of atmospheric depositions we computed the medians as well as
5% and 95% percentiles of the sensitivity coefficient frequency distributions for emission
reduction in each European country. Figure 5.3 shows this kind of statistical parameters for
all six heavy metals considered in the project. As seen from the figure heavy metal
depositions in Europe are the most sensitive to emissions reduction of Ukraine and Russia.
Among the countries significantly affecting depositions are also Germany, Poland, Italy,
France and the United Kingdom. However, one should keep in mind that general statistics
characterizes situation in Europe as a whole, whereas deposition sensitivity in particular
regions varies significantly.
33
Russia
34
1.E-01
1.E-01
Finland
Portugal
1.E-01
Albania
Albania
Portugal
Macedonia
Estonia
Lithuania
Croatia
Norway
Slovenia
Bosnia&Herz.
Latvia
Ireland
Luxemburg
Rep.Moldova
Finland
Denmark
Greece
Slovakia
Switzerland
Bulgaria
1.E-01
Luxemburg
Ireland
Luxemburg
Norway
Albania
Estonia
Latvia
Macedonia
Croatia
Bosnia&Herz.
Slovenia
Sweden
Lithuania
Greece
Rep.Moldova
Denmark
Serbia&Mont.
Bulgaria
Sweden
Serbia&Mont.
Luxemburg
Albania
Estonia
Portugal
Macedonia
Norway
Latvia
Lithuania
Croatia
Ireland
Finland
Bosnia&Herz.
Denmark
Slovenia
Rep.Moldova
Sweden
Switzerland
Greece
Serbia&Mont.
Austria
Netherlands
Turkey
Belgium
Belarus
Slovakia
Hungary
Bulgaria
Spain
CzechRep.
Romania
France
UK
Poland
Germany
Italy
1.E-01
Estonia
Macedonia
Estonia
Portugal
Croatia
Latvia
Bosnia&Herz.
Ireland
Lithuania
Rep.Moldova
Finland
Slovenia
Bulgaria
Norway
Denmark
Greece
Serbia&Mont.
Sweden
Belgium
Turkey
Austria
Turkey
Hungary
Belarus
Romania
Netherlands
CzechRep.
Spain
Belgium
Italy
UK
France
Germany
Poland
Russia
Ukraine
Sensitivity coefficient
Portugal
Albania
Luxemburg
Ireland
Lithuania
Macedonia
Croatia
Norway
Denmark
Bosnia&Herz.
Finland
Latvia
Switzerland
Slovenia
Rep.Moldova
Sweden
Greece
Bulgaria
Estonia
Serbia&Mont.
Austria
Netherlands
Slovakia
Hungary
Belgium
Romania
Turkey
CzechRep.
UK
Belarus
Spain
Italy
France
Poland
Germany
Russia
Ukraine
Sensitivity coefficient
1.E-01
Macedonia
Norway
Lithuania
Portugal
Latvia
Albania
Luxemburg
Ireland
Slovenia
Rep.Moldova
Croatia
Bosnia&Herz.
Denmark
Finland
Slovakia
Austria
Sweden
Serbia&Mont.
Belarus
Slovakia
Austria
Spain
Belarus
Netherlands
CzechRep.
Switzerland
UK
Romania
Italy
France
Slovakia
Hungary
Germany
Poland
Russia
Ukraine
Sensitivity coefficient
b
Bulgaria
Greece
Switzerland
Russia
Ukraine
Sensitivity coefficient
a
Hungary
Turkey
Hungary
Austria
Belgium
CzechRep.
d
Switzerland
CzechRep.
Belgium
Spain
Poland
UK
France
Germany
Italy
Romania
Netherlands
e
Spain
Belarus
Romania
France
UK
Poland
Germany
Italy
Russia
Ukraine
Sensitivity coefficient
c
Netherlands
Turkey
f
Ukraine
Sensitivity coefficient
1.E+00
1.E-02
As
1.E-03
1.E-04
1.E-05
1.E-06
1.E+00
1.E-02
Cd
1.E-03
1.E-04
1.E-05
1.E-06
1.E+00
1.E-02
Cr
1.E-03
1.E-04
1.E-05
1.E-06
1.E+00
1.E-02
Hg
1.E-03
1.E-04
1.E-05
1.E-06
1.E+00
1.E-02
Ni
1.E-03
1.E-04
1.E-05
1.E-06
1.E+00
1.E-02
Pb
1.E-03
1.E-04
1.E-05
1.E-06
Figure 5.3. Sensitivity coefficient of heavy metal deposition in Europe to emission reduction in
different European countries for: (a) – As, (b) – Cd, (c) – Cr, (d) – Hg, (e) – Ni, (f) – Pb. Dots present
medians of the frequency distribution functions; ranges show 90%-confidense intervals over the
European territory
6.
CONCLUSIONS
Modelling activities related to evaluation of atmospheric dispersion of six selected heavy
metals (As, Cd, Cr, Hg, Ni, and Pb) in Europe have been performed within framework of
WP04 of the project.
Particularly, the chemical transport model MSCE-HM has been reviewed and improved.
Major model parameterisations have been revised as well as all required input data have been
prepared for modelling process. New parameterisation of particle-bound heavy metal wind
re-suspension from soil and seawater has been developed.
Heavy metal airborne pollution levels in Europe have been evaluated for the base year 2000.
The modelling results have been compared with available measurements. It was obtained that
discrepancies between modelled and observed values mostly do not exceed a factor of two.
Besides, the intrinsic model uncertainty was estimated to be not higher than 50%.
Reduction of heavy metal deposition in Europe between 2000 and 2010 has been estimated
based on BAU and MFTR emission scenarios. It was obtained that the reduction varies both
for different heavy metals and in different parts of Europe. The highest deposition reduction
was predicted for Cd, Pb and Ni (about 20% and 30% for BAU and MFTR, respectively); the
lowest – for As and Cr (below 8% and 15% for BAU and MFTR, respectively). The value of
deposition reduction depends on reduction of anthropogenic emissions, contribution of
natural sources and wind re-suspension, and influence of the intercontinental atmospheric
transport.
Source-receptor matrices of heavy metal deposition sensitivity to emission reduction in
different European countries have been computed. This kind of matrices define a link
between emission reduction in each European country and heavy metal deposition to each
grid cell (or even to each land use type within a grid cell) of the model domain. Obtained
modelling results have been delivered to other work-packages for further use in the full chain
of the project.
35
7.
REFERENCES
Alfaro S.C. and L. Gomes [2001] Modeling mineral aerosol production by wind erosion: Emission intensities
and aerosol size distributions in source areas. Journal of Geophysical Research, vol. 106, No. D16,
pp.18,075 – 18,084.
Alfaro S.C., Gaudichet A., Gomes L. and M. Maillé [1998] Mineral aerosol production by wind erosion: aerosol
particle sizes and binding energies. Geophysical Research Letters, vol. 25. No. 7, pp.991– 994.
Alfaro S.C., Gaudichet A., Gomes L., Maillé M. [1997] Modeling the size distribution of a soil aerosol produced
by sandblasting. Journal of Geophysical Research, vol.102, pp.11239–11249.
Andersson M., Wängberg I., Gårdfeldt K., Munthe J. [2004] Investigation of the Henry’s low coefficient for
elemental mercury. Proceedings of the 7th Conference “Mercury as a global pollutant”. RMZ – Materials
and Geoenvironment, Ljubljana, June 2004.
Ariya P.A., Khalizov A., Gidas A. [2002] Reactions of gaseous mercury with atomic and molecular halogens:
kinetics, product studies, and atmospheric implications. J. Phys. Chem. 106, 7310-7320.
Behrenfeld M.J. and Falkowski P.G. [1997] Photosynthetic derived from satellite-based chlorophyll
concentration. Limnol. Oceanogr., 42(1), 1-20.
Bott A. [1989a] A positive definite advection scheme obtained by nonlinear renormalization of the advective
fluxes. Mon. Wea. Rev. 117, 1006-1015.
Bott A. [1989b] Reply to comment on “A positive definite advection scheme obtained by nonlinear
renormalization of the advective fluxes.” Mon. Wea. Rev. 117, 2633-2636.
Bott A. [1992] Monotone flux limitation in the area-preserving flux-form advection algorithm. Mon. Wea. Rev.
120, 2592-2602.
Carpi A, Lindberg S.E. [1998] Application of a teflonTM dynamic flux chamber for quantifying soil mercury
flux: tests and results over background soil. Atmos. Environ. 32(5), 873-882.
Fecan F., Marticorena B. and G.Bergametti [1999] Parameterization of the increase of the aeolian erosion
threshold wind friction velocity due to soil moisture for arid and semi-arid areas. Annales Geophysicae, vol.
17, pp. 149 – 157.
Gårdfeldt K., Sommar J., Strömberg D., Feng X. [2001] Oxidation of atomic mercury by hydroxyl radicals and
photoinduced decomposition of methylmercury in the aqueous phase. Atmos. Environ. 35, 3039-3047.
Gomes L., Rajot J.L., Alfaro S.C. and A. Gaudichet [2003] Validation of a dust production model from
measurements performed in semi-arid agricultural areas of Spain and Niger. Catena, vol. 52, pp. 257 – 271.
Gong S.L. [2003] A parameterization of sea-salt aerosol source function for sub- and super-micron particles.
Global Biogeochemical Cycles, vol. 17, No. 4, p. 1097.
Gong S.L., Zhang X.Y., Zhao T.L., McKendry I.G., Jaffe D.A., Lu N.M. [2003] Characterization of soil dust
aerosol inChina and its transport and distribution during 2001 ACE-Asia: 2. Model simulation and
validation. Journal of Geophysical Research, 108(D9), 4262.
Grell, G. A., J. Dudhia, and D. R. Stauffer [1995] A description of the fifth-generation Penn State/NCAR
Mesoscale Model (MM5). NCAR/TN-398+STR. NCAR Technical Note. Mesoscale and Microscale
Meteorology Division. National Center for Atmospheric Research. Boulder, Colorado. Pp. 122.
Grini A., Myhre G., Zender C.S. and I.S.A. Isaksen [2005] Model simulations of dust sources and transport in
the global atmosphere: Effects of soil erodibility and wind speed variability. Journal of Geophysical
Research, vol. 110, pp.D02205.
Gustin MS, Lindberg S, Marsik F, Casimir A, Ebinghaus R, Edwards G, Hubble-Fitzgerald C, Kemp R, Kock
H, Leonard T, London J, Majewski M, Montecinos C, Owens J, Pilote M, Poissant L, Rasmussen P,
Schaedlich F, Schneeberger D, Schroeder W, Sommar J, Turner R, Vette A, Wallschlaeger D, Xiao Z,
Zhang H. [1999] Nevada STORMS project: Measurement of mercury emissions from naturally enriched
surfaces. J. Geophys. Res. 104(D17), 21831-21844.
Hall B. [1995] The gas phase oxidation of mercury by ozone. WASP 80, 301-315.
Ilyin I., Travnikov O., Aas W., Uggerud H.Th. [2003] Heavy metals: transboundary pollution of the
environment. EMEP Status Report 2/2003, Meteorological Synthesizing Centre - East, Moscow, Russia.
Jacobson M. Z. [1999] Fundamentals of atmospheric modeling. Cambridge University Press. 656 p.
36
Kim K.-H., Lindberg S.E., Meyers T.P. [1995] Micrometeorological measurements of mercury vapor fluxes
over background forest souls in eastern Tennessee. Atmos. Environ. 29(2), 267-282.
Lamborg C. H., Fitzgerald W. F., O’Donnell J., Torgersen T. [2002] A non-steady-state compartmental model
of global-scale mercury biogeochemistry with interhemispheric atmospheric gradients. Geochimica et
Cosmochimica Acta 66(7), 1105-1118.
Lin C.-J., Pehkonen S. O. [1999] The chemistry of atmospheric mercury: a review. Atmos. Environ. 33, 20672079.
Lurie Yu. Yu. [1971] Handbook for Analytical Chemistry. Khimiya, Moscow, 454 p.
Marticorena B. and G. Bergametti [1995] Modeling the atmospheric dust cycle: 1. Design of a soil-derived dust
emission scheme. Journal of Geophysical Research, vol. 100, No. D8, pp. 16,415 – 16,430.
Marticorena B., Bergametti G, and B. Aumont [1997] Modeling the atmospheric dust cycle: 2. Simulation of
Saharan dust sources. Journal of Geophysical Research, vol. 102, No. D4, pp. 4387 – 4404.
Munthe J. [1992] The aqueous oxidation of elemental mercury by ozone. Atmos. Environ. 26A, 1461-1468.
Petersen G., Munthe J., Pleijel K., Bloxam R. and A.V.Kumar [1998] A comprehensive Eulerian modeling
framework for airborne mercury species: Development and testing of the tropospheric chemistry module
(TCM). Atmos. Environ. 32(5), 829-843.
Poissant L., Casimir A. [1998] Water–air and soil–air exchange rate of total gaseous mercury measured at
background sites. Atmos. Environ. 32(5), 883-893.
Ruijgrok W., Tieben H., Eisinga P. [1997] The dry deposition of particles to a forest canopy: A comparison of
model and experimental results. Atmos. Environ. 31, 399-415.
Ryaboshapko A. and Ilyin I. [2001] Mercury re-emission to the atmosphere in Europe. In: Transport and
chemical transformation in the troposphere. Proceedings of EUROTRAC Symposium 2000, GarmischPartenkirchen, Germany, 27-31 March 2000 (issued on enclosed compact disc).
Ryaboshapko A., Ilyin I., Bullock R., Ebinghaus R., Lohman K., Munthe J., Petersen G., Segneur C., Wangberg
I. [2001] Intercomparison study of numerical models for long-range atmospheric transport of mercury.
Stage I: Comparison of chemical modules for mercury transformations in a cloud/fog environment.
EMEP/MSC-E Technical report 2/2001, Meteorological Synthesizing Centre – East, Moscow, Russia.
Sander R. [1997] Henry's law constants available on the Web. EUROTRAC Newsletter 18, 24-25 (www.mpchmainz.mpg.de/~sander/res/henry).
Slinn W.G.N. [1982] Predictions for particle deposition to vegetative canopies. Atmos. Environ. 16, 1785-1794.
Sommar J., Gårdfeldt K., Strömberg D., Feng X. [2001] A kinetic study of the gas-phase reaction between the
hydroxyl radical and atomic mercury. Atmos. Environ. 35, 3049-3054.
Travnikov O. and I.Ilyin [2005] Regional Model MSCE-HM of Heavy Metal Transboundary Air Pollution in
Europe. EMEP/MSC-E Technical Report 6/2005, 59 pp.
Travnikov O., Ryaboshapko A. [2002]. Modelling of mercury hemispheric transport and depositions.
EMEP/MSC-E Technical Report 6/2002, Meteorological Synthesizing Centre - East, Moscow, Russia.
Walton J.J., MacCracken M.C. and Ghan S.J. [1988] A global-scale Lagrangian trace species model of
transport, transformation, and removal processes. J. Geophys. Res. 93(D7), 8339-8354.
Wesely M.L., Cook D.R., Hart R.L. [1985] Measurements and parameterization of particulate sulfur dry
deposition over grass. J. Geophys. Res. 90(D1), 2131-2143.
Williams R.M. [1982] A model for the dry deposition of particles to natural water surfaces. Atmos. Environ. 16,
1933-1938.
Zender C. S., Bian H. and D. Newman [2003] Mineral dust entrainment and deposition (DEAD) model:
description and 1990s dust climatology. Journal of Geophysical Research, vol. 108, No. D14, p. 4416.
Zhang H., Lindberg S.E., Marsik F.J., Keeler G.J. [2001] Mercury air/surface exchange kinetics of background
soils of the Tahquamenon River watershed in the Michigan upper peninsula. Water, Air and Soil Pollut
126, 151-169.
37