Chemosphere 67 (2007) 12871295

www.elsevier.com/locate/chemosphere

Climate change inuence on POPs distribution

and fate: A case study
M. Dalla Valle
b

a,b,*

, E. Codato b, A. Marcomini

a,b

a
Consorzio Venezia Ricerche, c/o VEGA, Via della Liberta` 5/12, 30175 Marghera VE, Italy
Department of Environmental Sciences, University of Venice, Calle Larga S.ta Marta, 2137, I-30123 Venice, Italy

Received 7 June 2006; received in revised form 22 October 2006; accepted 6 December 2006
Available online 26 January 2007

Abstract
Climate change has the potential of aecting the behaviour and distribution of organic pollutants, including POPs. Direct eects of
climate change, like temperature increase, modication of wind and precipitation patterns, sea level rise, snow and ice cover, may be very
eective in altering the partitioning of POPs among the environmental compartments. Other consequences of future climate scenarios
may imply the alteration of degradation rates, soil properties (and hence land use), air-particle partitioning of chemicals and so forth.
A case study is here presented to illustrate the major implications of climate change on the long term at the local scale. A dynamic multimedia model was applied to selected PCB and PCDD/F congeners to simulate the eects of climate change on their distribution and
uxes over the next 50 y in the Venice Lagoon (Italy). Dierent climate change scenarios were tested, nding noticeable variations in
POPs concentration even for minor environmental changes. PCBs and PCDFs environmental concentrations may dier by a factor
two in a moderate climate change scenario, compared to a situation with stable climate over the next 50 y. However, model results also
suggest that if global warming may have the potential of reducing the environmental levels of these chemicals, it would probably enhance
their mobility and hence their potential for long range atmospheric transport.
 2006 Elsevier Ltd. All rights reserved.
Keywords: Climate change; POPs; Model; Venice Lagoon

1. Introduction
The reconstruction of the historical climate trends of our
planet demonstrates how climate is constantly changing,
showing peculiar oscillations at dierent time scales. Climate is far from being constant, even in the short term,
as the alternation of glacial and post-glacial periods is
accompanied by less dramatic short-term climate oscillations that nonetheless can result in signicant alterations
of ecosystems and living organism distribution. In addition, interannual variations, which depend also on solar
activity (that varies by up to 0.1% in intensity, with an
11 y cycle), and other oscillations have been observed since

Corresponding author. Tel.: +39 0415093184; fax: +39 0415093074.

E-mail address: [email protected] (M. Dalla Valle).

0045-6535/$ - see front matter  2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.chemosphere.2006.12.028

experimental observation started (IPCC, 2001). Climate

change seems to have accelerated over the last century,
probably because of the increase of greenhouse gas emissions (e.g. CO2, CH4, HFCs, PFCs) and hence of their
atmospheric concentrations. Recent studies estimate that
the global mean temperature increased by 0.6 0.2 C during the 20th century, and predict a further increase between
1.4 C and 5.8 C by the year 2100 (IPCC, 2001).
POPs are known to undergo long range atmospheric
transport (LRAT), travelling according to their physical
chemical properties and to the characteristics of the environment that they encounter (soil properties, climate, wind
direction and speed and so forth), reaching remote regions
where they have never been produced or used. This process
can happen as a simple emission-transport-deposition
event or following a series of hops by deposition onto
a surface and subsequent emission (Wania and Mackay,

1288

M. Dalla Valle et al. / Chemosphere 67 (2007) 12871295

1996). The transport distance and the number of air-surface exchange episodes depend on the type of surface (soil,
water, vegetation, etc.) and on the physicalchemical properties of the given chemical. According to this principle,
persistent chemicals with higher volatility will undergo
LRAT followed by deposition in distant areas (e.g. arctic
regions), while those with lower vapour pressure will be
deposited preferably in areas closer to the emission source.
Many other mechanisms and factors are able to inuence the distribution of POPs in the atmosphere: the capacity of the environmental compartments to accumulate or to
degrade POPs; the atmospheric circulation patterns; the
spatial and temporal distribution of primary sources and
the kinetic of the air-surface exchange. POPs in air can
be degraded by photodegradation and via reactions with
hydroxyl radical (OH ), both depending, to a certain
extent, on the solar radiation intensity.
Temperature is one of the main factors able to inuence
the global distribution of POPs, even if several other processes play an important role. Processes directly depending
on temperature are: emission rates from primary and secondary sources, the gas-particle partitioning of POPs in
air, reactions rates (biodegradation, photolysis and oxidation in air), the air-surface exchange (Halsall et al., 1999).
The degradation rates, for example, are estimated to
increase by a factor of two to three for every 10 C increment (Lammel et al., 2001; Macdonald et al., 2005). In
addition, important atmospheric depletion mechanisms
for organic contaminants are photochemical degradation
and the degradation initiated by the hydroxyl radical,
whose concentration is related to solar activity and cloud
cover (Mandalakis et al., 2003).
In general, the capacity of an environmental compartment to retain and release chemicals, which depends on
several processes, is widely variable both spatially and temporally (according to daily, seasonal and long term trends).
The distribution of POPs among the environmental compartments depends also on their properties and on the
capacity of the exchange surfaces (vegetation, soil, water)
to retain them (Mackay, 2001). Therefore, the eects of
possible future climatic changes, which can imply, for
example, the acceleration of desertication, changes in land
use, the reduction of glacier surface, can signicantly inuence global POPs cycling (e.g. Dalla Valle et al., 2004;
Macdonald et al., 2005).
Recently, an attempt to estimate the spatial and temporal variability of such capacity has been made for the
oceans (Jurado et al., 2004), surcial soil (Dalla Valle
et al., 2005a) and vegetation (Dalla Valle et al., 2004).
The spatial variability of this capacity depends on the combination of several factors like organic carbon content (in
soil or within plant leaves), green biomass, or more generally on the ecosystem type. Those factors even can vary
remarkably within a limited space (e.g. along a mountain
slope), thus inuencing POPs distribution according to
their properties (Grimalt et al., 2004; Van Droodge et al.,
2004; Daly and Wania, 2005). The temporal variability is

mainly dependant on temperature (e.g. Hornbuckle and

Eisenreich, 1996; Halsall et al., 1999), therefore showing
daily and seasonal oscillations, but also on biomass trends
and biogeochemical cycles (e.g. phytoplankton cycles;
Dachs et al., 2002), which can be seasonal or exhibiting a
long term trend, and on soil use and properties. Another
complication is given by the wide range of physicalchemical properties of the dierent classes of chemicals.
Other climate related variables such as the frequency of
storm surges, precipitations, the seasonality and the intensity of the atmospheric events, can inuence the fate of
POPs in the environment. An increase of precipitations,
for example, can cause an increase of POPs deposition onto
the soil. More frequent storm surges may enhance the
mobilisation of chemicals stored in the soil compartment,
which can be transported by land runos, making them
available to the aquatic organisms.
It is important then, especially at the local scale, to take
fully into account all the possible environmental changes
when trying to predict the future environmental distribution of a given persistent chemical. Unfortunately, the possible inuences of climate change and of its human
responses, have not been extensively studied so far, despite
the great attention that climate change related subjects are
receiving from various disciplines of the scientic community. To our knowledge, only a few comprehensive studies
have been published so far and they are entirely dedicated
to the Arctic region (e.g. Macdonald et al., 2003, 2005).
Other recent studies highlighted the possible inuences of
climate change and of interannual climate variability on
POPs distribution and uxes (Ma et al., 2004; Eisenreich,
2005; Macleod et al., 2005). Here we provide a semi-quantitative assessment of the possible inuence of climate
change on the distribution and fate of some selected chemicals in a temperate region, the Venice Lagoon area. Therefore, the main objective of this study was to track the
distribution of some selected chemicals in the Lagoon
environment over the next 50 y, focusing on the inuence
of climate change by considering three dierent climate scenarios. This study is aimed mainly at demonstrating how
even relatively modest climate and environmental changes
can trigger a much more signicant variation in POPs distribution and behaviour by applying a reliable model to a
well known case study.
2. The case study area
The Venice Lagoon, a large coastal lagoon located along
the Adriatic Sea coast, is a complex and frail ecosystem
surrounded by a densely populated area and an important
industrial district.
This area has been the focus of several investigations
over the last decade and a wide range of environmental
characteristics have been monitored for several years. As
a result, the area is known in detail. The environmental distribution of POPs has also been studied by means of several
experimental (e.g. Bellucci et al., 2000; Frignani et al.,

M. Dalla Valle et al. / Chemosphere 67 (2007) 12871295

2001; Guerzoni et al., 2004) and modelling activities (Dalla

Valle et al., 2003, 2005b). The main POP sources to the
Lagoon are represented by the several kinds of human
activities located in and around the Lagoon itself, in particular the emissions from the industrial district of Porto
Marghera, the cities of Venice and Mestre, and the various
civil and agricultural activities in the drainage basin. The
semi-volatile pollutants emitted in the atmosphere may
undergo transport outside the region but can be signicantly deposited over the Lagoon area, according to their
chemical properties and release mode. On the other hand,
chemicals introduced directly into the Lagoon waters, from
industries or residential areas (e.g. the historical city of
Venice has no sewage system) can contaminate the bottom
sediment but also be distributed in the Lagoon waters and
enter the food chain. The Venice Lagoon has a supercial
area of approximately 550 km2, of which approximately
418 km2 is subject to the tides of the Adriatic Sea. The
Lagoon is connected to the sea through three inlets: Lido,
Malamocco and Chioggia. The average depth is about 1 m
and the tidal excursion is ca. 30 cm. Climatic changes
observed over the last century have had repercussions also
on the Venice Lagoon. In particular, the frequency of
storm surges in the historical centre of Venice has increased
considerably due to sea level rise and to the subsidence of
the historical centre itself. Recent studies estimated a relative sea level rise of about 26 cm over the last century (e.g.
Bonato et al., 2001). As far as the precipitations, the trend
observed in Northern Italy in the period from 1951 to 1996,
indicates a reduction of the number of days with precipitations (approximately 14% less), mainly in winter. On the
other hand, precipitation episodes have become more
intense. There is a lack of comprehensive studies on recent
temperature trends although there is some evidence that
the global trends reported by the IPCC (2001) are observed
also in the Venice area (Barbi et al., 2005).
The Venice Lagoon is a transition environment, hence
interested by a continuous natural evolutionary process
and it is surrounded by a densely populated area, that
has determined a long history of human interventions.
Future POPs distribution and uxes in the area will be thus
inuenced both by direct eects of climate change (e.g.
increased temperature, reduced precipitations, submersion
of marshlands) and by indirect eects (e.g. land use
changes).
The estimation of climate change impact on the distribution of POPs in the Lagoon area requires the analysis of the
eects caused by climate perturbations at the local level,
and then their possible interactions with POPs. POPs
behaviour, distribution, and their dependency on climate
may be studied and simulated with the aid of various modelling tools.
3. Methods
Fugacity based models are particularly useful to predict
the mobility and persistence of a given chemical, known its

1289

properties and the characteristics of the environmental

compartments. In order to assess the possible future
behaviour of some selected POPs in the Venice Lagoon,
a level IV dynamic model (Mackay, 2001) was developed
and applied to the area. The considered compounds were
PCB 118, PCB 180, 2,3,7,8-TCDF, 1,2,3,4,7,8-HCDF.
These were chosen among all PCB and PCDD/F congeners for which sucient data (i.e. concentration levels) was
available for the study area. In fact, for several PCB and
PCDD/F congeners, and in particular for the lighter
PCB ones, most of the measured concentrations were
below the detection limit and a comparison between measured and modelled data would be impossible. This model
represents an evolution of a previous dynamic model successfully used to reconstruct the historical trends of
selected POPs in the area (Dalla Valle et al., 2005b). The
model dominion was, like in the previous modelling exercise, limited to the central part of the Lagoon, because
of its relative homogeneity and because of its very detailed
characterisation. This area, comprised between the Malamocco and Lido channels, has a total surface of about
170 km2 and a mean depth of ca. 2.5 m and can be considered to be quite well mixed. The model considers ve compartments, namely air, soil, sediment, water and suspended
particulate matter (SPM). The SPM was considered as an
individual compartment since a previous study proved that
POPs found in water particles in the Lagoon area are not
in equilibrium with the water column, but rather with the
sediment (Dalla Valle et al., 2003). This is probably due to
very high re-suspension and deposition rates (up to some
kg m2 day1). In addition, this compartment, whose concentration is typically in the range of ca. 30300 mg/l,
seems to be responsible for the re-distribution of POPs
within the Lagoon area, in particular transporting the
chemicals from the inner part of the Lagoon (more polluted) towards the outer parts (Frignani et al., 2001; Dalla
Valle et al., 2003).
The model was run in non-equilibrium conditions and
inputs to the Lagoon have been estimated from experimental data obtained by recent investigations carried out in the
area (more details in Dalla Valle et al., 2005b). In particular, air emissions have been estimated by using data produced by a sampling campaign of atmospheric deposition
(Guerzoni et al., 2004), while water emissions have been
estimated from the results of a project devoted to the monitoring of the riverine input (Collavini et al., 2005). Despite
the constant eorts in reducing POP emissions following
international treaties and national laws, residual sources
like remaining spills from old transformers and landlls,
combustion processes, natural sources, etc., are still significant and cannot be ignored. For this reason, inputs to the
Lagoon were assumed to decrease linearly over the 50 y of
the simulation (from 2000 to 2050) by a factor 10. However, the aim of this study is not to provide a quantitative
and fully realistic concentration estimates for the studied
chemicals, but rather to test a modelling approach and to
evaluate the inuence of future climate change on POPs

1290

M. Dalla Valle et al. / Chemosphere 67 (2007) 12871295

dynamics, hence this assumption should be considered reasonable enough for the purposes of this study.
The model requires information on the relevant physicalchemical properties of the studied chemicals (e.g.
Kow, vapour pressure, Kaw) and on their degradation rates.
Detailed information on the Lagoon environment is also
needed (e.g. the Lagoon mean depth and surface area,
the percentage of organic carbon in the sediment, temperature, and so on) together with information on advective
uxes and on emission trends. Some relevant environmental parameters describing the model domain are shown in
Table 1 and have been previously discussed in detail by
Dalla Valle et al. (2003).
The model is described by ve mass balance equations
written in terms of fugacity. The fugacity values, and hence
the concentrations in the individual compartments, was
calculated by solving the following system of equations
(Eqs. (1)(5)):
Air:

SPM:

where V stands for volume, Z for fugacity capacity and f

for fugacity. The subscripts are: b for bulk phase, A for
air, W for water, Sed for sediment, SPM for suspended particulate matter, r for reaction or degradation, and j for
advection. Intermedia D values are denoted by Dij for
transport from compartment i to the compartment j. Emissions are denoted with E.
In order to estimate the impact of the climatic changes
on POPs in the Lagoon environment, three dierent scenarios (A, B, C) were tested. In scenario A, dened as the
control scenario, environmental conditions have been
considered constant along the 50 y of the simulation, with
constant temperature (288 K, the annual mean temperature
in Venice), precipitations (900 mm/y), solar radiance and
degradation rates. Scenario B assumes a modest and gradual evolution of the initial conditions along the simulation
run, with temperature increasing from 288 K to 289 K, precipitations decreasing by 5% of the degradation rate
increasing by 10%. In the last scenario (C), more dramatic
variations were assumed, with the mean annual temperature reaching 291 K, precipitations decreasing by 10%
and degradation rate increasing by 30%. All the assumptions are consistent with the climate change scenarios
envisaged by the latest IPCC climate change assessment
report (IPCC, 2001).
Degradation rates are dicult to estimate as they
depend on several factors and they are the sum of dierent
and competing reaction processes (e.g. photolysis, hydrolysis, redox reactions and microbial degradation). Normally
the inuence of temperature is estimated by calculating
the contribution of the single removal mechanisms and
applying the Arrhenius equation (Eq. (6)) to them:

V bSPM Z bSPM dfSPM = dt DWSPM fW DSedSPM fSed

k Ae RT

V bA Z bA dfA = dt EA DSA fS DWA fW

 DrA DASPM DAS Daw fA

Soil:
V bS Z bS dfS = dt DAS fA  DrS DSA DSW fS

Water:
V bW Z bW dfW = dt EW Daw fA DSW fS DSedW fSed
DSPM fSPM  DrW DjW
DWA DWSed DWSPM fW

Ea

DASPM fA DSSPM fS
 DrSPM DSPMW
DSPMSed fSPM

Sediment:
V bSed Z bSed dfSed = dt DWsed fw DSPMSed fSPM
 DrSed DSedW DSedSPM
Dburial fSed

Table 1
Relevant environmental parameters used in the model (Dalla Valle et al.,
2003)
Parameter

Value

Total surface
SPM concentration
Sediment deposition rate
Active sediment depth
Sediment organic carbon fraction
SPM organic fraction
Water residence time
Mean depth

132 km2
65 mg/l
7.45 105 m/h
15 cm
0.010
0.0165
5 days
2.5 m

where A is a constant and Ea is the activation energy. Here

we chose to adopt a simplied approach by assuming hypothetical yet realistic reaction rate changes for the B and C
scenarios (Lammel et al., 2001; Macdonald et al., 2005).
All the temperature dependant physicalchemical
parameters (Henrys law constant, vapour pressure, partition coecients) were assumed to vary according to the
temperature variation. For the PCBs, the equations used
for the calculation of the chemicalphysical parameters as
a function of temperature were (Eqs. (7)(9)):
log H AH  BH =T

log Ps AP  BP =T

log K OA AK  BK =T

where Ai and Bi are empirical dimensionless constants

(Harner and Bidleman, 1996; Paasivirta et al., 1999).
Since empirical data about the temperature dependency
of the Henrys law constant was not available for PCDFs,
the equation used was (Gusev et al., 2005; Eq. (10)):



1
1
H H 0 exp aH

10
T T0

M. Dalla Valle et al. / Chemosphere 67 (2007) 12871295

log K ow Az  Bz log S L

11

where Az and Bz are dimensionless empirical constant

(Paasivirta et al., 1999), while SL is the solubility
(mol/m3). Paasivirta et al. (1999) reported also the temperature dependence of solubility (Eq. (12)):
log S L AS  BS =T

12

4. Results and discussion

In this study, four PCB and PCDF congeners were considered: PCB 118, PCB 180, 1,2,3,4,7,8-HCDF and 2,3,7,8TCDF. Given the initial concentration of each congener in
the individual environmental compartments (Table 2), the
emission rates, the relevant properties of the chemical
and of the compartments, the nal concentrations (t =
50 y) were estimated. The initial concentrations in the compartments were either estimated from experimental data or
obtained by previous modelling exercises. Thus, the mean
sediment concentration measured during a recent sampling
campaign (Magistrato alle Acque di Venezia, 2000) was set
as initial concentration (t = 0) for the central Lagoon sediment. Water and SPM concentrations were calculated
assuming their equilibrium with the sediment, as demonstrated in a previous study (Dalla Valle et al., 2003).

Table 2
Concentrations at t = 0 and at t = 50 y for the three tested scenarios
Concentration
PCB 118
Initial (t = t0)
Scenario A (t = t0 + 50 y)
Scenario B (t = t0 + 50 y)
Scenario C (t = t0 + 50 y)

Sediment
(ng/kg)

Water
(pg/l)

SPM
(ng/kg)

1040
9.1
7.4
5.5

57
0.6
0.5
0.4

547
59
50
41

PCB 180
Initial (t = t0)
Scenario A (t = t0 + 50 y)
Scenario B (t = t0 + 50 y)
Scenario C (t = t0 + 50 y)

441
210
170
115

TCDF
Initial (t = t0)
Scenario A (t = t0 + 50 y)
Scenario B (t = t0 + 50 y)
Scenario C (t = t0 + 50 y)

4.5
1.9
1.6
1.0

0.27
0.10
0.08
0.06

2.5
0.97
0.79
0.52

1,2,3,4,7,8-HCDF
Initial (t = t0)
Scenario A (t = t0 + 50 y)
Scenario B (t = t0 + 50 y)
Scenario C (t = t0 + 50 y)

16.3
7.0
5.5
3.4

0.86
0.36
0.29
0.17

8.2
3.5
2.8
1.7

75
11
9.1
6.2

227
107
87
59

Finally, as experimental measurements were not available

for air concentration, its value was set to be the same calculated for the year 2000 by a previous modelling simulation (Dalla Valle et al., 2005b).
Before analysing the dierences between the three
scenarios (A, B an C), it is worth to look in more detail
at the model results for what we dened as the control
scenario (A). The only input parameter allowed to change
over time in this scenario is the chemical input into the
system, which is supposed to decrease linearly by a factor
10 over the 50 y of the model run. This implies a
rapid decrease in concentration in all the environmental
compartments.
A remarkable decrease in concentration over the 50 y of
the simulation was observed in all the compartments and
for all the studied chemicals. The compartment exhibiting
the slowest concentration decline is the Lagoon sediment,
with about a factor two decrease over the simulation run
for PCB 180, TCDF and HCDF. The concentration of
PCB 118 (Fig. 1) has a much more pronounced declining
trend (two orders of magnitude) during the same time
interval, probably because of its higher volatility and a faster degradation rate. We can therefore expect time trends
to vary widely according to the considered chemicals.
The most rapid decline was observed in air concentration,
because of its higher degradation rate compared to the
other compartments and because of advective ows. Concentration in air decreases by 9099.5% depending on the
chemical, although in this case the two PCBs show a slower
decline compared to PCDFs. Concentrations in SPM and
therefore in bulk water show a trend very similar to that
of the sediment due to a very rapid exchange with the sediment itself, determining a near-equilibrium condition
among these compartments. Water concentration decreases
linearly following the water input trend, showing how its
concentration essentially depends only on the inputs to
this compartment. Given the high hydrophobicity of
the chemicals here studied, without a direct input, water

1E-09
Air
Sediment
Soil
Water
SPM

1E-10

Fugacity (Pa)

where H0 is the Henrys law constant value (Pa m3 mol1)

at T0, aH is a constant, T is the temperature (K) and
T0 = 283.15 K. The temperature dependency of Kow was
obtained from the following empirical relationship (Eq.
(11)):

1291

1E-11

1E-12

1E-13

t (years)

Fig. 1. PCB 118 fugacity in the individual environmental compartments

for scenario A (logarithmic scale).

1292

M. Dalla Valle et al. / Chemosphere 67 (2007) 12871295

concentration would rapidly drop because of a rapid

exchange with the sediment.
What has been described so far reects well known facts
about POPs behaviour in the environment and previous
modelling results for the same area (e.g. Dalla Valle
et al., 2003). Interesting observations can be made by comparing the temporal trends predicted by the model in the
control scenario (A), against the two other scenarios
(B and C), which include a climate change perspective.
As explained in the previous section, scenario B introduces
only a limited and progressive climate evolution, while scenario C assumes more dramatic changes over the 50 y of
the simulation. The eects of temperature increase determine a signicant variation also of the fugacity capacity
of the compartments, which is an indicator of the capacity
of a given compartment to store a compound. The fugacity
capacity of sediment (Z) for PCB 118, for example,
decreases from ca. 65 000 mol Pa1 m3 to ca. 43 000
mol Pa1 m3 over the simulation run under the conditions
imposed in scenario C. A similar decrease is observed for
the fugacity capacity of soil and SPM. The fugacity capacity of the air compartment decreases only by 1% over the
same period. No signicant dierences are observed among
the four compounds.
The net result of all the changes in physicalchemical
properties and environmental characteristics is a relative
decrease in concentration for all the considered chemicals
in all the environmental compartments, with the exception
of air, in the scenarios B and C compared to scenario A.
The estimated sediment concentration for the C scenario
in the year 2050 is about 50% lower than that estimated
for the A scenario, with HCDF being the most sensitive
chemical to climate change. The concentration for scenario
B is somewhat intermediate, being ca. 20% lower compared
to the control scenario (Fig. 2 compares the concentration
of PCB 180 for the three scenarios). A similar behaviour is
observed for SPM and bulk water, while air concentration
is, at the end of the model run, highest in scenario C (+3

Table 3
Comparison between the concentrations predicted by the model under the
B and C scenarios against Scenario A at the end of the model run
(t = 50 y)

500

Concentration (ng/kg)

10% depending on the chemical) and slightly higher than in

scenario A for the B scenario (between +1% and 4%). This
is due to the increased volatility of the chemicals with
increasing temperature and it is more evident for PCB
118, the most volatile of the four chemicals. Water concentration shows only limited dierences among the scenarios,
decreasing between 6% and 9% in the more extreme conditions (C). Dierences among the studied chemicals are
quite limited, although in general HCDF shows a slightly
greater response to altered climate conditions, except for
the air compartment. Table 3 shows the concentration
ratios at t = 50 y of the scenarios B and C compared to
the control scenario (A).
The main reservoir of POPs in the Lagoon environment
is represented by the sediment, therefore concentration
trends in this compartment are representative of the overall
budget of such chemicals for the area.
The concentration trends in the sediment show a significantly dierent behaviour in the three scenarios. Concentrations are obviously equal at t0 (year 2000), but then
the curves dierentiate quite rapidly, especially between
the two extreme scenarios (A and C), for which the dierence is the greatest. In particular, the concentration is lowest for the scenario C, given a faster degradation process
and a more pronounced volatility. Although the concentration decline of PCB 118 (Fig. 2) is much faster if compared
to that of the other studied congeners, because of a higher
degradation rate and more signicant advective losses due
to a higher volatility, their trends are rather similar and
their response to climate change are roughly the same.
We can then infer that probably the behaviour of the
PCB and PCDD/F congeners not considered by this modelling exercise would be very similar, with a decrease in
concentration depending mainly on their degree of chlorination. Concentrations for the individual environmental
compartments at the beginning and at the end of the simulation runs are shown for the three scenarios and for all
the four chemicals here considered in Table 2.

400

300

200
Scenario A
Scenario B
Scenario C

100

0
0

10

20

30

40

50

t (years)

Fig. 2. PCB 180 concentration in sediment for the three hypothesized

climate scenarios.

Ratios (at t = 50 y)

Air

Sediment

Water

SPM

PCB 118
[PCB118]B/[PCB118]A
[PCB118]C/[PCB118]A

1.04
1.10

0.82
0.60

0.97
0.92

0.86
0.69

PCB 180
[PCB180]B/[PCB180]A
[PCB180]C/[PCB180]A

1.03
1.06

0.81
0.55

0.97
0.91

0.81
0.55

TCDF
[TCDF]B/[TCDF]A
[TCDF]C/[TCDF]A

1.03
1.08

0.81
0.54

0.97
0.92

0.81
0.54

1,2,3,4,7,8-HCDF
[HCDF]B/[HCDF]A
[HCDF]C/[HCDF]A

1.01
1.03

0.78
0.48

0.98
0.94

0.78
0.48

M. Dalla Valle et al. / Chemosphere 67 (2007) 12871295

5. Sensitivity analysis
A sensitivity analysis was performed in order to assess
the inuence on the model output of the single input
parameters. This can provide useful information on the
studied environment by providing an insight of the relative
importance of the individual environmental processes regulating chemicals behaviour and fate. The local sensitivity
(S) is normally dened as the relative deviation of the output value Y corresponding to a variation of an input
parameter Xi (Webster et al., 1998):
SX i oY =Y  X i =oX i
where oY is the absolute change of output value and oXi is
the input change. In some studies sensitivity is calculated
by varying more parameters at once, but in this case the
sensitivity associated to each input parameter has been calculated separately.
Therefore, sensitivity was calculated by doubling each
parameter and then comparing the corresponding output
concentration to the original model result. Here we report
the results for the following compartments: air, water, sediment and SPM (Table 4). Only the input parameters determining a non-negligible variation of the model output are
shown in the Table 4.
Only a few parameters exhibit a signicant inuence on
the model output, and in most cases the eect is more
evident only on one-two environmental compartments.
Degradation rates have dierent impacts on the compartments, mainly because these chemicals are stored for the

Table 4
Sensitivity analysis
Parameter

Degradation in air
Degradation in sediment
Degradation in soil
Degradation in water
Henrys law constant
Vapour pressure
Precipitations rate
Kow
Air-side airwater mtc
Burial rate
Deposition velocity
Sediment mixing rate
Sediment diusion
Sediment resunspension
rate
Washout ratio
Water residence time
Water runo
Water side airwater mtc
Water side water
sediment mtc

Output variation (%)

Air

Sediment

Water

SPM

0.0
0.0
7.5
0.0
4.6
100.0
41.3
0.0
4.7
0.0
22.9
2.3
0.1
0.0

0.0
61.9
2.2
0.0
1.4
0.1
0.0
21.7
1.3
3.9
0.0
1.7
0.0
0.2

0.0
0.2
0.0
0.0
61.8
0.1
0.0
0.1
19.4
0.0
0.0
0.0
0.4
0.0

0.0
60.9
2.2
0.0
1.4
0.1
0.0
22.1
1.3
3.8
0.0
1.7
0.0
98.0

41.3
8.0
0.0
2.4
0.0

0.0
2.4
0.0
0.7
0.0

0.0
32.6
0.0
10.0
0.0

0.0
2.4
0.0
0.7
0.0

The table shows the concentration variations obtained for each environmental compartment by doubling the individual input parameters.

1293

most part in the sediment and therefore only in this compartment degradation is a signicant depletion process.
For this reason, by doubling the degradation rate in sediment, a 62% decrease in sediment concentration is
observed, while a similar variation in water and air degradation rates has no signicant eect. The parameters with
the biggest inuence on the model outputs are, in addition
to the degradation rates in soil and sediment, the KOW of
the chemical, the air side airwater mass transfers coecient, the water residence time in the Lagoon, and temperature. Several parameters concur in determining air
concentration and the most important ones according to
the sensitivity analysis are: deposition velocity, the washout
ratio, the precipitation rate and the vapour pressure of the
chemical. It is important to remark that the variations
imposed have solely the objective of identifying the parameters that are more likely to aect the model output. This
means that also the uncertainty associated to the model
should be taken into account when prioritizing the eorts
made to improve the model accuracy. Some variables, for
example, may exhibit a great inuence on the model output
but they are known with good accuracy (e.g. water residence time, vapour pressure), while others may have a
minor inuence but they may be not suciently studied
or else they may have a great intrinsic variability (e.g. sediment mixing rate).
6. Conclusions
Despite the great uncertainty on the exact evolution of
current climate conditions, there is little doubt that even
relatively modest climate alterations will bring signicant
environmental changes. Chemical cycles will be also
aected in a variety of ways, and the interaction between
climate impacts and the possible adaptation/mitigation
options that will be adopted will increase the uncertainties.
In order to reduce the uncertainties and to have a realistic
prediction of chemicals behaviour in the medium-long
term, the development and application of suitable modelling tools is needed. As a rst exercise, the case study of
the Venice Lagoon was considered by simulating the
impact of climate change on the behaviour and the distribution of POPs in the area with the aid of a dynamic multimedia environmental model. This model proved to be a
useful and relatively simple tool to predict the distribution
of POPs in a complex system evolving over time. On the
basis of our model simulations, future POPs concentrations
in the environment, regardless the considered hypothetical
scenario, are bound to decrease quickly over the time.
However, remarkable concentration dierences among
the tested scenarios were observed, also for modest variations of the environmental conditions and without having
considered the possible secondary eects of climate change
(e.g. land use change, sea level rise, and so forth). While for
sediment and SPM substantial dierences are observed
among the three scenarios, with about a factor of two
dierence between nal concentrations in the two extreme

1294

M. Dalla Valle et al. / Chemosphere 67 (2007) 12871295

scenarios, water concentration decreases in the same way

regardless of the climatic conditions. This probably
because water and air concentration are still determined
mainly by direct inputs, and therefore there is no sucient
time to reach the equilibrium with the other compartments.
If the possible long term indirect eects of climate change
and their interactions with human intervention (mitigation
and adaptation policies) were taken into account, these
changes would probably have a far greater impact. This
will have implications also on the global circulation of
these chemicals. The net result of the tested scenarios,
which are however realistic, points towards a faster reduction of the overall POPs budget, but at the same time
an enhancement of their mobility. This may result in a
greater transfer to polar regions, where degradation and
removal from the environment is more dicult for such
chemicals, causing an increase in bioaccumulation and
biomagnication.
The studied PCB and PCDF congeners seem to be representative of their respective classes, however, the behaviour of other organic chemicals, covering a wider range
of properties should be investigated by future modelling
activities. The complex interaction of the properties of a
given chemical and the evolution of climate and environmental scenarios may, in fact, trigger dierent and not easily predictable behaviours.
In addition, future modelling activities in this eld
should include the adoption of spatially resolved data,
combining a detailed characterisation of the relevant environmental properties (e.g. land use, biomass production,
etc.) with their possible future evolution under dierent
future climate scenarios.
Acknowledgements
The authors gratefully acknowledge the Euro-Mediterranean Centre for Climate Change (CMCC; Lecce, Italy)
for nancial support.
References
Barbi, A., Millini, R., Monai, M., Soa, S., 2005. Meteo-climatic features
of the Venice Lagoon. In: Fletcher, C.A., Spencer, T. (Eds.), Flooding
and Environmental Challenges for Venice and its Lagoon: State of
Knowledge. Cambridge University Press, UK, p. 691.
Bellucci, L.G., Frignani, M., Raccanelli, S., Carraro, C., 2000. Polychlorinated dibenzo-p-dioxins and dibenzofurans in surcial sediment of
the Venice Lagoon (Italy). Marine Pollut. Bull. 40, 6576.
Bonato, N., Ferla, M., Umgiesser, G., Zen, G., 2001. Tidal observations in
the Venetian Lagoon. Update on sea level change from 1872 to 2000.
In: Proceedings of the nal workshop of Cost Action 40. Sea level in
Europe: observation, interpretation and exploitation. Dubrovnik,
Croatia, 1921 September 2001.
Collavini, F., Bettiol, C., Zaggia, L., Zonta, R., 2005. Pollutant loads from
the drainage basin to the Venice Lagoon (Italy). Environ. Int. 31, 939
947.
Dachs, J., Lohmann, R., Ockenden, W.A., Mejanelle, L., Eisenreich, S.J.,
Jones, K.C., 2002. Oceanic biogeochemical controls on global
dynamics of persistent organic pollutants. Environ. Sci. Technol 36,
42294237.

Dalla Valle, M., Marcomini, A., Sfriso, A., Sweetman, A.J., Jones, K.C.,
2003. Estimation of PCDD/F distribution and uxes in the Venice
Lagoon, Italy: combining measurement and modelling approaches.
Chemosphere 51, 603616.
Dalla Valle, M., Sweetman, A.J., Dachs, J., Jones, K.C., 2004. The
Maximum Reservoir Capacity for vegetation: implications for global
cycling. Global Biogeochem. Cy. 18, GB4032.
Dalla Valle, M., Jurado, E., Sweetman, A.J., Jones, K.C., Dachs, J.,
2005a. The Maximum Reservoir Capacity of global surface soils for
persistent organic pollutants: implications for global cycling. Environ.
Pollut. 134, 153164.
Dalla Valle, M., Marcomini, A., Jones, K.C., Sweetman, A.J., 2005b.
Reconstruction of historical trends of PCDD/Fs and PCBs in the
Venice Lagoon, Italy. Environ. Int. 31, 10471052.
Daly, G.L., Wania, F., 2005. Organic contaminants in mountains.
Environ. Sci. Technol 39, 385398.
Eisenreich, S.J. (Ed.), 2005. Climate change and the European water
dimension. EU Report, 21553. European Commission: Rome, Italy.
p. 253.
Frignani, M., Bellucci, L.G., Carraro, C., Favotto, M., 2001. Accumulation of polychlorinated dibenzo-p-dioxins and dibenzofurans in
sediments of the Venice Lagoon and the industrial area of Porto
Marghera. Marine Pollut. Bull 42, 533544.
Grimalt, J., Borghini, F., Sanchez-Hernandez, J.C., Barra, R., Torres
Garcia, C.J., Focardi, S., 2004. Temperature dependence of the
distribution of organochlorine compounds in the mosses of the
Andean mountains. Environ. Sci. Technol. 38, 53865392.
Guerzoni, S., Rossini, P., Molinaroli, E., Rampazzo, G., Raccanelli, S.,
2004. Measurement of atmospheric deposition of polychlorinated
dibenzo-p-dioxins and dibenzofurans in the Lagoon of Venice, Italy.
Chemosphere 54, 13091317.
Gusev, A., Mantseva, E., Shatalov, V., Strukov, B., 2005. Regional
Multicompartment Model MSCE-POP. EMEP/MSC-E Technical
Report 5/2005. Annex A.
Halsall, C.J., Gevao, B., Howsam, M., Lee, R.G.M., Ockenden, W.A.,
Jones, K.C., 1999. Temperature dependence of PCBs in the UK
atmosphere. Atmos. Environ. 33, 541552.
Harner, T., Bidleman, T.F., 1996. Measurements of octanol-air partition
coecients for polychlorinated biphenyls. J. Chem. Eng. Data 41, 895
899.
Hornbuckle, K.C., Eisenreich, S.J., 1996. Dynamic of gaseous semivolatile
organic compounds in a terrestrial ecosystem eects of diurnal and
seasonal climate variations. Atmos. Environ. 30, 39353945.
IPCC, 2001. Climate change 2001: the scientic basis. In: Houghton, J.T.,
Ding, Y., Griggs, D.J., Noguer, M., van der Linden, P.J., Dai, X.,
Maskell, K., Johnson,C.A., (Eds.), Contribution of Working Group I
to the Third Assessment Report of the Intergovenrmental Panel on
Climate Change. Cambridge University Press, Cambridge, United
Kingdom and New York, NY, USA, p. 881.
Jurado, E., Lohmann, R., Meijer, S., Jones, K.C., Dachs, J., 2004.
Latitudinal and seasonal capacity of the surface oceans as a reservoir
of polychlorinated biphenyls. Environ. Pollut. 128, 149162.
Lammel, G., Feichter, J., Leip, A., Semeena, V.S., 2001. Long-range
transport and environmental fate of modern pesticides Model
development and application, In: Lammel, G., (Ed.), Slowly degradable organics in the atmospheric environment and airsea exchange
Report Max Planck Institute for Meteorology No. 335. Hamburg,
p. 5156.
Ma, J., Hung, H., Blanchard, P., 2004. How do climate uctuations aect
persistent organic pollutant distribution in North America? Evidence
from a decade of air monitoring. Environ. Sci. Technol. 38, 25382543.
Macdonald, R.W., Harner, T., Fyfe, J., 2005. Recent climate change in the
Arctic and its impact on contaminant pathways and interpretation of
temporal trend data. Sci. Total Environ. 342, 586.
Macdonald, R.W., Harner, T., Fyfe, J., Loeng, H., Weingartner, T., 2003.
AMAP Assessment 2002: The inuence of global change on contaminant pathways to, within, and from the Arctic. Arctic Monitoring and
Assessment Programme (AMAP), Oslo, Norway. pp. xii+65.

M. Dalla Valle et al. / Chemosphere 67 (2007) 12871295

Mackay, D., 2001. Multimedia Environmental Models, second ed. In: The
Fugacity Approach Lewis Publishers, Boca Raton, FL, USA, pp. 261.
Macleod, M., Riley, W.J., McKone, T.S., 2005. Assessing the inuence of
climate variability on atmospheric concentrations of polychlorinated
biphenyls using a global-scale mass balance model (BETR-Global).
Environ. Sci. Technol. 39, 67496756.
Magistrato alle Acque di Venezia, 2000. The mapping of the pollutants in
the Lagoon bottom sediments (in Italian). Final report. Consorzio
Venezia Nuova, Venice, Italy.
Mandalakis, M., Berresheim, H., Stephanou, E.G., 2003. Direct evidence
for destruction of polychlorobiphenyls by OH radicals in the
subtropical troposphere. Environ. Sci. Technol. 37, 542547.

1295

Paasivirta, J., Sinkkonen, S., Mikkelson, P., Rantio, T., Wania, F., 1999.
Estimation of vapor pressures, solubilities and Henrys law constants
of selected persistent organic pollutants as functions of temperature.
Chemosphere 39, 811832.
Van Droodge, B., Grimalt, J., Camarero, L., Catalan, J., Stuchlik, E.,
Torres Garcia, C.J., 2004. Atmospheric semivolatile organochlorine
compounds in European high-mountain areas (central Pyrenees and
high Tatras). Environ. Sci. Technol. 38, 35253532.
Wania, F., Mackay, D., 1996. Tracking the distribution of persistent
organic pollutants. Environ. Sci. Technol. 37, 13521359.
Webster, E., Mackay, D., Wania, F., 1998. Evaluating environmental
persistence. Environ. Toxicol. Chem. 17, 13441351.