A Modeling Comparison of Mercury Deposition from Current

Article pubs.acs.org/est

A Modeling Comparison of Mercury Deposition from Current Anthropogenic Mercury Emission Inventories Francesco De Simone,† Christian N. Gencarelli,† Ian M. Hedgecock,*,† and Nicola Pirrone‡ †

CNR-Institute of Atmospheric Pollution Research, Division of Rende, UNICAL-Polifunzionale, 87036 Rende, Italy CNR-Institute of Atmospheric Pollution Research, Area della Ricerca di Roma 1, Via Salaria km 29,300, Monterotondo, 00015 Rome, Italy

‡

S Supporting Information *

ABSTRACT: Human activities have altered the biogeochemical cycle of mercury (Hg) since precolonial times, and anthropogenic activities will continue to perturb the natural cycle of Hg. Current estimates suggest the atmospheric burden is three to five times greater than precolonial times. Hg in the upper ocean is estimated to have doubled over the same period. The Minamata convention seeks to reduce the impact human activities have on Hg releases to the environment. A number of the Articles in the Convention concern the development of detailed inventories for Hg emissions and releases. Using the global Hg chemical transport model, ECHMERIT, the influence of the anthropogenic emission inventory (AMAP/UNEP, EDGAR, STREETS) on global Hg deposition patterns has been investigated. The results suggest that anthropogenic Hg emissions contribute 20−25% to present-day Hg deposition, and roughly two-thirds of primary anthropogenic Hg is deposited to the world’s oceans. Anthropogenic Hg deposition is significant in the North Pacific, Mediterranean and Arctic. The results indicate immediate reductions in Hg emissions would produce benefits in the short term, as well as in the long term. The most impacted regions would be suitable to assess changes in Hg deposition resulting from implementation of the Minamata convention. are currently freely available, AMAP/UNEP (2010),4 EDGAR,6 and STREETS.7 The STREETS inventory is based on the AMAP/UNEP (2005) inventory,8,9 which has been scaled according to the total Hg emissions in Streets et al.,10 for use in the GEOS-Chem model (http://geos-chem.org/). Due to the differences between the AMAP/UNEP (2005)9 and AMAP/ UNEP (2010)4 inventories, in terms of the methodology used11 as well as the relative importance of coal combustion for power generation and artisanal and small scale gold mining (ASGM),4 The AMAP/UNEP (2005)9 inventory has also been considered for this study. Hereafter these inventories are referred to as AMAP-2010, EDGAR, STREETS and AMAP-2005, the links where the emissions data may be obtained can be found in the Acknowledgments. The distribution of Hg emissions between Hg0(g), and reactive Hg species, oxidized Hg compounds and Hg associated with particulate matter (HgII(g) and HgP), as well as the emission height, differ between the inventories, (see Section 2.3). As these characteristics influence Hg deposition patterns their impact has also been investigated. Recently a number of modified anthropogenic inventories have been proposed, to better match observations using an inverse modeling method,12 to understand all time Hg releases,13 or to explain trends

1. INTRODUCTION Much of the Hg present in the environment today is the result of past anthropogenic activity,1 with coal combustion, metal refining, gold and silver extraction, and a variety of industrial processes all having made a significant impact. Current levels of Hg in the atmosphere and oceans are significantly perturbed with respect to precolonial times: the atmospheric burden is estimated to be 300−500% higher, and the concentrations in the upper ocean 200% higher.2−4 Consequently the reductions in the use and emissions of Hg proposed in the Minamata Convention (http://www.mercuryconvention.org/) will take place within a context of the continuous recycling of previously emitted Hg between environmental compartments. Most Hg input into ecosystems is the result of atmospheric deposition, both dry and wet. Hence the atmosphere is fundamental in distributing Hg, and the lifetime of the elemental form (Hg0(g)) of 8−12 months ensures that this occurs on a global scale. To assess the impact of future changes in anthropogenic emissions of Hg on the overall environmental burden, it is important to constrain the uncertainties associated with Hg emissions to and deposition from the atmosphere. In a previous article5 the impact of the uncertainties in biomass burning inventories on Hg emission and subsequent deposition were investigated using a global chemical transport model (CTM), ECHMERIT. Adopting a similar approach, the influence of the choice of anthropogenic Hg emission inventory used in the model has been investigated. Three recent global Hg emission inventories © XXXX American Chemical Society

Received: February 9, 2016 Revised: April 22, 2016 Accepted: April 27, 2016

A

DOI: 10.1021/acs.est.6b00691 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

ASGM mean this source is difficult to define. The AMAP-2005 inventory (on which the Streets inventory is based), included emissions from ASGM but used the “distribution mask” from large scale gold production to distribute the emissions.9 The AMAP-2010 inventory uses a more realistic distribution as does EDGAR, and this results in proportionally increased emissions in the Tropics and the Southern Hemisphere, as can be seen in the side panel of Figure 1. A discussion of the major uncertainties in currently available inventories, can be found in Kwon and Selin.17 2.2. Model Setup and Simulations. The global Hg chemical transport model ECHMERIT19,20 is based on the fifth generation General Circulation Model ECHAM5.21,22 ECHMERIT was run using T42 horizontal resolution (roughly 2.8° × 2.8° at the equator) and 19 vertical levels up to 10 hPa. The AMAP-2010 inventory was converted to the NetCDF format, the EDGAR, STREETS and AMAP-2005 inventories are available in NetCDF. The Climate Data Operators (CDO) mass conserving remapping tool23 was then used to interpolate the inventories onto the ECHMERIT grid. A first set of simulations, referred to hereafter as Full, was performed using as input all Hg emissions, which include anthropogenic, marine and biomass burning emissions. Monthly biomass burning Hg emissions were mapped to CO emissions from the FINNv1 inventory24 using a globally averaged enhancement ratio,25 as described in De Simone et al.5 Emissions from oceans were calculated online in the model, as described in De Simone et al.,20 and prompt re-emission of deposited Hg was included as in Selin et al.26 A spin-up period of 4 years was employed and the results from the fifth year, 2010, were used for this study,regardless of the nominal reference year of each inventory. As there remains some uncertainty concerning atmospheric Hg oxidation pathways,27−29 simulations were run using an O3/OH oxidation scheme and an alternative Br based oxidation mechanism. Oxidant fields were imported from the Mozart model30 for the O3/OH simulations, and from pTomcat31,32 for the Bromine fields. In addition to these simulations two tracer simulations assuming a fixed atmospheric lifetime for Hg0(g)5 were performed, as were simulations in which the height distribution was varied as described in Section 2.3. A summary of the simulations performed can be found in Supporting Information Table S1. The results of these simulations were used for comparison with observed atmospheric Hg concentrations and wet deposition flux measurements. To evaluate the Hg deposition which results directly from human activities, further simulations were performed which included only anthropogenic emissions, and in which the prompt re-emission mechanism was not used. These simulations are referred to hereafter as Only Anthr.. 2.3. Height Distribution of Emissions. For Hg sources that predominantly emit elemental Hg, (Hg0(g)) such as biomass burning, the vertical distribution of the emissions in the model levels does not have a significant effect on the eventual Hg deposition pattern, see De Simone et al.5 For shorter lived Hg species, HgII(g) and HgP, the model level into which they are placed is likely to influence deposition. The three inventories used here all take different approaches, AMAP/UNEP gives three height ranges, up to 50 m, between 50 and 150 m and above 150 m. The EDGAR inventory provides emissions divided into sectors following the Intergovernmental Panel on Climate Change (IPPC) identification code (http://edgar.jrc. ec.europa.eu/faq6.php), without specific information on the height distribution. The IPCC sectors were mapped to

observed in measurements by including emissions from commercial products containing Hg,14 and by constraining emissions and speciation.15 Adaptations of the global inventories have been also proposed for specific regions, to consider specific technologies and procedures, leading to significant differences in domestic emissions estimates for China as well as for other areas.16,17 However, these inventories are either not readily available in a gridded format or not entirely tested by the scientific community. Moreover the primary focus of this study is the major differences found in publicly available inventories widely used in the Hg modeling community. Most human exposure to Hg is through the consumption of fish and therefore the differences in the calculated deposition fluxes to the oceans is assessed. It goes without saying that a wider monitoring network for Hg in precipitation would be beneficial to constrain uncertainties in Hg fluxes, and thus also in emissions. Ultimately better constrained models will lead to better estimates of the time scales of environmental response times to future reductions in anthropogenic Hg emissions.

2. MATERIALS AND METHODS 2.1. Anthropogenic Emission Inventories. The details of the three principal inventories used have been described in detail in the literature,4,6,7,10 including the emission sector definitions considered by each one. A detailed comparison of the sectors included in the AMAP-2010 and EDGAR inventories, where they overlap and where sector groupings differ can be found in Supporting Information Table S9 of Muntean et al.6 A summary of the major characteristics of the inventories, relevant to the scope of this study can be found in Table 1. As can be seen they are broadly similar in most Table 1. Emission Inventory Characteristics inventory AMAP2010 EDGAR STREETS

resolution

total Hg emitted Mg

Hg0(g)

HgII(g)

HgP

2010

0.5° × 0.5°

1960

81%

14%

5%

2008 2005

0.1° × 0.1° 1° × 1°

1287 1900

72% 58%

22% 36%

6% 6%

reference year

respects. Emission height and speciation have been identified as being important parameters in regional modeling studies, but on a global scale they tend to be less so. The differences in Hg speciation can lead to marked differences in local concentrations of HgII(g) and HgP particularly in regional models.18 In global models where the spatial resolution is more coarse the effect is less noticeable, and will depend on the height distribution employed to partition the emissions between model levels, see Section 2.3. Only the AMAP/UNEP inventories give a vertical distribution for the emissions, they are divided into three height ranges and differentiate between 0 II Hg(g) , Hg(g) and HgP. The EDGAR inventory provides speciated emissions divided into sectors identified using standard IPCC codes (see http://edgar.jrc.ec.europa.eu/faq6. php), and therefore a height distribution can be inferred, see section 2.3. The spatial distribution of the emission inventories does differ as can be seen in Figure 1. One of the major reasons for this, is the uncertainty involved in the estimation of emissions from ASGM. Hg emission from ASGM is assumed to be Hg0(g), which given its lifetime, can be transported far away from sources, the unregulated and at times temporary nature of B


Article


Figure 1. Geographical distribution and latitudinal profile of annual Hg anthropogenic emissions as estimated by STREETS, EDGAR and AMAP2010 inventories. The units for both figures are μ g m−2 y−1.

the vertical emission distribution on the final deposition fields, see Supporting Information Table S1.

Standardized Nomenclature for Air Pollutants (SNAP) categories and the emission height table used in the Unified EMEP model33 was followed. The STREETS inventory gives no information regarding emission height or emission sector, they are therefore distributed throughout the BL as in the GEOS-Chem model (www.geos-chem.org). The distributions above are referred to as the “native” distributions. In addition simulations in which different vertical distributions were imposed were performed, in order to evaluate the impact of

3. RESULTS 3.1. Emission Inventory Comparison. Figure 1 shows the geographical distribution and the latitudinal profile of the total anthropogenic Hg emissions from the AMAP-2010, EDGAR, and STREETS inventories. The total annual Hg emission is similar in the AMAP-2010 and STREETS inventories at around 1900 Mg, whereas the EDGAR inventory has a noticeably C


Article

Environmental Science & Technology Table 2. Comparison of the Full Simulations with Global Observations for 2010 TGM

wet deposition

regression

stats

regression

stats

inventory

investigation

intercept

slope

r

NRMSE (%)

intercept

slope

r

NRMSE (%)

AMAP-2010 EDGAR STREETS AMAP-2005 AMAP-2010 EDGAR STREETS

O3/OH O3/OH O3/OH O3/OH bromine bromine bromine

0.4 0.3 0.2 0.1 −0.2 −0.1 −0.1

0.85 0.83 0.96 1.12 0.82 0.82 0.94

0.86 0.83 0.87 0.85 0.83 0.79 0.84

13.1 14.3 14.3 18.0 14.7 16.7 15.4

4.5 4.6 5.4 4.7 7.1 6.9 7.8

0.23 0.18 0.27 0.27 0.08 0.06 0.15

0.23 0.18 0.21 0.24 0.07 0.06 0.11

12.8 12.1 16.6 14.5 15.1 14.3 18.4

events.20,22 None of the three inventories significantly (α = 0.05) outperforms the others with respect to the regression slopes. However, the STREETS inventory has the greatest error in both oxidation scenarios. Differently to the comparison with measured TGM, the performance of the models depends significantly on the oxidation mechanisms. These results clearly indicate that the deposition is the most appropriate variable to measure and analyze to assess ecosystem impact and model accuracy. 3.3. Modeled Hg Deposition. 3.3.1. Latitudinal Variation. While the latitudinal profile of the emissions shows separate peaks and two cut-offs at the latitudinal limits of significant industrial human activity (45°S and 75°N) (Figure 2(a)), the profile of the total deposition never reaches zero at any latitude (Figure 2(b) and (c)). The latitudinal profiles of Hg deposition due to anthropogenic emissions (Only Anthr.) simulations are very similar in the Northern Hemisphere (NH) above 30°N, with a broad peak centered around the latitude of the major NH industrial regions (≈40°N). This is the case for both the O3/OH and the Br oxidation mechanisms, and also whether considering solely anthropogenic or total Hg emissions. South of 30°N the oxidation mechanism used in the simulations has a marked influence on the deposition profile, as can be seen in Figures 2(b) and (c). The O3/OH simulations give peaks in deposition almost equal to the NH industrial region peak just above and below the equator (all sources, the peaks are less pronounced in the anthropogenic only simulations) with a fairly steady decrease southwards toward the Antarctic region. The deposition profile in the SH in the Br simulation differs because the Br concentration profile is more constant with latitude when compared to the O3/OH concentration, and in particular does not peak around the equator. The broad peak in the Full simulation is a result of the combination of marine evasion of Hg and significant Br concentrations between roughly 15 and 60°S. The latitudinal deposition profiles when only anthropogenic Hg emissions are considered begin to differ below roughly 40°N. Below this latitude the AMAP-2010 inventory predicts roughly twice the deposition flux compared to the simulations using the STREETS inventory, with the EDGAR inventory resulting in fluxes between the two, irrespective of the oxidation mechanism employed. The simulations using the O3/OH mechanism however simulate proportionally higher deposition south of 40°N. The differences in the latitudinal profiles when all Hg emissions sources are included highlight even more markedly the influence of the oxidation mechanism used in the simulations. Using the O3/OH mechanism gives results with peaks either side of the equator, Figure 2(b). In Figure 2 the deposition profiles have been normalize to the value of the peak

lower total of 1300 Mg. However, as the most recent AMAP/ UNEP report4 estimates that anthropogenic emissions could be between 1010 and 4070 Mg annually, the EDGAR total is well within the range of uncertainty.6 The geographical locations of the emissions are similar among the inventories, with the greatest sources distributed over the industrialized areas of North America, Europe, South and East Asia. The most notable difference is the higher emissions in the AMAP-2010 inventory at tropical latitudes in Africa and western South America. This difference is the direct result of the revised estimation of emissions due to ASGM in this inventory, which are globally estimated to be higher than those from coal combustion.4 However, in an updated study concerning anthropogenic Hg emissions levels from China Zhang et al.,16 suggested that the AMAP-2010 inventory for this region significantly overestimates the ASGM sector and underestimates the coal industrial consumption one, due to the lack of official data about the activity levels of such two sectors. The latitudinal profiles in the right-hand panel of Figure 1 show the differences between the three inventories clearly. The AMAP-2010 and EDGAR inventories have broader distributions, the AMAP2010 inventory having higher emissions in equatorial and tropical regions due to its higher estimation of emissions from ASGM. The STREETS inventory has almost all emissions in the northern hemisphere centered around 40°N. The correlation (Pearson r) between the latitudinal profiles of the inventories is ≥0.86 in all cases, however the agreement map in Supporting Information Figure S1 shows how much they differ in terms of the location of areas of higher emissions. Supporting Information Figure S1 indicates the areas in the world where the emissions are greater than μ + σ (average + the standard deviation). With the exception of the industrial regions of the northern hemisphere the three inventories agree on the areas of highest emissions in very few instances. 3.2. Model Performance. The results from the Full simulations (Section 2.2) for the year 2010 were compared to available measurement data (see Supporting Information S4 and S3). A statistical summary of the comparison for Total Gaseous Mercury (TGM, the sum of Hg0(g) and HgII(g)) and for Hg wet deposition from the main runs can be found in Table 2. The results of the comparison indicate a good agreement between measured and simulated TGM in the different model runs. The inclusion in the model of AMAP-2010, EDGAR, or STREETS inventories does not lead to statistically significant (α = 0.05) different regression slopes, even when using a different oxidation mechanism. However, looking at the Hg wet deposition fluxes, the results of the comparison show a lower correlation than with TGM concentrations, due to the wellknown difficulties models using coarse horizontal and/or vertical resolution exhibit in simulating correctly precipitation D


Article


China,36 data from the tropics would also give an insight into the atmospheric oxidation mechanisms which determine the atmospheric cycle of Hg. To illustrate this point Figure 3 shows the simulated latitudinal distribution of wet deposition using the different

Figure 3. Latitudinal profile of the simulated wet deposition flux using STREETS, EDGAR and AMAP-2010 inventories in the Full runs for both oxidation mechanisms adopted: Bromine and O3+OH.

inventories and the two oxidation mechanisms Full runs. Notwithstanding the differences in the inventories the profiles make it quite clear that wet deposition fluxes would be expected to be roughly similar in the tropics and NH if Br is globally responsible for atmospheric Hg0(g) oxidation. If it is the case that O3/OH are the major oxidants then the Hg wet deposition flux would be roughly twice as high in the tropics than in northern midlatitudes. 3.3.2. Geographical Distribution. The spatial distribution of the simulated total Hg deposition flux is shown in Supporting Information Figure S2 for the Full simulations whereas the same maps for the Only Anthr. simulations can be found in Supporting Information Figure S3. The differences between the inventories appear limited, using the STREETS inventory produces higher deposition in Europe, while overall the results using the EDGAR inventory are generally lower everywhere due to the lower total emission estimate. The STREETS inventory also results in somewhat lower deposition in the SH generally. The major differences in the deposition fields depend on the choice of the oxidation mechanism. In particular the simulations with Br as the oxidant have distinctly higher deposition fluxes over the Northern Pacific and somewhat higher over the North Atlantic. The deposition to the equatorial Pacific is lower in the Br simulations compared to the O3/OH simulations. The simulated fluxes to the ocean basins is discussed further in Section 3.4. The agreement maps in Figure 4 are used to highlight the similarities and differences in the spatial distribution of the regions of high Hg deposition, defined here as greater than the average plus one standard deviation (>μ + σ). The primary colors highlight those areas where only one of the inventories simulated fluxes greater than μ + σ, whereas the gray areas indicate the regions where the results of the simulations with all three inventories predict particularly high Hg deposition. In the Only Anthr. simulations the AMAP-2010 inventory shows more regions where only it predicts high deposition, and these are mostly in the equatorial region. These simulations have regions in which all the inventories agree, East Asia, Europe and the Eastern U.S., independent of the oxidation mechanism employed. This is

Figure 2. Latitudinal profiles of mercury emissions (a) and normalized total deposition that result from model runs adopting O3/OH (b) and Bromine (c) driven oxidation mechanism for the different inventories. Panels (b) and (c) show the deposition obtained from Full and Only Anthr. simulations. Deposition profiles are not to scale.

at 35°N which coincides with the peak in the anthropogenic emission profiles. These peaks are higher than the deposition peak at 40°N in the case of the AMAP-2010 and EDGAR inventories and only slightly lower using the STREETS inventory. By comparison in the simulations using the Br oxidation mechanism the highest deposition south of the equator is between 45 and 60% of the NH peak deposition value. The O3/OH oxidation mechanism gives a deposition profile where relative to the emissions profile, the peak is smoothed and spread significantly further south throughout the whole of the Tropics, this is particularly evident in the case of the STREETS inventory. The deposition profile in the case of the Br oxidation mechanism reflects the emissions profile more closely in the NH, but in the SH creates a broad deposition between 15 and 60°S reflecting the relatively higher oxidant levels at these latitudes. Both oxidation mechanisms simulate deposition beyond the latitudinal range of the emissions, with similar relative fractions in the Arctic, see Section 3.4. The Br mechanism gives a much higher relative fraction below 60°S over the Southern Ocean and Antarctica. The case for the expansion of monitoring Hg in precipitation to cover equatorial and tropical regions is clear, and not only to quantify the impact that Hg deposition has on ecosystems in this region. Combined with the ongoing monitoring in North America,34 Europe35 and E


Article


Figure 4. Agreement maps of Hg deposition fields obtained by including STREETS, EDGAR, and AMAP-2010 inventories in Full (a and c) and Only Anthr. simulations (b and d), for both oxidation mechanisms. The maps show the areas where deposition is > μ + σ.

have a distinct influence on the simulated Hg deposition (as has been observed with regional18,37,38 and global models26,39). Therefore, a number of sensitivity runs were performed, varying the oxidation mechanism, the emission inventory, speciation and height distribution, see Supporting Information Table S1. In order to obtain an estimate of the impact current anthropogenic emissions have on the Hg deposition flux the results from a number of the sensitivity simulations were combined in an ’ensemble’. The similarity of the deposition fields resulting from the sensitivity runs was assessed using the horizontal pattern correlation method40,41 and the nonparametric Kolmogorov−Smirnov two-sample test. The results are shown in Supporting Information Table S2. Values of ProbKS‑test ≤ 0.05 indicate that it is improbable that the simulated Hg deposition fields belong to the same distribution. The simulations used to construct the ensemble are indicated in Supporting Information Table S2. The ProbKS‑test values showed Only Anthr. simulations demonstrate significant differences in the modeled deposition fields, for all the parameters that were changed. However, to construct the ensemble the less realistic simulations; all emissions in a single model layer, fixed lifetime, and all emissions as HgII(g), were excluded. The ensemble deposition field of the Only Anthr. simulations is shown in Figure 5(a). Figure 5(a) clearly shows that the major impact from anthropogenic emissions is deposition downwind of the major source regions, and is particularly significant over East Asia, and the northern Pacific and Atlantic oceans. Figure 5(b) compares the ensemble of the Only Anthr. simulations with the ensemble of the same Full simulations. It can be seen that the anthropogenic contribution to Hg deposition in the Tropics is around 15% and reaches 25% over significant areas of the northern hemisphere. Due to

predominantly due to reactive species emitted directly from anthropogenic sources, which are rapidly deposited over these regions (see Supporting Information Figure S4). However, while both mechanisms predict high deposition over the northwest Pacific this region extends to the western Pacific in the Br simulations but not in the O3+OH simulations. Considering all Hg emissions leads to generally broader agreement, with the exception of the STREETS inventory which gives higher deposition in the NH compared to the other two inventories. The AMAP-2010 and EDGAR inventories predict higher deposition in the southern Indian Ocean, the far south Atlantic and the western south Pacific than does the STREETS inventory. The analysis of the geographical distribution of the deposition supports the findings considering the latitudinal profiles (Section 3.3.1) about the fate of Hg emitted from anthropogenic activities under the different oxidation scenarios. Bromine allows Hg emitted in the tropical areas to be transported toward higher latitudes in both hemisphere before being deposited. In the case of the O3/ OH oxidation mechanism deposition occurs mostly either over areas near emission hot spots or immediately downwind of them (see Figure 4) or within the tropics area, where oxidant concentrations are high. 3.4. Uncertainty and Impact Assessment. A previous study of Hg emissions from biomass burning showed that the factors which had the most impact on Hg deposition fields were the emission inventory and the gas phase oxidation mechanism used in the simulations.5 These factors also influence the simulated Hg deposition fields in this study, however in the case of biomass burning5 Hg emissions were assumed to be Hg0(g). When considering anthropogenic emissions the speciation of the emissions combined with the emission height also F


Article


Mediterranean, has been calculated and is summarized in Table 3. In spite of all the uncertainties and unknowns related to the mercury, the ensemble of simulations suggest that 20−25% of the deposition over the Arctic is due to the anthropogenic activities. The Arctic and the Mediterranean Seas are also the basins where the anthropogenic emission impact is proportionally greatest.

4. DISCUSSION The three inventories in this study agree on the geographical locations of the anthropogenic Hg emission hotspots over industrialized regions of North America, Europe, South and East Asia, although regional studies show some differences in emission estimates and source sector attribution.16,42,43 The EDGAR and AMAP-2010 inventories have the most similar geographical distributions, and in comparison to the STREETS inventory have a higher proportion of emissions at more southerly latitudes. This is due mostly to the recent reevaluation of the importance of ASGM as a source of Hg to the atmosphere which was not so evident in the AMAP-2005 inventory on which STREETS is based although, a recent study suggests that the ASGM contribution in the AMAP-2010 inventory could be overestimated with respect to industrial coal combustion.16 The inclusion of Hg releases from commercial activities is another source which should be taken into consideration.14 The oxidation mechanism used in the simulations has an important influence on the fate of anthropogenic Hg, enhancing the deposition and the differences among the inventories in tropical regions for O3/OH, or allowing the transport of Hg to higher latitudes in the case of the Bromine oxidation mechanism. Primary Hg anthropogenic emissions are responsible for up to 23% of the total Hg deposition in some areas where the ecosystem is already under significant anthropogenic and/or climate pressure, such as the Arctic and the Mediterranean Basin. The northern hemisphere is unsurprisingly more influenced by anthropogenic emissions it is also therefore the region which will benefit most from emission reduction. This seems to the case in the North Atlantic where it is believed that the Ocean is already responding to a reduction in riverine inputs which occurred from the 1970s onward.44 Although on a global scale the

Figure 5. Geographical distribution of the total Hg deposition from anthropogenic emissions only obtained from an ensemble of simulations for the year 2010 (a) in terms of the average (μ) and standard deviation σ of the ensemble. The comparison of the Only Anthr. and Full simulations (b) shows the contribution to total deposition from anthropogenic sources.

the importance of the Oceans in the Hg cycle and because the major human route of Hg exposure is through fish consumption, the deposition to Ocean basins and also the Table 3. Annual Mercury Deposition (Mg) to the Oceansa

basins

a

%

North

South

North

South

Indian

Med.

Southern

Arctic

O3/OH

Atlantic

Atlantic

Pacific

Pacific

Ocean

Sea

Ocean

Ocean

AMAP-2010 EDGAR streets O3/OH Full ensemble avg. impact (%)

149 112 171 842 17

104 62 74 723 11

425 317 511 2392 17

268 160 192 1828 11 basins

190 117 147 1185 13

9 7 14 44 23

15 8 10 106 10

34 26 43 149 23

land

sea

35 37 39 30

65 63 61 70 %

North

South

North

South

Indian

Med.

Southern

Arctic

Br

Atlantic

Atlantic

Pacific

Pacific

Ocean

Sea

Ocean

Ocean

AMAP-2010 EDGAR streets Br Full ensemble avg. impact (%)

158 120 181 882 17

113 66 77 774 11

466 352 554 2574 18

249 144 172 1660 11

180 106 134 1084 13

12 9 16 54 22

38 21 23 306 9

34 26 42 195 17

land

sea

33 35 38 28

67 65 62 72

The Arctic has been defined as the region north of 66°N, and the Southern Ocean as south of 60°S. G


Article


europa.eu/edgar_v4tox1/index.php STREETS, ftp://ftp.as. harvard.edu/gcgrid/data/ExtData/HEMCO/MERCURY/ v2014-09/STREETS/ AMAP-2005, http://eccad.sedoo.fr.

emission height and speciation of Hg emissions have a relatively limited effect on the simulations, they are far more important on a regional scale and should be better constrained for the assessment of local versus regional/hemispheric sources for regions or ecosystems at particular risk. As this study and a previous study of the impact of Hg emissions from biomass burning the oxidation mechanism is of great importance in determining the spatial distribution of Hg deposition. Continued monitoring, particularly at tropical latitudes and particularly of Hg in precipitation have the potential to give greater understanding of the mechanisms driving the global atmospheric Hg cycle. The relatively high deposition from anthropogenic sources seen in the Arctic and Mediterranean regions, and also to a slightly lesser extent, the North Pacific, suggest that these regions would be sensible choices for monitoring the eventual impact of the implementation of the Minamata Convention. Article 22 of the Convention is entitled “Effectiveness Evaluation”, point 2 states “To facilitate the evaluation, the Conference of the Parties shall, at its f irst meeting, initiate the establishment of arrangements for providing itself with comparable monitoring data on the presence and movement of mercury and mercury compounds in the environment as well as trends in levels of mercury and mercury compounds observed in biotic media and vulnerable populations”. In order to identify trends it would clearly be auspicious to begin monitoring Hg, ideally in precipitation as soon as possible, and the regions identified in this study would be expected to respond most strongly to changes in anthropogenic emissions.

■

■

(1) Amos, H. M.; Sonke, J. E.; Obrist, D.; Robins, N.; Hagan, N.; Horowitz, H. M.; Mason, R. P.; Witt, M.; Hedgecock, I. M.; Corbitt, E. S.; Sunderland, E. M. Observational and Modeling Constraints on Global Anthropogenic Enrichment of Mercury. Environ. Sci. Technol. 2015, 49, 4036−4047. (2) Soerensen, A. L.; Sunderland, E. M.; Holmes, C. D.; Jacob, D. J.; Yantosca, R. M.; Skov, H.; Christensen, J. H.; Strode, S. A.; Mason, R. P. An Improved Global Model for Air-Sea Exchange of Mercury: High Concentrations over the North Atlantic. Environ. Sci. Technol. 2010, 44, 8574−8580. (3) Mason, R. P.; Choi, A. L.; Fitzgerald, W. F.; Hammerschmidt, C. R.; Lamborg, C. H.; Soerensen, A. L.; Sunderland, E. M. Mercury biogeochemical cycling in the ocean and policy implications. Environ. Res. 2012, 119, 101−117. (4) AMAP/UNEP. Technical Background Report for the Global Mercury Assessment 2013;2013. http://www.unep.org/ hazardoussubstances/Mercury/Informationmaterials/ ReportsandPublications/tabid/3593/Default.aspx. (5) De Simone, F.; Cinnirella, S.; Gencarelli, C. N.; Yang, X.; Hedgecock, I. M.; Pirrone, N. Model Study of Global Mercury Deposition from Biomass Burning. Environ. Sci. Technol. 2015, 49, 6712−6721. (6) Muntean, M.; Janssens-Maenhout, G.; Song, S.; Selin, N. E.; Olivier, J. G.; Guizzardi, D.; Maas, R.; Dentener, F. Trend analysis from 1970 to 2008 and model evaluation of EDGARv4 global gridded anthropogenic mercury emissions. Sci. Total Environ. 2014, 494−495, 337−350. (7) Corbitt, E. S.; Jacob, D. J.; Holmes, C. D.; Streets, D. G.; Sunderland, E. M. Global source-receptor relationships for mercury deposition under present-day and 2050 emissions scenarios. Environ. Sci. Technol. 2011, 45, 10477−10484. (8) Pacyna, E.; Pacyna, J.; Sundseth, K.; Munthe, J.; Kindbom, K.; Wilson, S.; Steenhuisen, F.; Maxson, P. Global emission of mercury to the atmosphere from anthropogenic sources in 2005 and projections to 2020. Atmos. Environ. 2010, 44, 2487−2499. (9) AMAP/UNEP. Technical Background Report to the Global Atmospheric Mercury Assessment;2008. http://www.unep.org/ hazardoussubstances/Mercury/Informationmaterials/ ReportsandPublications/tabid/3593/Default.aspx. (10) Streets, D. G.; Zhang, Q.; Wu, Y. Projections of Global Mercury Emissions in 2050. Environ. Sci. Technol. 2009, 43, 2983−2988. (11) Steenhuisen, F.; Wilson, S. J. Identifying and characterizing major emission point sources as a basis for geospatial distribution of mercury emissions inventories. Atmos. Environ. 2015, 112, 167−177. (12) Song, S.; et al. Top-down constraints on atmospheric mercury emissions and implications for global biogeochemical cycling. Atmos. Chem. Phys. 2015, 15, 7103−7125. (13) Streets, D. G.; Devane, M. K.; Lu, Z.; Bond, T. C.; Sunderland, E. M.; Jacob, D. J. All-time releases of mercury to the atmosphere from human activities. Environ. Sci. Technol. 2011, 45, 10485−10491. (14) Horowitz, H. M.; Jacob, D. J.; Amos, H. M.; Streets, D. G.; Sunderland, E. M. Historical mercury releases from commercial products: Global environmental implications. Environ. Sci. Technol. 2014, 48, 10242−10250. (15) Zhang, Y.; Jacob, D. J.; Horowitz, H. M.; Chen, L.; Amos, H. M.; Krabbenhoft, D. P.; Slemr, F., St.; Louis, V. L.; Sunderland, E. M. Observed decrease in atmospheric mercury explained by global decline in anthropogenic emissions. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, 526−531. (16) Zhang, L.; Wang, S.; Wang, L.; Wu, Y.; Duan, L.; Wu, Q.; Wang, F.; Yang, M.; Yang, H.; Hao, J.; Liu, X. Updated emission inventories for speciated atmospheric mercury from anthropogenic sources in China. Environ. Sci. Technol. 2015, 49, 3185−3194.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.est.6b00691. Tables summarizing the simulations performed with ECHMERIT, the spatial correlation and ProbKS‑test values between simulations, and the measurements of TGM air concentrations and of the Hg wet deposition fluxes used for the comparison with model simulation. Maps of total deposition from anthropogenic only and full emission simulations using both oxidation mechanisms. Agreement maps of total Hg emissions and of deposition due to Hg reactive species only (PDF)

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*Phone: +39 0984 493213; fax: +39 0984 493215; e-mail: ian. [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We are grateful to Sebastian Rast and his staff at the Max Planck Institute for Meteorology in Hamburg, Germany for the distribution of their software ECHAM5 and for providing the access to the processed ERA-INTERIM data. We are grateful to Xin Yang for providing the Br/BrO fields from p-Tomcat. The research was performed in the framework of the EU project GMOS (FP7−265113). Tables S3 and S4 are based on tables in a GMOS report prepared by Oleg Travnikov (Meteorological Synthesizing Centre-East (MSC-E) of EMEP, Moscow, Russia). Data Sources: AMAP-2010, http://www.amap.no/ mercury-emissions/datasets) EDGAR, http://edgar.jrc.ec. H


Article

Environmental Science & Technology (17) Kwon, S. Y.; Selin, N. E. Uncertainties in Atmospheric Mercury Modeling for Policy Evaluation. Current Pollution Reports 2016, 1−12. (18) Bieser, J.; De Simone, F.; Gencarelli, C.; Geyer, B.; Hedgecock, I.; Matthias, V.; Travnikov, O.; Weigelt, A. A diagnostic evaluation of modeled mercury wet depositions in Europe using atmospheric speciated high-resolution observations. Environ. Sci. Pollut. Res. 2014, 21, 9995−10012. (19) Jung, G.; Hedgecock, I. M.; Pirrone, N. ECHMERIT v1.0 − a new global fully coupled mercury-chemistry and transport model. Geosci. Model Dev. 2009, 2, 175−195. (20) De Simone, F.; Gencarelli, C. N.; Hedgecock, I. M.; Pirrone, N. Global atmospheric cycle of mercury: a model study on the impact of oxidation mechanisms. Environ. Sci. Pollut. Res. 2014, 21, 4110−4123. (21) The atmospheric general circulation model ECHAM 5. PART I: Model description, MPI-Report No. 349,2003. https://www.mpimet. mpg.de/fileadmin/publikationen/Reports/max_scirep_349.pdf. (22) Roeckner, E.; Brokopf, R.; Esch, M.; Giorgetta, M.; Hagemann, S.; Kornblueh, L.; Manzini, E.; Schlese, U.; Schulzweida, U. Sensitivity of Simulated Climate to Horizontal and Vertical Resolution in the ECHAM5 atm Model. J. Clim. 2006, 19, 3771−3791. (23) Climate Data Operators. 2014; Max-Plank Institute fur Meteorologie,2014. https://code.zmaw.de/projects/cdo. (24) Wiedinmyer, C.; Akagi, S. K.; Yokelson, R. J.; Emmons, L. K.; Al-Saadi, J. A.; Orlando, J. J.; Soja, A. J. The Fire INventory from NCAR (FINN): a high resolution global model to estimate the emissions from open burning. Geosci. Model Dev. 2011, 4, 625−641. (25) Friedli, H. R.; Arellano, A. F.; Cinnirella, S.; Pirrone, N. Initial estimates of mercury emissions to the atmosphere from global biomass burning. Environ. Sci. Technol. 2009, 43, 3507−3513. (26) Selin, N. E.; Jacob, D. J.; Yantosca, R. M.; Strode, S.; Jaeglé, L.; Sunderland, E. M. Global 3-D land-ocean-atmosphere model for mercury: Present-day versus preindustrial cycles and anthropogenic enrichment factors for deposition. Global Biogeochem. Cycles 2008, 22, GB2011. (27) Hynes, A. J.; Donohoue, D. L.; Goodsite, M. E.; Hedgecock, I. M. Mercury Fate and Transport in the Global Atmosphere: Emissions, Measurements and Models; Pirrone, N., Mason, R. P., Eds.; Springer, 2009; Chapter 14, pp 427−457. (28) Subir, M.; Ariya, P. A.; Dastoor, A. P. A review of uncertainties in atmospheric modeling of mercury chemistry I. Uncertainties in existing kinetic parameters − Fundamental limitations and the importance of heterogeneous chemistry. Atmos. Environ. 2011, 45, 5664−5676. (29) Subir, M.; Ariya, P. A.; Dastoor, A. P. A review of the sources of uncertainties in atmospheric mercury modeling II. Mercury surface and heterogeneous chemistry − A missing link. Atmos. Environ. 2012, 46, 1−10. (30) Emmons, L. K.; et al. Description and evaluation of the Model for Ozone and Related chemical Tracers, version 4 (MOZART-4). Geosci. Model Dev. 2010, 3, 43−67. (31) Yang, X.; Cox, R. A.; Warwick, N. J.; Pyle, J. A.; Carver, G. D.; O’Connor, F. M.; Savage, N. H. Tropospheric bromine chemistry and its impacts on ozone: A model study. J. Geophys. Res. 2005, 110, 1984−2012. (32) Yang, X.; Pyle, J. A.; Cox, R. A.; Theys, N.; Van Roozendael, M. Snow-sourced bromine and its implications for polar tropospheric ozone. Atmos. Chem. Phys. 2010, 10, 7763−7773. (33) Simpson, D.; et al. The EMEP MSC-W chemical transport model − technical description. Atmos. Chem. Phys. 2012, 12, 7825− 7865. (34) Gay, D. A.; Schmeltz, D.; Prestbo, E.; Olson, M.; Sharac, T.; Tordon, R. The Atmospheric Mercury Network: measurement and initial examination of an ongoing atmospheric mercury record across North America. Atmos. Chem. Phys. 2013, 13, 11339−11349. (35) Tørseth, K.; Aas, W.; Breivik, K.; Fjæraa, A. M.; Fiebig, M.; Hjellbrekke, A. G.; Lund Myhre, C.; Solberg, S.; Yttri, K. E. Introduction to the European Monitoring and Evaluation Programme (EMEP) and observed atmospheric composition change during 1972− 2009. Atmos. Chem. Phys. 2012, 12, 5447−5481.

(36) Fu, X. W.; Zhang, H.; Yu, B.; Wang, X.; Lin, C.-J.; Feng, X. B. Observations of atmospheric mercury in China: a critical review. Atmos. Chem. Phys. 2015, 15, 9455−9476. (37) Gencarelli, C. N.; De Simone, F.; Hedgecock, I. M.; Sprovieri, F.; Yang, X.; Pirrone, N. European and Mediterranean mercury modelling: Local and long-range contributions to the deposition flux. Atmos. Environ. 2015, 117, 162−168. (38) Kos, G.; Ryzhkov, A.; Dastoor, A.; Narayan, J.; Steffen, A.; Ariya, P. A.; Zhang, L. Evaluation of discrepancy between measured and modelled oxidized mercury species. Atmos. Chem. Phys. 2013, 13 (9), 4839−4863. (39) Amos, H. M.; et al. Gas-particle partitioning of atmospheric Hg(II) and its effect on global mercury deposition. Atmos. Chem. Phys. 2012, 12, 591−603. (40) Santer, B. D.; Taylor, K. E.; Wigley, T. M.; Penner, J. E.; Jones, P. D.; Cubasch, U. Towards the detection and attribution of an anthropogenic effect on climate. Clim. Dyn. 1995, 12, 77−100. (41) Santer, B. D.; Taylor, K. E.; Wigley, T. M. L.; Johns, T. C.; Jones, P. D.; Karoly, D. J.; Mitchell, J. F. B.; Oort, A. H.; Penner, J. E.; Ramaswamy, V.; Schwarzkopf, M. D.; Stouffer, R. J.; Tett, S. A search for human influences on the thermal structure of the atmosphere. Nature 1996, 382, 39−46. (42) Wu, Y.; Wang, S.; Streets, D. G.; Hao, J.; Chan, M.; Jiang, J. Trends in anthropogenic mercury emissions in China from 1995 to 2003. Environ. Sci. Technol. 2006, 40, 5312−5318. (43) Wu, Q.; Wang, S.; Zhang, L.; Song, J.; Yang, H.; Meng, Y. Update of mercury emissions from China’s primary zinc, lead and copper smelters, 2000−2010. Atmos. Chem. Phys. 2012, 12, 11153− 11163. (44) Soerensen, A. L.; Jacob, D. J.; Streets, D. G.; Witt, M. L. I.; Ebinghaus, R.; Mason, R. P.; Andersson, M.; Sunderland, E. M. Multidecadal decline of mercury in the North Atlantic atmosphere explained by changing subsurface seawater concentrations. Geophys. Res. Lett. 2012, 39, L21810.

I


A Modeling Comparison of Mercury Deposition from Current

Recommend Documents