On the Contribution of Biomass Burning to POPs (PAHs and PCDDs

Article pubs.acs.org/est

On the Contribution of Biomass Burning to POPs (PAHs and PCDDs) in Air in Africa G. Lammel,*,†,‡ A. Heil,§ I. Stemmler,†,∥ A. Dvorská,‡,⊥ and J. Klánová‡ †

Max Planck Institute for Chemistry, Hahn-Meitner-Weg 1, 55128 Mainz, Germany Masaryk University, Research Centre for Toxic Compounds in the Environment, Kamenice 5, 62500 Brno, Czech Republic § Helmholtz Research Centre Jülich, Institute for Energy and Climate Research, 52428 Jülich, Germany ‡

S Supporting Information *

ABSTRACT: Forest, savannah, and agricultural fires in the tropics and subtropics are sources for widespread pollution and release many organic substances into the air and soil, including persistent organic pollutants, i.e., polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) and polycyclic aromatic hydrocarbons (PAHs). The significance of this source for the exposure of humans and the environment in Africa toward phenanthrene, fluoranthene, pyrene, benzo(a)pyrene, 2,3,7,8-tetrachlorodibenzo-p-dioxin, 1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin, and octachlorodibenzo-p-dioxin is studied using daily global emissions from vegetation fires observed by satellite and a global multicompartment chemistry-transport model. Near-ground atmospheric concentrations of model-predicted vegetation fire related concentrations of PAHs and PCDDs were in the 10−1000 and 10−5−10−3 pg m−3 ranges, respectively. Vegetation fires in Africa are found to emit 180 ± 25 kg yr−1 of PCDD/Fs. By comparison with observations, it is found that fires explain 1−10% of the PCDD (5% of 2,3,7,8- tetrachlorodibenzo-p-dioxin) concentrations in the rural and background atmosphere of sub-Saharan Africa. The contribution of vegetation fires to exposure to PAH is probably >10%, but cannot be quantified due to lack of knowledge with regard to both emission factors and photochemistry. A sensitivity analysis suggests that the heterogeneous reaction of PAHs with ozone is effectively limiting atmospheric lifetime and long-range transport.

1. INTRODUCTION Biomass burning is a significant source of air pollution on various continents.1 Among the emitted trace substances, there are also persistent organic pollutants (POPs), i.e., polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) and polycyclic aromatic hydrocarbons (PAHs2). POPs are resisting rapid degradation in environmental compartments, partition among environmental compartments (multicompartmental substances), and most POPs are semivolatile (vapor pressures at 298 K in the range from 10−6 to 10−2 Pa). Although POPs criteria do not necessarily apply for many PAHs, PAHs are often considered POPs. PCDD/Fs and PAHs are a hazard for human health in Africa due to their toxicity,3 food contamination, and (in the case of PAH) inhalation.4 PCDD/Fs were found in soils of rural and remote locations of western and eastern Africa 5 and in air masses coming off western Africa.6 Both PCDD/Fs and PAHs are formed during combustion processes, and these represent their major source type in the atmospheric environment, except close to petrogenic, aluminum, iron, and steel industries (PAHs) and agriculture, as an impurity of pesticides, or are formed from pesticides under sunlight (PCDD/Fs7,8). In Africa, large areas of savannah, grass-, and cropland and forest burn every year. Most of the fires are anthropogenic fires, ignited for agricultural management practices with the aim to remove unwanted vegetation and to ameliorate soils for cropping and grazing.9 The fires are a major source of © 2013 American Chemical Society

regional-scale air pollution in tropical and subtropical Africa.10,11 In 2008, a year with average fire activity in Africa, 2.1 Pg of biomass (on a dry basis) was consumed by vegetation fires according to two satellite-borne vegetation fire emission inventories, the Global Fire Database (GFED3) 1 and the Global Fire Assimilation System (GFAS).12 According to these estimates, the fires in Africa contributed around 55% of the global amount of biomass consumed by vegetation fires in 2008. The vast majority (around 84−91%) of the biomass burned in Africa in 2008 was in savannah, wood, or grassland vegetation. Another approximately 9−14% were from forest fires. Agricultural fires (crop residue burning) contributed only around 1%. Burning of forest,13,14 crop residuals,15 and grassland/ savannah2 has been found to be a primary source of PCDD/ F and the PAH naphthalene in the extratropics2 and in the tropics.15,16 Long-range atmospheric transport of biomass burning plumes from Africa lead to elevated pollutant levels over the Atlantic,17,18 including PAHs19 and PCDD/Fs.6 Furthermore, secondary emissions of semivolatile organics can be triggered by forest fires20 through heating of soils. Longrange transport of such secondary emissions had been observed Received: Revised: Accepted: Published: 11616

April 6, 2013 September 20, 2013 September 24, 2013 September 24, 2013 dx.doi.org/10.1021/es401499q | Environ. Sci. Technol. 2013, 47, 11616−11624

Environmental Science & Technology

Article

and 0.4 m3 d−1 for HpCDD and OCDD. PAS is diffusion driven, and the sampler geometry is shielding, but efficient sampling volumes are sensitive to wind velocity, temperature,25 and also aerosol properties (size distribution, chemical composition). The effect of temperature on the sampling efficiency is particularly significant for semivolatile substances (including PAHs, not yet verified for PCDDs29,30), because of gas-particle partitioning being temperature dependent, and the sampling efficiency of the passive sampler being much lower for particles than for gaseous molecules.28 No calibration studies to determine the PAS effective equivalent sampling volume covering the high temperatures relevant for this study have yet been performed and no temperature dependencies are available. On the basis of the variabilities found in a 9-year calibration studies in Kosetice, Czech Republic (unpublished data), we estimate the uncertainty coming with these virtual sampling volumes to be ±50% for gaseous substances (PHE) but a factor of 3 for substances being significantly associated with the particulate phase. This high uncertainty estimate, conservatively, includes a safety margin (of 2) which accounts for the nonlinearity of the dependency of the PAH particulate mass fraction on temperature (doubling per −13 K in the temperature range 290−309 K)30,31 and unknown substance and particulate matter composition specificities of the gasparticle partitioning. 2.3. Air-Sheds. The regional provenance of advected air influencing the samples from the 10 sites was determined by three-dimensional 120 h-back-trajectories (every 4 h, 200 m above ground arrival height, HYSPLIT model 32). These airsheds were then checked for possible influences by large fires by comparison with monthly fire activity maps from the Global Fire Assimilation System (GFAS) (see Modeling below).

to be linked to forest fires in northern midlatitudes and in Africa.21,22 Other major POPs sources in Africa are pesticide usage, industries (metallurgical, pulp, chemical), waste incineration, and technical combustion processes in the energy and transport sectors.3,23,24 In this study, we test the hypothesis that biomass burning may contribute significantly to observed elevated atmospheric levels of PCDD and PAH in sub-Saharan Africa, i.e., the Sahel, equatorial, and southern Africa.25 A multicompartment model is applied rather than an atmospheric model only, as the substances studied are semivolatile and persistent in ground compartments, hence, may revolatilize from soil following deposition. In the tropics, this can even be significant for PAHs with low vapor pressure.26 Modelpredicted fire-related and observed atmospheric concentrations for the year 2008 are presented and discussed with the aim to draw conclusions with regard to the sources of the levels observed and the state of knowledge on the POP source biomass burning.

2. METHODS 2.1. Choice of Substances. A limited set of tracers were selected according to their abundance in the pilot study 25 and in field studies on biomass burning in the tropics in general, coverage of a wide range of physicochemical properties, and availability of such data. These consist of four PAHs, phenanthrene (PHE), fluoranthene (FLT), pyrene (PYR), and benzo(a)pyrene (BAP), and three PCDDs, 2,3,7,8tetrachlorodibenzo-p-dioxin (TCDD), 1,2,3,4,6,7,8-heptachlorodibenzo-p-dioxin (HpCDD) and octachlorodibenzo-p-dioxin (OCDD). For computational cost reasons, PCDF could not be included. The physicochemical properties and degradation rates are listed in the Supporting Information (SI), Table S1. Most of the studied substances, PHE, FLT, PYR, TCDD, and HpCDD, are semivolatile (vapor pressures at 298 K in the range p = 10−2−10−6 Pa), while BAP and OCDD with vapor pressures close to or below 10−6 Pa are nonvolatile. The selected PCDDs differ in their dominant removal mechanism: While the main fate of TCDD is atmospheric degradation together with uptake by the ocean, the main fate of HpCDD and OCDD is considered to be wet deposition.27 2.2. Measurements. During the period from January to July 2008, the atmospheric levels of the 16 2−6-ring PAHs, the US EPA priority listed substances, were determined at 4 continental background sites (i.e., Tombouctou (Mali), Mt. Kenya (Kenya), Molopo (Rep. of South Africa) and Barberspan (Rep. of South Africa)), 6 rural or suburban background sites (i.e., Niono (Mali), Bamako Airport (Mali), Koumakonda (Togo), Kwabenya (Ghana), Sheda (Nigeria), and Lusaka Airport (Zambia)) (monthly sampling), and the atmospheric levels of 17 PCDD/Fs at 7 of these sites (not at Niono, Bamako Airport, and Barberspan, 3-month samples 25). In addition, measurements at two urban or industrial sites, i.e., Brazzaville (Congo) and Dakar (Senegal), are included in the analysis for reference purposes (site information is listed in the SI, Table S2). The sampling method was passive air sampling (PAS 28). The results, provided in ng per sample, were converted into atmospheric concentrations using conversion factors (effective passive sampling volumes) derived from active and passive sideby-side sampling (September−December 2011 in Brno, Czech Republic), namely 4.1 m3 d−1 for PHE, 3.4 m3 d−1 for FLT, 2.8 m3 d−1 for PYR, 0.14 m3 d−1 for BAP, 1.9 m3 d−1 for TCDD,

3. MODELING 3.1. Emissions. Global emission of PAHs and PCDDs into air were based on emission factors applied to daily spatially resolved dry vegetation combustion rates. Since PAHs and PCDDs so far have not been included into the comprehensive reviews of emission factors from biomass burning to be used in modeling,33,34 a detailed literature study has been performed here. Emission factors for PAHs are based on geometric means of individual measurement results (details in SI S1.3.1). For PCDDs recommended emission factors for the sum of PCDD/Fs are used,35,36 recently updated37,38 and individual species’ emission factors scaled according to PCDD/ F patterns determined in open vegetation fires (details in SI S1.3.2). Emissions into the soil (e.g., in the ash) were neglected as these are far lower than the release into air.38 Daily real-time fire data (g fuel burnt day−1) are based on satellite-observed fire radiative power (GFAS FRP) measured by MODIS 39 and the 24 h means are uniformly distributed (details in SI, S1.4). MODIS, with an infrared channel resolution of 1 km at nadir omits small fires below around 10 MW FRP corresponding to approximately 0.01−0.1 ha effective fire area, dependent on the combustion stage. Therefore, open but small (e.g., agricultural fires such as e.g., stubble burning, or burning of piled agricultural wastes), smoldering (e.g., waste dumps), and indoor fires (e.g., biomass fuels in households) are not included. The satellite data do cover the entire simulated period. 3.2. Model. Tracer distributions in air upon emission were predicted using the multicompartment chemistry-transport model MPI-MCTM.40 It was developed from the general 11617

dx.doi.org/10.1021/es401499q | Environ. Sci. Technol. 2013, 47, 11616−11624


Article

Figure 1. Model-predicted Jan-Jun 2008 mean total atmospheric concentrations (sum gas and particulate fractions, pg m−3) of open fire related PYR, BAP, and HpCDD (from left to right) ∼1500 m height (upper row) and ground-level (lower row).

Figure 2. Model-predicted continental monthly mean total atmospheric concentrations (sum gas and particulate fractions, ng m−3, from left to right January through June) of (a) FLT and (b) HpCDD from vegetation fires at ∼1500 m height (upper row) and ground-level (lower row). The time series of the vertical distribution over selected stations show that the maximum is sometimes higher than 4 km and sometimes near the ground (SI, Figure S1).

circulation model ECHAM5 with aerosol submodel HAM 41 by including 2-dimensional surface compartments (topsoil, vegetation surfaces, ocean surface mixed layer,40 and land and sea ice,42 and the respective surface exchange processes). The model and its substance specific parametrizations had been applied and validated for studies on the global cycling of organochlorine pesticides,40 semivolatile and nonvolatile PAHs,26,43 and aerosol components in general.41,44

Atmospheric chemistry of PAHs and PCDDs is considered by reaction with the hydroxyl radical and ozone in the gasphase (kinetic data in SI, Table S1), while the reactivity of the particulate phase mass fraction of the substances is accounted for in the context of a sensitivity study (SI, S1.5.1). For PAHs, such a simplification appears justified as reactivity in the particulate phase must be significantly less than in the gaseous phase in order to explain levels at remote sites.26 In fact, 11618



Article

Figure 3. PAHs (ng m−3, left) and PCDDs (pg m−3, right) predicted open fire related concentrations at ground level and observations various types of site. Values below limit of quantification (LOQ) were set equal to LOQ, i.e., 0.003 ng m−3 and 0.001 pg m−3 for PAHs and PCDDs, respectively. Three-monthly mean basis.

Temperature dependent substance degradation in soil, on vegetation surfaces, and in seawater is assumed to follow firstorder kinetics40 (kinetic data in SI, Table S1). The tracers are released in gaseous form into the lowest atmospheric level and distribute in the first time step between the phases of the aerosol. Gas-particle partitioning was parametrized considering absorption into OM (using the octanol-air partitioning coefficient, Koa model50) for PCDDs and both absorption into OM and adsorption to BC 51 for PAHs, i.e., the gas-particle partitioning models which explain observations with regard to the respective substance class best.27,43 3.3. Model Setup and Simulation. The model meteorology and chemistry was spin-up over 4 years, while the simulation is of the fifth year, 2008. The model large-scale

reactivity in the particulate phase is far from being understood: Recent laboratory studies45−47 suggest that the reactions of semivolatile and nonvolatile PAH molecules sorbed to particles with ozone, the OH radical and NO2 may significantly limit the atmospheric lifetime of semivolatile and nonvolatile species in plumes. However, particle chemical composition and morphology matter and may also prevent reactivity through inaccessibility of the PAH molecule within the bulk particle for oxidant attack.48 The related uncertainty is covered in a sensitivity study (SI, S1.5.1, S2.5), while in the case of the PCDDs we refrain from considering this chemistry for lack of better knowledge. The hydroxyl radical concentrations are taken from a comprehensive chemistry model (EMAC, 69 species, 178 reactions, standard distributions49) run. 11619



Article

At continental background stations, predicted fire related PAH levels exceed the observations by up to a factor of ∼30, most pronounced for FLT and BAP (often 25% and >125%, respectively) and the station Bamako International Airport (>80%), where only upper limits of PCDDs could be determined (see also SI, Figure S4). Interestingly, PCDDs were also 95% of TCDD remains unexplained by vegetation fires, as the highest share predicted attributable to fires whenever quantified by the measurements (at 6 sites) was 5% and 0.2% at the equatorial sites Brazzaville (urban) and Mt. Kenya (continental background), respectively.

meteorology (temperature surface pressure, divergence, and vorticity) was constrained to the years 2003−08 by nudging ECHAM5 to the ECMWF ERA-Interim reanalysis data.52 Radical concentrations for the year 2004 (monthly data, daylight-dependent diurnal cycles) are adopted for all simulated years. The model resolution was T63, i.e., 96 × 192 grid cells (∼1.9° × 1.9°) in the horizontal, and 31 vertical levels (∼1000−10 hPa). The time step was 12 min.

4. RESULTS 4.1. Co-incidences of Sample Air-Sheds and Open Vegetation Fires. The comparison of monthly fire count maps and monthly air-sheds indicated coincidences, i.e., likely influences of large fires in Ghana in January, in Sierra Leone and Guinea in April−-May and in the Congo in June on samples collected in Mali, Ghana, and Republic of Congo, respectively (SI, Figure S1). These coincidences are not reflected in the time series of observed concentrations, which were mostly elevated in January−March as compared to April−June (SI, Figure S5). The variation of PYR concentrations (not shown) was similar to the one of FLT. 4.2. Multicompartmental Modeling. 4.2.1. Predicted Atmospheric POP Levels Related to Vegetation Fires. Predicted open fire-related distributions of the studied PAHs and PCDDs are shown in Figures 1, 2, S3, and S5 (SI). Half-year (Jan−June 2008) mean near-ground atmospheric concentrations for the continent (defined as 18°W−55°E, 39°N−36°S) are 0.0076, 0.51, and 3.3 fg m−3 of TCDD, HpCDD, and OCDD, respectively, and 0.80, 0.48, 0.24, and 0.022 ng m−3 of PHE, FLT, PYR and BAP, respectively (distributions in Figure 1 for PYR, BAP, and HpCDD; the other tracer distributions are similar). The distributions originating from biomass burning plumes are dislocated according to prevailing winds, stretching far into the Gulf of Guinea and the equatorial Atlantic Ocean (Figures 1 and 2). The temporal variation is characterized by a minimum in April (Figure 2): HpCDD concentrations decline from 0.79 fg m−3 (January mean) steadily to 0.31 fg m−3 (April mean) and then increases to 0.58 fg m−3 (June mean), and FLT concentrations decline from 0.81 ng m−3 (January mean) to 0.24 ng m−3 (April mean) and then raised to 0.50 ng m−3 (June mean). Much is stored in higher altitudes of the planetary boundary layer. The regional half-year mean concentrations in ∼1500 m height correspond to 27, 52, and 44% of TCDD, HpCDD, and OCDD, respectively, and of 12, 23, 20, and 0.2% of PHE, FLT, PYR, and BAP, respectively, of those at the ground level (Figures 1 and 2) and the 50th percentile of the vertical mass distribution (vertical center of gravity) are located at approximately 3400, 3400, and 3150 m for TCDD, HpCDD, and OCDD, respectively, and at approximately 2200, 2600, 2550, and 900 m for PHE, FLT, PYR, and BAP, respectively. 4.2.2. Comparison of Predictions with Observations. The analysis is limited to background sites in sub-Saharan Africa. For reference purposes, two urban or industrial sites in the region are included, which were identified as being located in the regions possibly influenced by large vegetation fires in 2008 (“Co-incidences of sample air-sheds and open vegetation fires”, above), i.e., Dakar (Senegal) and Brazzaville (Republic of the Congo). The January−June means of predicted open fire related concentrations at ground level and observed levels are compared in Figure 3 and the related time series in the SI (Figure S3).

5. DISCUSSION Data uncertainties of this study contribute to both the uncertainties of the predicted concentration (from open fires) and the observed levels. 5.1. Observational Data. Measured data were based on passive air sampling whose efficiency is dependent on wind velocity and through its sensitivity to the phase equilibrium also on temperature and aerosol properties. Sampling of the total (gaseous and particulate) substance concentrations was probably underestimated, as effective passive sampling volumes had been based on measurements performed at lower temperatures, which means that the gas-particle partitioning was biased in the calibration data set toward high particulate mass fractions. In conclusion, the observed levels of all semivolatile substances (i.e., substances having p = 10−6−10−2 Pa at 298 K, hence, all substances addressed except PHE, see SI, Table S1) may be overestimated. The possible overestimation is limited for those substances which are mostly gaseous in temperate climates, i.e., PYR and FLT (particulate mass fraction θ = 0.05−0.3; range spanned by results of numerous field studies in the literature) and TCDD (θ = 0.13 (53)) and as expected most pronounced for BAP, HpCDD, and OCDD. However, overestimated effective sampling volumes cannot explain the systematic overpredictions: Among these substances BAP and OCDD are clearly overpredicted (January−Jun mean of cpred/cobs is 5.7 and 2.4, respectively), while HpCDD is not (cpred/cobs = 0.3). As the more volatile PAHs are similarly overpredicted as BAP and OCDD (for PHE, PYR, and FLT cpred/cobs is 2.8, 5.2, and 3.5, respectively). 11620



Article

5.2. Fire Related POP Emissions. The global sums of vegetation fire related PAH emissions in 2008 are 18.7 Gg PHE, 6.81 Gg FLT, 6.38 Gg PYR, and 1.38 Gg BAP (SI, Table S5). For PHE and FLT these values compare very well with global emission estimates for PAHs from open biomass burning (sum for the year 199626), i.e., 16.8 Gg PHE and 6.85 Gg FLT. However, with 2.36 Gg BAP was estimated significantly higher. Emissions in Africa account for 41−52% of the global emissions (SI, Table S5). They peak in January and September, corresponding to the burning seasons in the Sahel and Southern Africa, respectively, being 1 order of magnitude stronger than in months of minimum emission, i.e. April and October (SI, Figure S2). The emissions of PHE and FLT from all sources in continental Africa in 1996 were estimated 6436 and 5030 t yr−1, respectively.26 This means that the open vegetation fire related fluxes (SI, Table S5) would account for ∼82 and ∼55%, respectively, of the total emissions. The global sums of vegetation fire related PCDD emissions in 2008 are 0.18 kg TCDD, 6.83 kg HpCDD and 41.2 kg OCDD. These numbers correspond to ∼0.004, ∼0.3, and ∼0.5% of all global emissions of TCDD, HpCDD, and OCDD, respectively (based on estimates for the 1990s27). Emissions in Africa account for 41−65% of the global vegetation fire emissions (SI, Table S5). The emission flux of total PCDD/Fs can be roughly estimated by multiplying the emission fluxes of TCDD, HpCDD, and OCDD by 890, 26, and 5.0, respectively. These factors are based on PCDD/F patterns determined in open vegetation fires53−55 and a typical mix of savannah and forest fuels of 87 ± 3 in Africa (above). This leads to 180 ± 25 kg yr−1 of PCDD/F from open vegetation fires in Africa. In the 1990s, a PCDD/F production of 1.3 t yr−1 had been estimated for domestic waste combustion in sub-Saharan Africa 4 and ∼0.3 t yr−1 from all sources in west Africa6 (using a factor of 60 to convert from toxicity equivalents (TEQ) to total PCDD/F 5). 5.3. Model Predictions. Predicted PAH atmospheric concentrations from vegetation fires alone exceed the observed levels (quantified above). Levels in air also result from other sources, such as industrial and household combustion processes and waste burning, which are not included into the model predictions. Exceedances could be explained by too high PAH emission factors or underestimation of atmospheric sinks, i.e. photochemical degradation in the gaseous or particulate phase or wet deposition. 5.4. Deposition. Dry deposition of gaseous and particulate phase trace substances is described by validated state-of-the-art parametrizations,40,41,56 and hence, is not expected to be a significantly underestimated sink process. Wet deposition of the tracers studied is significantly more effective in the particulate phase, hence, dependent on gas-particle partitioning. There is a tendency for exceedances being more frequent for the least volatile substances, BAP and OCDD. Modeled monthly precipitation was compared with precipitation observations from the Tropical Rainfall Measuring Mission (TRMM) satellite. The TRMM satellite has several rainfall instruments onboard, measuring precipitation with radar, microwaves and in visible/infrared channels. The 3B42 V6 product used here is a combined sensor product, which is described in more detail in Huffman et al.57 The product has been validated for Africa with rain gauge data and shows generally a good performance.58 The agreement was generally good with some overestimates south and underestimates north

of the equator. Discrepancies found did hardly overlap with the airsheds of the samples, with two exceptions, which suggest underestimates of wet deposition only in the airshed of samples collected in Bamako and Niono in April−May and overestimates in the airshed of samples collected in Brazzaville in February−March (not shown). Exceedances are simulated for Bamako and Niono for BAP and OCDD (SI, Figures S4, S5). The hypothesis that the overpredictions of all PAHs studied and OCDD at several sites can be attributed to wet deposition can be rejected. 5.5. Photochemistry. Neglect of degradation in the particulate phase would not affect the PHE predictions, which were similarly off at most of the stations, at least not better than those of PYR (SI, Figure S4). In a sensitivity study, we account for this uncertainty and include the oxidation with ozone on aerosol surfaces (SI S1.5.1). The results suggest that this uncertainty could explain most, but not all of the overestimates. A significant underestimate of photochemical degradation of the gaseous compounds is unlikely, as the kinetic data are well established. Also concentration ratios of pairs of tracers are tested: Ratios may be source-specific (and are then called ‘diagnostic’ 59), if conservative during transport due to similar degradability. It is found during periods of possible influence of emissions of fires on the sample compositions (sites Bamako International Airport, Kwabenya and Brazzaville) that HCDD/ (HpCDD + OCDD) was lower, whereas FLT/(FLT + PYR) and FLT/(FLT + 1000 × OCDD) were elevated. Because of different photochemical degradability of these pairs of substances, however, these ratios are not conservative. For example, the rate coefficient for the reaction of OCDD with the hydroxyl radical, kOH, is more than 2 orders of magnitude lower than kOH of FLT and PYR (SI, Table S1) such that FLT/(FLT + 1000 × OCDD) provides actually an indication for age of air. Predicted values of FLT/(FLT + 1000 × OCDD) (>0.97) mostly exceed observed values (0.36−0.99). This is not the case assuming high photochemical degradation rates (0.14−0.91; sensitivity analysis, SI, S1.5.1). This indicates that the degradation rate of FLT was underestimated by neglecting heterogeneous reactivity. Similarly, BAP/(BAP + 1000 OCDD) values in the range 0.00−0.95 are predicted, but no values exceeding 0.70 are observed. These findings can be explained by the neglect of the ozone reactions of FLT (kO3 = 1.5 × 10−17 cm3 molec−1 s−1 on silica particles 45) and BAP (kO3 = 1.4 × 10−16 cm3 molec−1 s−1 on solid organic and salt aerosols 60) and the reaction of FLT with the OH radical (kOH = 3.2 × 10−13 cm3 molec−1 s−1 on silica particles46). The difference between kO3 of FLT and BAP suggests that PAH photochemical sinks are more underestimated than FLT’s, which is supported by the comparison of predicted and observed levels (Figures 3, S3, and S4 of the SI). It had been pointed out that the knowledge of chemical kinetics of particle-associated and also of gaseous PAHs is limiting source attribution and understanding of transport and fate on large spatial scales.26,59 For this study, substances were selected with at least consolidated reaction rate coefficients with the hydroxyl radical and ozone in the gas-phase. The results of the sensitivity analysis suggest that the heterogeneous reactions of PHE, FLT, PYR, and BAP with ozone are effectively limiting atmospheric lifetime and long-range transport potential. 5.6. Biomass Burning Emissions. Substantial progress in the characterization of emission factors from biomass burning has been achieved for inorganic and volatile organic species and 11621



Article

fuel types.33,34 Yet, emission factors for most species remain uncertain due to the small number of representative field measurements and the intrinsic strong heterogeneity in the combustion conditions and fuel properties. This is particularly true for PAHs: Field measurements of PAH emission factors are generally missing and available (simulated) open burning measurements of emission factors do not comprise tropical forest, savannah, or peat fuels. In the absence of other information, the emission factors used here had to be based on studies for different fuel types instead, which varied largely (typically 2 orders of magnitude; SI, S1.3.1). It is particularly unclear how representative measured emission factors for crop residue burning are for savannah/grassland fires of Africa. The geometric means derived could only be based on a few studies, which had characterized combustion conditions realistic for open fires. In conclusion, the PAH emission factors are uncertain by 1 order of magnitude. Hereby, the geometric means derived appear to represent lower intermediate estimates. Overestimated emission factors by 1 order of magnitude could explain the exceedances at all sites, except Mt. Kenya (exceedances in the range cpred/cobs = 30−60, 3monthly mean basis, see also SI, Figure S5). This data analysis also implies that PAHs in the background would be dominated by biomass burning. In a sensitivity study, we account for this uncertainty and include a lower estimate of PAH emission factors based on a (SI S1.5.2). The results suggest that this uncertainty could explain most, but not all of the overestimates. An existing global PAH emission inventory 61 bases forest and savannah fire emission factors on values from one study only,62 i.e., fuels firewood and rice straw, respectively. These emission factors are lower, too, typically by a factor of 2−3. Field measurements of PAH emission factors in biomass burning plumes are needed to establish a representative estimate of PAH emissions from vegetation fires in various ecosystems. This would allow a quantitative assessment of the contribution of biomass burning to the atmospheric cycling of PAHs and the long-range transport potential of PAHs. Effective passive sampling volumes need to be determined under conditions representative for subtropical and tropical climates. PCDD emission factors from open biomass burning had recently been downward corrected significantly (i.e., from 5 to 1 pgTEQ (g fuel burnt−1, using the 2005 WHO-TEF scheme) for forest fires, from 300 to 40 pgTEQ (g fuel burnt−1) for waste burning), considered in this study.37,38 With uncorrected emission factors, the predicted levels would have been much higher, by a factor of 5−10.37 The comparison with observed levels, which indicates that open fires explain 1−10% of PCDD levels in the continental and rural background and the results of the sensitivity study (higher factors by a factor of 10; SI S1.5.2, S2.5), are supporting this downward correction. Note that even with the uncorrected factors, PCDD levels in the background would not be dominated by biomass burning. However, emissions used in this study did not include small fires (≲0.01 ha) including open domestic waste burning, which has high PCDD emission factors (40 pgTEQ (g fuel burnt−1).38 Therefore, levels downwind of urban areas are eventually underestimated here. No such indication was found at the urban sites included in this study, i.e., Dakar (PCDDs < LOQ) or Brazzaville (open fire emissions also explained 0.1−1%, similar to the background). 5.7. Dynamic Processes. The dynamic processes covered in the atmosphere general circulation model (included in our MCTM) are necessarily incomplete. Not all relevant

contributions to convection are considered: In this study, biomass burning emissions are injected into the lowest model layer. Due to the fire-induced plume-rise, biomass burning plumes typically rise higher into the atmosphere.63 It had been shown by Guan et al. 64 that the thermal energy of large fires in Southern Africa, when accounted for by modeling, would explain a higher vertical transport of biomass burning tracers into the free troposphere and a subsequent enhanced horizontal transport. The concentrations in the free troposphere and the long-range transport potential of PAHs and PCDDs emitted from biomass burning may therefore be underestimated. In conclusion, this first modeling study so far on the contribution of emissions from vegetation fires on ambient POP levels suggests that this source type contributes significantly to the exposure of the African rural and continental environment to PAH and PCDDs: Outside urban areas the contribution of vegetation fires to PCDDs is in the range 1− 10%. Obviously, PCDD levels in air in Africa are dominated by other sources (combustion, pesticide application). The results support a recent revision of the recommended emission factors, as the previous ones would suggest obvious overestimates, at least for HpCDD and OCDD. The contribution of biomass burning to PAH levels in air is expected to be higher than 10% at background sites, but cannot be quantified due to uncertainties of input data. A sensitivity analysis suggests that such a contribution would be explained by a combination of higher estimates of photochemical degradation (namely taking into account the heterogeneous PAH loss by reaction with ozone) and lower open fire emission factors, both within the uncertainty ranges of these input data. PAH and PCDD heterogeneous chemical sinks are the major uncertainties. Studies are needed which target other sources of PAHs and PCDDs in Africa, such as waste incineration and industrial and household combustion.

■

ASSOCIATED CONTENT

S Supporting Information *

Substance physicochemical properties and degradation rate coefficients, selection of emission factors, sensitivity analysis, results of the sample airsheds, emission fluxes, model predicted levels, and comparison with observed levels. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Phone +49-6131-305-4055; fax +49-6131-305-4009; e-mail g. [email protected]. Present Addresses ∥

University of Hamburg, CEN, Institute for Hydrobiology and Fisheries Science, Grosse Elbstr. 133, 22767 Hamburg, Germany. ⊥ Academy of Sciences of the Czech Republic, Global Change Research Centre, Bělidla 4a, 60300 Brno, Czech Republic. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors are grateful to M.G. Schultz for valuable discussions. The model runs were performed on the IBM Power6 computer of the German Climate Computing Centre (DKRZ). This research was supported by the Granting Agency 11622



Article

G. W.; Singh, H. B.; Talbot, R. W. Ozone and aerosol distributions and air mass characteristics over the South Atlantic basin during the burning season. J. Geophys. Res. 1996, 101, 24043−24068. (19) Nizzetto, L.; Lohmann, R.; Gioia, R.; Jahnke, A.; Temme, C.; Dachs, J.; Herckes, P.; di Guardo, A.; Jones, K. C. PAHs in air and seawater along a north-south Atlantic transect: Trends, processes and possible sources. Environ. Sci. Technol. 2008, 42, 1580−1585. (20) Black, R. R.; Meyer, C. R.; Yates, A.; van Zwieten, L.; Chittim, B. G.; Mueller, J. F. Release of PCDD/PCDF to air and land during open burning of sugarcane and forest litter over soil fortified with mass labelled PCDD/PCDF. Atmos. Environ. 2012, 59, 125−130. (21) Eckhardt, S.; Breivik, K.; Manö, S.; Stohl, A. Record high peaks in PCB concentrations in the Arctic atmosphere due to long-range transport and biomass burning emissions. Atmos. Chem. Phys. 2007, 7, 4537−4536. (22) Gioia, R.; Eckhardt, S.; Breivik, K.; Jaward, F. M.; Prieto, A.; Nizzetto, L.; Jones, K. C. Evidence for major emissions of polychlorinated biphenyls in the West African region. Environ. Sci. Technol. 2011, 45, 1349−1355. (23) Mansour, S. A. Persistent organic pollutants (POPs) in Africa: Egyptian scenario. Hum. Exp. Toxicol. 2009, 28, 531−66. (24) Linderholm, L.; Biague, A.; Månsson, F.; Norrgren, H.; Bergman, A.; Jakobsson, K. Human exposure to persistent organic pollutants in West AfricaA temporal trend study from GuineaBissau. Environ. Int. 2010, 36, 75−82. (25) Klánová, J.; Č upr, P.; Holoubek, I.; Borůvková, J.; Přibylová, P.; Kareš, R.; Tomšej, T.; Ocelka, T. Persistent organic pollutants in Africa. Part 1: Passive air sampling across a continent-wide network 2008. J. Environ. Mon. 2009, 11, 1952−1963. (26) Lammel, G.; Sehili, A. M.; Bond, T. C.; Feichter, J.; Grassl, H. Gas/particle partitioning and global distribution of polycyclic aromatic hydrocarbonsA modelling approach. Chemosphere 2009, 76, 98− 106. (27) Lohmann, R.; Jurado, E.; Dachs, J.; Lohmann, U.; Jones, K. C. Quantifying the importance of the atmospheric sink for polychlorinated dioxins and furans relative to other global loss processes. J. Geophys. Res. 2006, 111, D21303 DOI: 10.1029/2005JD006923. (28) Klánová, J.; Č upr, P.; Kohoutek, J.; Harner, T. Assessing the influence of meteorological parameters on the performance of polyurethane foam-based passive air samplers. Environ. Sci. Technol. 2008, 42, 550−555. (29) Bidleman, T. F. Atmospheric processes: wet and dry deposition of organic compounds are controlled by their vapor-particle partitioning. Environ. Sci. Technol. 1988, 22, 361−367. (30) Lammel, G.; Klánová, J.; Ilić, P.; Kohoutek, J.; Gasić, B.; Kovacić, I.; Škrdlíková, L. Polycyclic aromatic hydrocarbons in air on small spatial and temporal scalesII. Mass size distributions and gasparticle partitioning. Atmos. Environ. 2010, 44, 5022−5027. (31) Pankow, J. F.; Bidleman, T. F. Interdependence of the slopes and intercepts from log-log correlations of measured gas-particle partitioning and vapour pressureI. Theory and analysis of available data. Atmos. Environ. 1992, 26A, 1071−1080. (32) Draxler, R. R.; Rolph, G. D. HYSPLIT (HYbrid Single-Particle Lagrangian Integrated Trajectory) Model access via NOAA ARL READY Website, NOAA Air Resources Laboratory; NOAA: Silver Springs, MD, USA, 2003, http://www.arl.noaa.gov/ready/hysplit4. html. (33) Andreae, M. O.; Merlet, P. Emission of trace gases and aerosols from biomass burning. Global Biogeochem. Cyc. 2001, 15, 955−966. (34) Akagi, S. K.; Yokelson, R. J.; Wiedinmyer, C.; Alvarado, M. J.; Reid, J. S.; Karl, T.; Crounse, J. D.; Wennberg, P. O. Emission factors for open and domestic biomass burning for use in atmospheric models. Atmos. Chem. Phys. 2011, 11, 4039−4072. (35) UNEP. Standardized toolkit for identification and quantification of dioxin and furan releases, UNEP Chemicals, ed. 2.1, 253 pp., Geneva, 2005, http://www.chem.unep.ch/POPs/pcdd_activities/ toolkit/default.htm. (36) Kakareka, S. V. Sources of persistent organic pollutants emission on the territory of Belarus. Atmos. Environ. 2002, 36, 1407−1419.

of the Czech Republic (Project No. P503/11/1230) and the European Commission, FP7 (Project No. 226534, ArcRisk, and No. 218793, MACC).

■

REFERENCES

(1) van der Werf, G. R.; Randerson, J. T.; Giglio, L.; Collatz, G. J.; Mu, M.; Kasibhatla, P. S.; Morton, D. C.; de Fries, R. S.; Jin, Y.; van Leeuwen, T. T. Global fire emissions and the contribution of deforestation, savanna, forest, agricultural, and peat fires (1997−2009). Atmos. Chem. Phys. 2010, 10, 11707−11735. (2) Gullett, B.; Touati, A.; Oudejans, L. PCDD/F and aromatic emissions from simulated forest and grassland fires. Atmos. Environ. 2008, 42, 7997−8006. (3) Fiedler, H. Dioxins and Furans (PCDD/PCDF). In The Handbook of Environmental Chemistry; Hutzinger, O., Fiedler, H., Eds.; Springer: Berlin, 2003; Vol. 30, pp 123−201. (4) UNEP. Regionally Based Assessment of Persistent Toxic Substances (RBA PTS) Project for Region V, Sub-Sahara Africa, 2002, 132 pp. (5) Brzuzy, L. P.; Hites, R. A. Global mass balance for polychlorinated dibenzo-p-dioxins and dibenzofurans. Environ. Sci. Technol. 1996, 30, 1797−1804. (6) Lohmann, R.; Ockenden, W. A.; Shears, J.; Jones, K. C. Atmospheric distribution of polychlorinated dibenzo-p-dioxins, dibenzofurans (PCDD/Fs), and non-ortho biphenyls (PCBs) along a North-South Atlantic transect. Environ. Sci. Technol. 2001, 35, 4046− 4053. (7) Holt, E.; Weber, R.; Stevenson, G.; Gaus, C. Polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) impurities in pesticides: A neglected source of contemporary relevance. Environ. Sci. Technol. 2010, 44, 5409−5415. (8) Holt, E.; Weber, R.; Stevenson, G.; Gaus, C. Formation of dioxins during exposure of pesticide formulations to sunlight. Chemosphere 2012, 88, 364−370. (9) Trollope, W. S. W. Personal perspectives on commercial versus communal African fire paradigms when using fire to manage rangelands for domestic livestock and wildlife in southern and east African ecosystems. Fire Ecology 2011, 7, 57−73. (10) Lacaux, J. B.; Delmas, R.; Cros, B.; Lefeivre, B.; Andreae, M. O. Influence of biomass burning on precipitation chemistry in the equatorial forests of Africa. In Global Biomass Burning; Levine, J., Ed.; MIT Press: Cambridge, MA, 1991, pp 167−173. (11) Lacaux, J. B.; Delmas, R.; Jambert, C.; Kuhlbusch, T. A. J. NOx emissions from African savannah fires. J. Geophys. Res. 1996, 101, 23585−23595. (12) Kaiser, J. W.; Heil, A.; Andreae, M. O.; Benedetti, A.; Chuborova, N.; Jones, L.; Morcrette, J. J.; Razinger, M.; Schultz, M. G.; Suttie, M.; van der Werf, G. R. Biomass burning emissions estimated with a global fire assimilation system based on observed fire radiative power. Biogeosci. 2012, 9, 527−554. (13) Gullett, B.; Touati, A. PCDD/F emissions from burning wheat and rice field residue. Atmos. Environ. 2003, 35, 4893−4899. (14) Ward, T. J.; Lincoln, E. Concentrations of PM2.5-associated OC, EC, and PCDD/Fs measured during the 2003 wildfire season in Missoula, Montana. Environ. Monit. Assess. 2006, 115, 39−50. (15) Shih, S. I.; Lee, W. J.; Lin, L. F.; Huang, J. Y.; Su, J. W.; ChangChien, G. P. Significance of biomass open burning on the levels of polychlorinated dibenzo-p-dioxins and dibenzofurans in the ambient air. J. Hazardous Mater. 2008, 153, 276−284. (16) Prange, J. A.; Gaus, C.; Weber, R.; Rapke, O.; Müller, J. F. Assessing forest fire as a potential PCDD/F source in Queensland, Australia. Environ. Sci. Technol. 2003, 37, 4325−4329. (17) Fishman, J.; Fakhruzzaman, K.; Cros, B.; Nganga, D. Identification of widespread pollution in the southern hemisphere deduced from satellite analyses. Science 1991, 252, 1693−1696. (18) Browell, E. V.; Fenn, M. A.; Butler, C. F.; Grant, W. B.; Clayton, M. B.; Fishman, J.; Bachmeier, A. S.; Gregory, G. L.; Fuelberg, H. E.; Bradshaw, J. T.; Sandholm, S. T.; Blake, D. R.; Heikes, B. G.; Sachse, 11623



Article

R. K. An emission budget for dioxins from crop and bush fires in Australia, Technical Report #1. Australian Government Dept. of Environment and Heritage: Canberra, 193 pp., 2004. (55) Meyer, C. P.; Müller, J. F.; Symons, R. K. Determination of the source and level of dioxin emissions from bushfires in Australia, CSIRO Rep., 81 pp., 2009. (56) Stier, P.; Feichter, J.; Kloster, S.; Kloster, S.; Vignati, E.; Wilson, J. Emission-induced non-linearities in the global aerosol system: Results from the ECHAM5-HAM aerosol-climate model. J. Clim. 2006, 19, 3845−3862. (57) Huffman, G. J.; Adler, R. F.; Bolvin, D. T.; Gu, G.; Nelkin, E. J.; Bowman, K. P.; Hong, Y.; Stocker, E. F.; Wolff, D. B. The TRMM Multi-satellite Precipitation Analysis: Quasi-global, multi-year, combined-sensor precipitation estimates at fine scale. J. Hydrometeor. 2007, 8, 38−55. (58) Nicholson, S. E.; Some, B.; McCollum, J.; Nelkin, E.; Klotter, D.; Berte, Y.; Diallo, B. M.; Gaye, I.; Kpabeba, G.; Ndiaye, O.; Noukpozounkou, J. N.; Tanu, M. M.; Thiam, A.; Toure, A. A.; Traore, A. K. Validation of TRMM and other rainfall estimates with a highdensity gauge dataset for West Africa. Part II: Validation of TRMM rainfall products. J. Appl. Meteorol. 2003, 42, 1355−1368. (59) Dvorská, A.; Lammel, G.; Klánová, J. Use of diagnostic ratios for studying source apportionment and reactivity of ambient polycyclic aromatic hydrocarbons over Central Europe. Atmos. Environ. 2011, 45, 420−427. (60) Kwamena, N. O. A.; Thornton, J. A.; Abbatt, J. P. D. Kinetics of surface-bound benzo[a]pyrene and ozone on solid organic and salt aerosols. J. Phys. Chem. A 2004, 108, 11626−11634. (61) Zhang, Y. X.; Tao, S. Global atmospheric emission inventory for polycyclic aromatic hydrocarbons (PAHs) for 2004. Atmos. Environ. 2009, 43, 812−819. (62) Jenkins, B. M.; Jones, A. D.; Turn, S. Q.; Williams, R. B. Emission factors for polycyclic aromatic hydrocarbons from biomass burning. Environ. Sci. Technol. 1996, 30, 2462−2469. (63) Val Martin, M.; Kahn, R. A.; Logan, J. A.; Paugam, R.; Wooster, M.; Ichoku, C. Space-based observational constraints for 1-D fire smoke plume-rise models. J. Geophys. Res. 2012, 117, D22204 DOI: 10.1029/2012JD018370. (64) Guan, H.; Chatfield, R. B.; Freitas, S. R.; Bergström, R. W.; Longo, K. M. Modeling the effect of plume-rise on the transport of carbon monoxide over Africa with NCAR CAM. Atmos. Chem. Phys. 2008, 8, 6801−6812.

(37) Black, R. R.; Meyer, C.; Touati, A.; Gullett, B. K.; Fiedler, H.; Mueller, J. F. Emission factors for PCDD/PCDF and dl-PCB from open burning of biomass. Environ. Int. 2012, 38, 62−66. (38) UNEP. Hazardous chemicals from open burning of waste in developing countries, UNEP Chemicals, 153 pp., Geneva, 2010. (39) Heil, A.; Kaiser, J. W.; van der Werf, G. R.; Wooster, M. J.; Schultz, M. G.; Denier van der Gon, H. Assessment of the real-time fire emissions (GFASv0) by MACC, ECMWF Tech. Memo No. 628, ECMWF, Reading, UK, 2010. (40) Semeena, V. S.; Feichter, J.; Lammel, G. Significance of regional climate and substance properties on the fate and atmospheric longrange transport of persistent organic pollutantsexamples of DDT and γ-HCH. Atmos. Chem. Phys. 2006, 6, 1231−1248. (41) Stier, P.; Feichter, J.; Kinne, S.; Kloster, S.; Vignati, E.; Wilson, J.; Ganzeveld, L.; Tegen, I.; Werner, M.; Balkanski, Y.; Schulz, M.; Boucher, O.; Minikin, A.; Petzold, A. The aerosol climate model ECHAM5-HAM. Atmos. Chem. Phys. 2005, 5, 1125−1156. (42) Hofmann, L.; Stemmler, I.; Lammel, G. The impact of organochlorines cycling in the cryosphere on their global distributions and fate2. Land ice and temporary snow cover. Environ. Pollut. 2012, 162, 482−488. (43) Sehili, A. M.; Lammel, G. Global fate and distribution of polycyclic aromatic hydrocarbons emitted from Europe and Russia. Atmos. Environ. 2007, 41, 8301−8315. (44) Roelofs, G. J.; Stier, P.; Feichter, J.; Vignati, E.; Wilson, J. Aerosol activation and cloud processing in the global aerosol-climate model ECHAM5-HAM. Atmos. Chem. Phys. 2006, 6, 2389−2399. (45) Estève, W.; Budzinski, H.; Villenave, E. Relative rate constants for the heterogeneous reactions of OH, NO2 and NO radicals with polycyclic aromatic hydrocarbons adsorbed on carbonaceous particles. Part 1: PAHs adsorbed on 1−2 μm calibrated graphite particles. Atmos. Environ. 2004, 38, 6063−6072. (46) Perraudin, E.; Budzinski, H.; Villenave, E. Identification and quantification of ozonation products of anthracene and phenanthrene adsorbed on silica particles. Atmos. Environ. 2007, 41, 6005−6017. (47) Miet, K.; Le Menach, K.; Flaud, P. M.; Budzinski, H.; Villenave, E. Heterogeneous reactivity of pyrene and 1-nitropyrene with NO2: Kinetics, product yields and mechanism. Atmos. Environ. 2009, 43, 837−843. (48) Zhou, S.; Lee, A. K. Y.; McWhinney, R. D.; Abbatt, J. P. D. Burial effects of organic coatings on the heterogeneous reactivity of particle-borne benzo[a]pyrene (BaP) toward ozone. J. Phys. Chem. A 2012, 116, 7050−7055. (49) Taraborrelli, D.; Lawrence, M. G.; Butler, T. M.; Sander, R.; Lelieveld, J. Mainz Isoprene Mechanism 2 (MIM2): An isoprene oxidation mechanism for regional and global atmospheric modelling. Atmos. Chem. Phys. 2009, 9, 2751−2777. (50) Finizio, A.; Mackay, D.; Bidleman, T.; Harner, T. Octanol-air partitioning coefficient as a predictor of partitioning of semi-volatile organic chemicals to aerosols. Atmos. Environ. 1997, 31, 2289−2296. (51) Dee, D. P.; Uppala, S. M.; Simmons, A. J.; Berrisford, P.; Poli, P.; Kobayashi, S.; Andrae, U.; Balmaseda, M. A.; Balsamo, G.; Bauer, P.; Bechtold, P.; Beljaars, A. C. M.; van de Berg, L.; Bidlot, J.; Bormann, N.; Delsol, C.; Dragani, R.; Fuentes, M.; Geer, A. J.; Haimberger, L.; Healy, S. B.; Hersbach, H.; Holm, E. V.; Isaksen, L.; Kallberg, P.; Kohler, M.; Matricardi, M.; McNally, A. P.; Monge-Sanz, B. M.; Morcrette, J. J.; Park, B. K.; Peubey, C.; de Rosnay, P.; Tavolato, C.; Thepaut, J. N.; Vitart, F. The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Q. J. Roy. Meteor. Soc. 2011, 137, 553−597. (52) Lohmann, R.; Lammel, G. Adsorptive and absorptive contributions to the gas-particle partitioning of polycyclic aromatic hydrocarbons: State of knowledge and recommended parametrization for modeling. Environ. Sci. Technol. 2004, 38, 3793−3803. (53) Eitzer, B. D.; Hites, R. A. Polychlorinated dioxins and furanes in the ambient atmosphere of Bloomington, Indiana. Environ. Sci. Technol. 1989, 23, 1389−1395. (54) Meyer, C. P.; Beer, T.; Müller, J. F.; Gillett, R.; Weeks, I.; Powell, J.; Tolhurst, K. G.; McCaw, L.; Cook, G.; Marney, D., Symons, 11624


On the Contribution of Biomass Burning to POPs (PAHs and PCDDs

Recommend Documents