Model and input uncertainty in multi-media fate modeling: benzo[a]pyrene concentrations in Europe

The multi-media environmental behavior of heavy metals around tailings under the influence of precipitation

Precipitation can lead to significant leaching of heavy metals from abandoned tailings，resulting in a decline in the quality of the surrounding environment. This study aimed to simulate and quantify the migration patterns and fate of heavy metals in tailings caused by precipitation in various environmental media (tailings, air, water, soil, and sediments) using leaching tests, source apportionment, and a fugacity model. Results revealed that the average contents of Cd, Cu, As, Pb, Zn, and Cr in the un-weathered tailings were 3.43, 495.56, 160.70, 138.94, 536.57, and 69.52 mg/kg, respectively. The ecological risk factors in the tailings as well as in sediments and soils, were in the following order: Cd >Cu >As >Pb >Zn >Cr. A fugacity model based on the mass-balance methods was established, achieving a good agreement between simulation and measured values. The total amounts of Cd, Cu, As, Pb, and Zn leached from abandoned tailings over the 30-year evaluation period were estimated to be 1.09, 62.44, 0.16, 0.94, and 102.12 t, respectively. Soil and sediments are important reservoirs for heavy metals. The sum of the As, Cd, Cu, Pb, and Zn storage capacities in the soil and sediment accounted for 77.28%, 75.63%, 73.94%, 69.39%, and 57.80% of the total storage capacity, respectively. This study could provide the means for the establishment of a targeted pollution control plan, a guide for restoration projects, and will aid in controlling pollution risk and improving the surrounding environment.

Analysis of the multi-media environmental behavior of polycyclic aromatic hydrocarbons (PAHs) within Haizhou Bay using a fugacity model

In this study, we aimed to quantify the transport and fate of PAHs in different environmental phases (air, seawater, soil, sediment and fish), verify application of the Level III fugacity model in a bay simulation, and understand the transport and fate of PAHs in the bay environment on a macroscopic scale. The simulated average concentrations of ∑₁₆PAH in the air and soil (23.8 ng/m³ and 1080.91 ng/g, respectively), which is as a background reference data for the Haizhou Bay. In addition, the soil (307 t), fish (29.4 t), and sediment (9.72 t) phases were found to be important reservoirs in the Haizhou Bay. Emissions from road vehicles (658 t) accounted for the largest share of PAH emissions in the area, and atmospheric deposition contributed most to the input of PAHs to the polluted area in the region. Whereas the contribution of river runoff input was small, and degradation loss was the main output pathway.

Spatially resolved environmental fate models: A review

Spatially resolved environmental models are important tools to introduce and highlight the spatial variability of the real world into modeling. Although various spatial models have been developed so far, yet the development and evaluation of these models remain a challenging task due to several difficulties related to model setup, computational cost, and obtaining high-resolution input data (e.g., monitoring and emission data). For example, atmospheric transport models can be used when high resolution predicted concentrations in atmospheric compartments are required, while spatial multimedia fate models may be preferred for regulatory risk assessment, life cycle impact assessment of chemicals, or when the partitioning of chemical substances in a multimedia environment is considered. The goal of this paper is to review and compare different spatially resolved environmental models, according to their spatial, temporal and chemical domains, with a closer insight into spatial multimedia fate models, to achieve a better understanding of their strengths and limitations. This review also points out several requirements for further improvement of existing models as well as for their integration.

Influence of monitoring data selection for optimization of a steady state multimedia model on the magnitude and nature of the model prediction bias

SimpleBox is an important multimedia model used to estimate the predicted environmental concentration for screening-level exposure assessment. The main objectives were (i) to quantitatively assess how the magnitude and nature of prediction bias of SimpleBox vary with the selection of observed concentration data set for optimization and (ii) to present the prediction performance of the optimized SimpleBox. The optimization was conducted using a total of 9604 observed multimedia data for 42 chemicals of four groups (i.e., polychlorinated dibenzo-p-dioxins/furans (PCDDs/Fs), polybrominated diphenyl ethers (PBDEs), phthalates, and polycyclic aromatic hydrocarbons (PAHs)). The model performance was assessed based on the magnitude and skewness of prediction bias. Monitoring data selection in terms of number of data and kind of chemicals plays a significant role in optimization of the model. The coverage of the physicochemical properties was found to be very important to reduce the prediction bias. This suggests that selection of observed data should be made such that the physicochemical property (such as vapor pressure, octanol-water partition coefficient, octanol-air partition coefficient, and Henry's law constant) range of the selected chemical groups be as wide as possible. With optimization, about 55%, 90%, and 98% of the total number of the observed concentration ratios were predicted within factors of three, 10, and 30, respectively, with negligible skewness.

Watershed-scale modeling on the fate and transport of polycyclic aromatic hydrocarbons (PAHs)

PAHs are potentially carcinogenic substances that are persistent in the environment. Increasing concentrations of PAHs were observed due to rapid urbanization, thus; monitoring PAHs concentrations is necessary. However, it is expensive to conduct intensive monitoring activities of a large number of PAHs. This study addressed this issue by developing a multimedia model coupled with a hydrological model (i.e., Soil and Water Assessment Tool (SWAT)) for Taehwa River (TR) watershed in Ulsan, the industrial capital of South Korea. The hydrologic module of the SWAT was calibrated, and further used to simulate the fate and transport of PAHs in soil and waterbody. The model demonstrated that the temporal or seasonal variation of PAHs in soil and waterbody can be well reproduced. Meanwhile, the spatial distribution of PAHs showed that urban areas in TR watershed have the highest PAH loadings compared to rural areas. Sensitivity analyses of the PAH soil and PAH water parameters were also able to determine the critical processes in TR watershed: degradation, deposition, volatilization, and wash off mechanism. We hope that this model will be able to aid the stakeholders in: regulating PAH concentrations emitted by various sources; and also apply the model to other Persistent Organic Pollutants (POPs).

Development of a multimedia model (POPsLTEA) to assess the influence of climate change on the fate and transport of polycyclic aromatic hydrocarbons in East Asia

A dynamic multimedia model (POPsLTEA) for an East Asia region was developed and evaluated to quantitatively assess how climate change (CC) alters the environmental fate and transport dynamics of 16 polycyclic aromatic hydrocarbons (PAHs) in air, water, soil, and sediment. To cover the entire model domain (25°N-50°N and 98°E-148°E) where China, Japan, and South and North Koreas are of primary concern, a total of 5000 main cells of 50km×50km size were used while 1008 cells of a finer spatial resolution (12.5km×12.5km) was nested for South Korea (33°N-38°N and 126°E-132°E). Most of the predicted concentrations agreed with the observed values within one order of magnitude with a tendency of overestimation for air and sediment. Prediction of the atmospheric concentration was statistically significant in both coincidence and association, suggesting the model's potential to successfully predict the fate and transport of the PAHs as influenced by CC. An example study of benzo(a)pyrene demonstrates that direction and strength of the CC influence on the pollution levels vary with the location and environmental media. As compared to the five year period of 2011 to 2015, the changes across the model domain in the annual geometric mean concentration over the years of 2021 through 2100 were predicted to range from 88% to 304%, from 84% to 109%, from 32% to 362%, and from 49% to 303%, in air, soil, surface water, and sea water, respectively, under the scenario of RCP8.5.

Urban and rural transport of semivolatile organic compounds at regional scale: A multimedia model approach

Urban areas are generally regarded as major sources of some semivolatile organic compounds and other persistent organic pollutants (POPs) to the surrounding regions. Huge differences in contaminant emissions between urban and rural areas directly affect their fate in environmental media. Little is known about POPs behavior between urban and rural areas at a regional scale. A spatially resolved Berkeley-Trent-Urban-Rural Fate Model (BETR-UR) was designed by coupling land cover information to simulate the transport of POPs between urban and rural areas, and the Bohai Rim was used as a case study to estimate Polycyclic Aromatic Hydrocarbon (PAH) fate. The processes of contaminant fate including emission, inter-compartmental transfer, advection and degradation in urban and rural areas were simulated in the model. Simulated PAH concentrations in environmental media of urban and rural areas were very close to measured values. The model accuracy was highly improved, with the average absolute relative error for PAH concentrations reduced from 37% to 3% compared with unimproved model results. PAH concentrations in urban soil and air were considerably higher than those in rural areas. Sensitivity analysis showed temperature was the most influential parameter for Phen rather than for Bap, whose fate was more influenced by emission rate, compartment dimension, transport velocity and chemical persistence. Uncertainty analysis indicated modeled results in urban media had higher uncertainty than those in rural areas due to larger variations of emissions in urban areas. The differences in urban and rural areas provided us with valuable guidance on policy setting for urban-rural POP control.

A first European scale multimedia fate modelling of BDE-209 from 1970 to 2020

The European Variant Berkeley Trent (EVn-BETR) multimedia fugacity model is used to test the validity of previously derived emission estimates and predict environmental concentrations of the main decabromodiphenyl ether congener, BDE-209. The results are presented here and compared with measured environmental data from the literature. Future multimedia concentration trends are predicted using three emission scenarios (Low, Realistic and High) in the dynamic unsteady state mode covering the period 1970-2020. The spatial and temporal distributions of emissions are evaluated. It is predicted that BDE-209 atmospheric concentrations peaked in 2004 and will decline to negligible levels by 2025. Freshwater concentrations should have peaked in 2011, one year after the emissions peak with sediment concentrations peaking in 2013. Predicted atmospheric concentrations are in good agreement with measured data for the Realistic (best estimate of emissions) and High (worst case scenario) emission scenarios. The Low emission scenario consistently underestimates measured data. The German unilateral ban on the use of DecaBDE in the textile industry is simulated in an additional scenario, the effects of which are mainly observed within Germany with only a small effect on the surrounding areas. Overall, the EVn-BTER model predicts atmospheric concentrations reasonably well, within a factor of 5 and 1.2 for the Realistic and High emission scenarios respectively, providing partial validation for the original emission estimate. Total mean MEC:PEC shows the High emission scenario predicts the best fit between air, freshwater and sediment data. An alternative spatial distribution of emissions is tested, based on higher consumption in EBFRIP member states, resulting in improved agreement between MECs and PECs in comparison with the Uniform spatial distribution based on population density. Despite good agreement between modelled and measured point data, more long-term monitoring datasets are needed to compare predicted trends in concentration to determine the rate of change of POPs within the environment.

Uncertainty in multi-media fate and transport models: a case study for TNT life cycle assessment

Life cycle assessment (LCA) is an evaluation method used by decision-makers to help assess the relative environmental impacts of various industrial processes. Despite that many LCA methods remain sensitive to uncertain input data, which can reduce the utility of their results, uncertainty arising from constituent LCA models remains poorly understood. Here, we begin to address this problem by evaluating the extent to which parameter-value uncertainty affects the SimpleBox 2.0 fate and transport model, which serves as a backbone for many LCA ecotoxicological impact categories. Two Monte Carlo type sampling methods were used to evaluate dispersion in steady-state concentration values for three chemicals involved in grenade production: toluene, 2,4-dinitrotoluene (2,4-DNT), and 2,4,6-trinitrotoluene (TNT). Parameters were first sampled stochastically one-at-a-time, then by randomly exploring a local patch of the parameter space. We confirmed that global temperatures contribute primarily to the overall variance of model results, which at most spanned approximately 8 decades in magnitude. These results are consistent with previous results obtained for the whole of the LCA method. LCA methods carry out calculations iteratively; a reduction in the error of a single component, such as the fate and transport model, may therefore improve its performance and utility as a decision-making aid.

Need of accurate model prediction of variability of the concentration ratio for testing coherence among environmental quality objectives: a case study of polycyclic aromatic hydrocarbons

Maintaining coherence among environmental quality objectives (EQOs) should be an important consideration for the EQOs to be met simultaneously. The objectives of the present work were to demonstrate the need of accurate variability prediction by models and to present considerations in selecting models for testing coherence of the EQOs. SimpleBox and POPsME were chosen as the two different types of models to compare the prediction variability and its influence on the results of coherence test among the maximum permissible concentrations (MPCs) of polycyclic aromatic hydrocarbons (PAHs) in South Korea. False calls by these models on coherence were found to occur often due to inaccurate prediction of variability in the concentration ratio at steady state, strongly suggesting that models for coherence test should be accurate in predicting not only the point value representing the concentration ratio but the variability of the value. It was demonstrated that spatially resolved dynamic models would have an intrinsic advantage over one box steady state models in reducing the rate of false negative call.

Evaluating officially reported polycyclic aromatic hydrocarbon emissions in the Athabasca oil sands region with a multimedia fate model

Emissions of organic substances with potential toxicity to humans and the environment are a major concern surrounding the rapid industrial development in the Athabasca oil sands region (AOSR). Although concentrations of polycyclic aromatic hydrocarbons (PAHs) in some environmental samples have been reported, a comprehensive picture of organic contaminant sources, pathways, and sinks within the AOSR has yet to be elucidated. We sought to use a dynamic multimedia environmental fate model to reconcile the emissions and residue levels reported for three representative PAHs in the AOSR. Data describing emissions to air compiled from two official sources result in simulated concentrations in air, soil, water, and foliage that tend to fall close to or below the minimum measured concentrations of phenanthrene, pyrene, and benzo(a)pyrene in the environment. Accounting for evaporative emissions (e.g., from tailings pond disposal) provides a more realistic representation of PAH distribution in the AOSR. Such indirect emissions to air were found to be a greater contributor of PAHs to the AOSR atmosphere relative to reported direct emissions to air. The indirect pathway transporting uncontrolled releases of PAHs to aquatic systems via the atmosphere may be as significant a contributor of PAHs to aquatic systems as other supply pathways. Emission density estimates for the three PAHs that account for tailings pond disposal are much closer to estimated global averages than estimates based on the available emissions datasets, which fall close to the global minima. Our results highlight the need for improved accounting of PAH emissions from oil sands operations, especially in light of continued expansion of these operations.

Using gridded multimedia model to simulate spatial fate of Benzo[α]pyrene on regional scale

Predicting the environmental multimedia fate is an essential step in the process of assessing the human exposure and health impacts of chemicals released into the environment. Multimedia fate models have been widely applied to calculate the fate and distribution of chemicals in the environment, which can serve as input to a human exposure model. In this study, a grid based multimedia fugacity model at regional scale was developed together with a case study modeling the fate and transfer of Benzo[α]pyrene (BaP) in Bohai coastal region, China. Based on the estimated emission and in-site survey in 2008, the BaP concentrations in air, vegetation, soil, fresh water, fresh water sediment and coastal water as well as the transfer fluxes were derived under the steady-state assumption. The model results were validated through comparison between the measured and modeled concentrations of BaP. The model results indicated that the predicted concentrations of BaP in air, fresh water, soil and sediment generally agreed with field observations. Model predictions suggest that soil was the dominant sink of BaP in terrestrial systems. Flow from air to soil, vegetation and costal water were three major pathways of BaP inter-media transport processes. Most of the BaP entering the sea was transferred by air flow, which was also the crucial driving force in the spatial distribution processes of BaP. The Yellow River, Liaohe River and Daliao River played an important role in the spatial transformation processes of BaP. Compared with advection outflow, degradation was more important in removal processes of BaP. Sensitivities of the model estimates to input parameters were tested. The result showed that emission rates, compartment dimensions, transport velocity and degradation rates of BaP were the most influential parameters for the model output. Monte Carlo simulation was carried out to determine parameter uncertainty, from which the coefficients of variation for the estimated BaP concentrations in air and soil were computed, which were 0.46 and 1.53, respectively. The model output-concentrations of BaP in multimedia environment can be used in human exposure and risk assessment in the Bohai coastal region. The results also provide significant indicators on the likely dominant fate, influence range of emission and transport processes determining behavior of BaP in the Bohai coastal region, which is instrumental in human exposure and risk assessment in the region.

Quantifying the trade-off between parameter and model structure uncertainty in life cycle impact assessment

To enhance the use of quantitative uncertainty assessments in life cycle impact assessment practice, we suggest to quantify the trade-off between parameter uncertainty, i.e. any uncertainty associated with data and methods used to quantify the model parameters, and model structure uncertainty, i.e. the uncertainty about the relations and mechanisms being studied. In this paper we show the trade-off between the two types of uncertainty in a case of maize production with a focus on freshwater ecotoxicity due to pesticide application in The Netherlands. Parameter uncertainty in pesticide emissions, chemical-specific data, effect and damage data, and fractions of metabolite formation of degradation products was statistically quantified via probabilistic simulation, i.e. Monte Carlo simulation. Model structure uncertainties regarding the concentration-response model to be included, the selection of the damage model, and the inclusion of pesticide transformation products were assessed via discrete choice analysis. We conclude that to arrive at a minimum level of overall uncertainty the linear concentration-response model is preferable, while the transformation products may be excluded. Selecting the damage model has a relatively low influence on the overall uncertainty. Our study shows that quantifying the trade-off between different types of uncertainty can help to identify optimal model complexity from an uncertainty point of view.

Description and propagation of uncertainty in input parameters in environmental fate models

Today, chemical risk and safety assessments rely heavily on the estimation of environmental fate by models. The key compound-related properties in such models describe partitioning and reactivity. Uncertainty in determining these properties can be separated into random and systematic (incompleteness) components, requiring different types of representation. Here, we evaluate two approaches that are suitable to treat also systematic errors, fuzzy arithmetic, and probability bounds analysis. When a best estimate (mode) and a range can be computed for an input parameter, then it is possible to characterize the uncertainty with a triangular fuzzy number (possibility distribution) or a corresponding probability box bound by two uniform distributions. We use a five-compartment Level I fugacity model and reported empirical data from the literature for three well-known environmental pollutants (benzene, pyrene, and DDT) as illustrative cases for this evaluation. Propagation of uncertainty by discrete probability calculus or interval arithmetic can be done at a low computational cost and gives maximum flexibility in applying different approaches. Our evaluation suggests that the difference between fuzzy arithmetic and probability bounds analysis is small, at least for this specific case. The fuzzy arithmetic approach can, however, be regarded as less conservative than probability bounds analysis if the assumption of independence is removed. Both approaches are sensitive to repeated parameters that may inflate the uncertainty estimate. Uncertainty described by probability boxes was therefore also propagated through the model by Monte Carlo simulation to show how this problem can be avoided.

Uncertainty in environmental health impact assessment: quantitative methods and perspectives

Environmental health impact assessment models are subjected to great uncertainty due to the complex associations between environmental exposures and health. Quantifying the impact of uncertainty is important if the models are used to support health policy decisions. We conducted a systematic review to identify and appraise current methods used to quantify the uncertainty in environmental health impact assessment. In the 19 studies meeting the inclusion criteria, several methods were identified. These were grouped into random sampling methods, second-order probability methods, Bayesian methods, fuzzy sets, and deterministic sensitivity analysis methods. All 19 studies addressed the uncertainty in the parameter values but only 5 of the studies also addressed the uncertainty in the structure of the models. None of the articles reviewed considered conceptual sources of uncertainty associated with the framing assumptions or the conceptualisation of the model. Future research should attempt to broaden the way uncertainty is taken into account in environmental health impact assessments.

Assessing model uncertainty of bioaccumulation models by combining chemical space visualization with a process-based diagnostic approach

As models describing human exposure to organic chemicals gain wider use in chemical risk assessment and management, it becomes important to understand their uncertainty. Although evaluation of parameter sensitivity/uncertainty is increasingly common, model uncertainty is rarely assessed. When it is, the assessment is generally limited to a handful of chemicals. In this study, a strategy for more comprehensive model uncertainty assessment was developed. A regulatory model (EUSES) was compared with a research model based on more recent science. Predicted human intake was used as the model end point. Chemical space visualization techniques showed that the extent of disagreement between the models varied strongly with chemical partitioning properties. For each region of disagreement, the primary human exposure vector was determined. The differences between the models' process algorithms describing these exposure vectors were identified and evaluated. The equilibrium assumption for root crops in EUSES caused overestimations in daily intake of superhydrophobic chemicals (log K(OW) > 11, log K(OA) > 10), whereas EUSES's approach to calculating bioaccumulation in fish prey resulted in underestimations for hydrophobic compounds (log K(OW) ∼ 6-8). Uptake of hydrophilic chemicals from soil and bioaccumulation of superhydrophobic chemicals in zooplankton were identified as important research areas to enable further reduction of model uncertainty in bioaccumulation models.

Modeled and monitored variation in space and time of PCB-153 concentrations in air, sediment, soil and aquatic biota on a European scale

We evaluated various modeling options for estimating concentrations of PCB-153 in the environment and in biota across Europe, using a nested multimedia fate model coupled with a bioaccumulation model. The most detailed model set up estimates concentrations in air, soil, fresh water sediment and fresh water biota with spatially explicit environmental characteristics and spatially explicit emissions to air and water in the period 1930-2005. Model performance was evaluated with the root mean square error (RMSE(log)), based on the difference between estimated and measured concentrations. The RMSE(log) was 5.4 for air, 5.6-6.3 for sediment and biota, and 5.5 for soil in the most detailed model scenario. Generally, model estimations tended to underestimate observed values for all compartments, except air. The decline in observed concentrations was also slightly underestimated by the model for the period where measurements were available (1989-2002). Applying a generic model setup with averaged emissions and averaged environmental characteristics, the RMSE(log) increased to 21 for air and 49 for sediment. For soil the RMSE(log) decreased to 3.5. We found that including spatial variation in emissions was most relevant for all compartments, except soil, while including spatial variation in environmental characteristics was less influential. For improving predictions of concentrations in sediment and aquatic biota, including emissions to water was found to be relevant as well.

Interrelationship of Pyrogenic Polycyclic Aromatic Hydrocarbon (PAH) Contamination in Different Environmental Media

Interrelationships between pyrogenic polycyclic aromatic hydrocarbons (PAHs) were assessed in air, soil, water, sediment, and tree leaves by using multi-media monitoring data. Concurrent concentration measurements were taken bimonthly for a year for the multi-media at urban and suburban sites. PAH level correlations between air and other media were observed at the urban site but were less clear at the suburban site. Considering a closer PAHs distribution/fate characteristics to soil than suspended solids, contamination in sediment seemed to be governed primarily by that in soil. The partitioning of PAHs in waters could be better accounted for by sorption onto black carbon and dissolved organic carbon.