Application of gas chromatographic data and 2D molecular descriptors for accurate global mobility potential prediction

Mobility is a key feature affecting the environmental fate, which is of particular importance in the case of persistent organic pollutants (POPs) and emerging pollutants (EPs). In this study, the global mobility classification artificial neural networks-based models employing GC retention times (RT) and 2D molecular descriptors were constructed and validated. The high usability of RT was confirmed based on the feature selection step performed using the multivariate adaptive regression splines (MARS) tool. Although RT was found to be the most important, according to Kruskal-Wallis ANOVA analysis, it is insufficient to build a robust model, which justifies the need to expand the input layer with 2D descriptors. Therefore the following molecular descriptors: MPC10, WTPT-2, AATS8s, minaaCH, GATS7c, RotBtFrac, ATSC7v and ATSC1p, which were characterized by a high predicting potential were used to improve the classification performance. As a result of machine learning procedure ten of the most accurate neural networks were selected. The external validation showed that the final models are characterized by a high general accuracy score (85.71-96.43%). The high predicting abilities were also confirmed by the micro-averaged Matthews correlation coefficient (MAMCC) (0.73-0.88). To evaluate the applicability of the models, new retention times of selected POPs and EPs including pesticides, polycyclic aromatic hydrocarbons, pharmaceuticals, fragrances and personal care products were measured and used for mobility prediction. Further, the classifiers were used for photodegradation and chlorination products of two popular sunscreen agents, 2-ethyl-hexyl-4-methoxycinnamate and 2-ethylhexyl 4-(dimethylamino)benzoate.

The Emissions Fractions Approach to Assessing the Long-Range Transport Potential of Organic Chemicals

The assessment of long-range transport potential (LRTP) is enshrined in several frameworks for chemical regulation such as the Stockholm Convention. Screening for LRTP is commonly done with the OECD Pov and LRTP Screening Tool employing two metrics, characteristic travel distance (CTD) and transfer efficiency (TE). Here we introduce a set of three alternative metrics and implement them in the Tool's model. Each metric is expressed as a fraction of the emissions in a source region. The three metrics quantify the extent to which the chemical (i) reaches a remote region (dispersion, ϕ1), (ii) is transferred to surface media in the remote region (transfer, ϕ2), and (iii) accumulates in these surface media (accumulation, ϕ3). In contrast to CTD and TE, the emissions fractions metrics can integrate transport via water and air, enabling comprehensive LRTP assessment. Furthermore, since there is a coherent relationship between the three metrics, the new approach provides quantitative mechanistic insight into different phenomena determining LRTP. Finally, the accumulation metric, ϕ3, allows assessment of LRTP in the context of the Stockholm Convention, where the ability of a chemical to elicit adverse effects in surface media is decisive. We conclude that the emission fractions approach has the potential to reduce the risk of false positives/negatives in LRTP assessments.

Moving persistence assessments into the 21st century: A role for weight-of-evidence and overall persistence

Assessing the persistence of chemicals in the environment is a key element in existing regulatory frameworks to protect human health and ecosystems. Persistence in the environment depends on many fate processes, including abiotic and biotic transformations and physical partitioning, which depend on substances' physicochemical properties and environmental conditions. A main challenge in persistence assessment is that existing frameworks rely on simplistic and reductionist evaluation schemes that may lead substances to be falsely assessed as persistent or the other way around-to be falsely assessed as nonpersistent. Those evaluation schemes typically assess persistence against degradation half-lives determined in single-compartment simulation tests or against degradation levels measured in stringent screening tests. Most of the available test methods, however, do not apply to all types of substances, especially substances that are poorly soluble, complex in composition, highly sorptive, or volatile. In addition, the currently applied half-life criteria are derived mainly from a few legacy persistent organic pollutants, which do not represent the large diversity of substances entering the environment. Persistence assessment would undoubtedly benefit from the development of more flexible and holistic evaluation schemes including new concepts and methods. A weight-of-evidence (WoE) approach incorporating multiple influencing factors is needed to account for chemical fate and transformation in the whole environment so as to assess overall persistence. The present paper's aim is to begin to develop an integrated assessment framework that combines multimedia approaches to organize and interpret data using a clear WoE approach to allow for a more consistent, transparent, and thorough assessment of persistence. Integr Environ Assess Manag 2022;18:868-887. © 2021 ExxonMobil Biomedical Sciences, Inc. Integrated Environmental Assessment and Management published by Wiley Periodicals LLC on behalf of Society of Environmental Toxicology & Chemistry (SETAC).

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.

Introducing a nested multimedia fate and transport model for organic contaminants (NEM)

Some organic contaminants, including the persistent organic pollutants (POPs), have achieved global distribution through long range atmospheric transport (LRAT). Regulatory efforts, monitoring programs and modelling studies address the LRAT of POPs on national, continental (e.g. Europe) and/or global scales. Whereas national and continental-scale models require estimates of the input of globally dispersed chemicals from outside of the model domain, existing global-scale models either have relatively coarse spatial resolution or are so computationally demanding that it limits their usefulness. Here we introduce the Nested Exposure Model (NEM), which is a multimedia fate and transport model that is global in scale yet can achieve high spatial resolution of a user-defined target region without huge computational demands. Evaluating NEM by comparing model predictions for PCB-153 in air with measurements at nine long-term monitoring sites of the European Monitoring and Evaluation Programme (EMEP) reveals that nested simulations at a resolution of 1°× 1° yield results within a factor of 1.5 of observations at sites in northern Europe. At this resolution, the model attributes more than 90% of the atmospheric burden within any of the grid cells containing an EMEP site to advective atmospheric transport from elsewhere. Deteriorating model performance with decreasing resolution (15°× 15°, 5°× 5° and 1°× 1°), manifested by overestimation of concentrations across most of northern Europe by more than a factor of 3, illustrates the effect of numerical diffusion. Finally, we apply the model to demonstrate how the choice of spatial resolution affect predictions of atmospheric deposition to the Baltic Sea. While we envisage that NEM may be used for a wide range of applications in the future, further evaluation will be required to delineate the boundaries of applicability towards chemicals with divergent fate properties as well as in environmental media other than air.

Georeferenced multimedia environmental fate of volatile methylsiloxanes modeled in the populous Tokyo Bay catchment basin

We investigated the multimedia fate of decamethylcyclopentasiloxane (D5) and dodecamethylcyclohexasiloxane (D6) in the densely populated catchment basin of Tokyo Bay, Japan, by using a georeferenced multimedia model. We estimated the daily per person consumption rate of these compounds in Japan according to literature. Emissions to the atmosphere accounted for almost all of the emissions of these compounds to the environment. The majority of these compounds was predicted to be distributed in the atmosphere (about 60%) and sediment (about 40%). The advective flows in and out of the atmosphere over the Tokyo Bay catchment basin dominated the flows of these compounds. The sewerage systems contributed considerably to the transport and fate of D5 and D6 in water. They transported these compounds from households to discharge outlets of sewage treatment plants (STPs), which in turn accounted for approximately one quarter of the emission of these compounds to rivers and to Tokyo Bay. The wastewater treatment plants also effectively removed these compounds from the wastewater. The overall persistence of D5 and D6 in the catchment basin was estimated to be 3.8-9.5 days. The horizontal distributions of these compounds were similar among environmental compartments; high concentrations were generally observed in populated areas. A sensitivity analysis demonstrated that the D5 discharge rate to Tokyo Bay excluding the direct discharge from STPs and the D5 mass in the river compartment were sensitive to changes of the organic-carbon-water partition coefficient. Comparison with the concentrations in rivers measured recently in the target area showed that the model captured overall trends of low to high concentrations in rivers. However, there was some variability and a bias toward underprediction. The model provided a better fit to measurements for D5 than for D6. One potential factor contributing to the bias toward underprediction was underestimation of the consumption rates.

Multiscale Spatial Modeling of Human Exposure from Local Sources to Global Intake

Exposure studies, used in human health risk and impact assessments of chemicals, are largely performed locally or regionally. It is usually not known how global impacts resulting from exposure to point source emissions compare to local impacts. To address this problem, we introduce Pangea, an innovative multiscale, spatial multimedia fate and exposure assessment model. We study local to global population exposure associated with emissions from 126 point sources matching locations of waste-to-energy plants across France. Results for three chemicals with distinct physicochemical properties are expressed as the evolution of the population intake fraction through inhalation and ingestion as a function of the distance from sources. For substances with atmospheric half-lives longer than a week, less than 20% of the global population intake through inhalation (median of 126 emission scenarios) can occur within a 100 km radius from the source. This suggests that, by neglecting distant low-level exposure, local assessments might only account for fractions of global cumulative intakes. We also study ∼10 000 emission locations covering France more densely to determine per chemical and exposure route which locations minimize global intakes. Maps of global intake fractions associated with each emission location show clear patterns associated with population and agriculture production densities.

Modelling exposure of oceanic higher trophic-level consumers to polychlorinated biphenyls: pollution 'hotspots' in relation to mass mortality events of marine mammals

Marine mammals in the past mass mortality events may have been susceptible to infection because their immune systems were suppressed through the bioaccumulation of environmental pollutants such as polychlorinated biphenyls (PCBs). We compiled mortality event data sets of 33 marine mammal species, and employed a Finely-Advanced Transboundary Environmental model (FATE) to model the exposure of the global fish community to PCB congeners, in order to define critical exposure levels (CELs) of PCBs above which mass mortality events are likely to occur. Our modelling approach enabled us to describe the mass mortality events in the context of exposure of higher-trophic consumers to PCBs and to identify marine pollution 'hotspots' such as the Mediterranean Sea and north-western European coasts. We demonstrated that the CELs can be applied to quantify a chemical pollution Planetary Boundary, under which a safe operating space for marine mammals and humanity can exist.

Inter-model comparison of the landscape determinants of vector-borne disease: implications for epidemiological and entomological risk modeling

Extrapolating landscape regression models for use in assessing vector-borne disease risk and other applications requires thoughtful evaluation of fundamental model choice issues. To examine implications of such choices, an analysis was conducted to explore the extent to which disparate landscape models agree in their epidemiological and entomological risk predictions when extrapolated to new regions. Agreement between six literature-drawn landscape models was examined by comparing predicted county-level distributions of either Lyme disease or Ixodes scapularis vector using Spearman ranked correlation. AUC analyses and multinomial logistic regression were used to assess the ability of these extrapolated landscape models to predict observed national data. Three models based on measures of vegetation, habitat patch characteristics, and herbaceous landcover emerged as effective predictors of observed disease and vector distribution. An ensemble model containing these three models improved precision and predictive ability over individual models. A priori assessment of qualitative model characteristics effectively identified models that subsequently emerged as better predictors in quantitative analysis. Both a methodology for quantitative model comparison and a checklist for qualitative assessment of candidate models for extrapolation are provided; both tools aim to improve collaboration between those producing models and those interested in applying them to new areas and research questions.

A new metric for long-range transport potential of chemicals

We propose a new metric for long-range transport potential (LRTP), GIF, based on source-receptor analyses and evaluate the LRTP and persistence of a wide variety of chlorinated and brominated organic compounds using GIF and overall persistence (POV), respectively. We calculated GIF and POV using our global 3D dynamic multimedia model (FATE). Physicochemical properties were obtained from quantitative structure-property relationship (QSPR) models. The FATE-QSPR combined model enabled us to systematically investigate the LRTP and persistence of a wide variety of chemical substances. On average, the estimated GIF and POV for chlorinated compounds were larger than those for their brominated counterparts, with the largest and smallest values found for polychlorinated biphenyls and polybrominated dibenzodioxins, respectively. We also compared GIF with four differently defined LRTP metrics and two LRTP metrics obtained from a simple model. The results of our analyses indicate that the LRTP ranks can differ considerably among LRTP metrics, the differences being dependent on the governing environmental processes, relevant physicochemical properties, and multimedia model.

Understanding quantitative structure-property relationships uncertainty in environmental fate modeling

In cases in which experimental data on chemical-specific input parameters are lacking, chemical regulations allow the use of alternatives to testing, such as in silico predictions based on quantitative structure-property relationships (QSPRs). Such predictions are often given as point estimates; however, little is known about the extent to which uncertainties associated with QSPR predictions contribute to uncertainty in fate assessments. In the present study, QSPR-induced uncertainty in overall persistence (POV ) and long-range transport potential (LRTP) was studied by integrating QSPRs into probabilistic assessments of five polybrominated diphenyl ethers (PBDEs), using the multimedia fate model Simplebox. The uncertainty analysis considered QSPR predictions of the fate input parameters' melting point, water solubility, vapor pressure, organic carbon-water partition coefficient, hydroxyl radical degradation, biodegradation, and photolytic degradation. Uncertainty in POV and LRTP was dominated by the uncertainty in direct photolysis and the biodegradation half-life in water. However, the QSPRs developed specifically for PBDEs had a relatively low contribution to uncertainty. These findings suggest that the reliability of the ranking of PBDEs on the basis of POV and LRTP can be substantially improved by developing better QSPRs to estimate degradation properties. The present study demonstrates the use of uncertainty and sensitivity analyses in nontesting strategies and highlights the need for guidance when compounds fall outside the applicability domain of a QSPR.

High-resolution inventory of Japanese anthropogenic mercury emissions

Heavy metals like mercury that are emitted into the environment remain there indefinitely, posing a long-term threat to both the environment and human health. Elemental mercury is volatile and is in gaseous form, and because of the long residence time, transported over long distances. Comprehensive control of mercury emissions therefore remains an important international issue. The crucial steps for designing effective approaches for such control include the quantification of mercury emissions by sources and the identification of geographical characteristics of the emissions. In this study a detailed, high-resolution inventory of Japanese mercury emissions in 2005 was developed to improve understanding of their geographical distribution. Proceeding from a national emissions inventory per source category, emissions were spatially allocated with increasing geographical resolution in a stepwise procedure using statistics from geographic information resources, yielding mercury emissions per prefecture, per municipality and per grid cell of approximately 1 × 1 km. The five prefectures with the highest emissions were Fukuoka, Yamaguchi, Hyogo, Oita, and Hokkaido, accounting for 35.2% of all emissions. In each prefecture a small number of municipalities account for a major share of emissions. Distribution by grid cell is characterized by a concentration of 50% of all emissions in a mere 32 of the 255 954 grid cells over which emissions are distributed in this study. It was also quantitatively confirmed that use of larger grid cells leads to greater uncertainty in emissions distribution. Problems with data collection are clarified and measures to improve the accuracy of future estimation are proposed.

Are reductions in industrial organic contaminants emissions in rich countries achieved partly by export of toxic wastes?

Recent studies show that PCB (polychlorinated biphenyl) air concentrations remain surprisingly high in parts of Africa and Asia. These are regions where PCBs were never extensively used, but which are implicated as recipients of obsolete products and wastes containing PCBs and other industrial organic contaminants, such as halogenated flame retardants (HFRs). We hypothesize that there may be different trends in emissions across the globe, whereby emissions of some industrial organic contaminants may be decreasing faster in former use regions (due to emission reductions combined with uncontrolled export), at the expense of regions receiving these substances as obsolete products and wastes. We conclude that the potential for detrimental effects on the environment and human health due to long-range transport by air, water, or wastes should be of equal concern when managing and regulating industrial organic contaminants. This calls for a better integration of life-cycle approaches in the management and regulation of industrial organic contaminants in order to protect environmental and human health on a global scale. Yet, little remains known about the amounts of industrial organic contaminants exported outside former use regions as different types of wastes because of the often illicit nature of these operations.

On the replacement of empirical parameters in multimedia mass balance models with QSPR data

Official OECD recommendations give the highest priority to application of purely empirical phys/chem data (partition coefficients, environmental half-live times etc.) in multimedia mass balance modeling of environmental overall persistence and long-range transport for potentially hazardous chemicals. We have demonstrated that the replacement of the empirical data with those predicted by employing Quantitative Structure-Property Relationships (QSPR) technique did not significantly decrease the performance of The Tool 2.0--the OECD multimedia mass balance model. To prove this, we compared each other the output results (overall persistence -P(OV); characteristic traveling distance -CTD and transport efficiency TE) obtained from 6 of multimedia models. The models utilized combinations of experimentally determined and QSPR-predicted values of the partition coefficients and half-live times. For predicting phys/chem data, we utilized 2 QSPRs developed in our laboratory and the EPI Suite package (US EPA). We did not observe any statistically significant (p<0.05) differences between the models. This conclusion is important, because it leads to reducing time and costs of laboratory studies required during the risk assessment procedure. Moreover, regarding the obtained results, we proposed to replace the single-threshold approaches established by majority of international regulations to screen substances for persistence, bioaccumulation and long-range transport potential with the approaches taking into account uncertainty of the results and/or probability of passing a given threshold.

Spatially explicit multimedia fate models for pollutants in Europe: state of the art and perspectives

A review by Hollander et al. (in preparation), discusses the relative potentials, advantages and shortcomings of spatial and non spatial models of chemical fate, highlighting that spatially explicit models may be needed for specific purposes. The present paper reviews the state of the art in spatially explicit chemical fate and transport modeling in Europe. We summarize the three main approaches currently adopted in spatially explicit modeling, namely (1) multiple box models, (2) numerical solutions of simultaneous advection-dispersion equations (ADE) in air, soil and water, and (3) the development of meta-models. As all three approaches experience limitations, we describe in further detail geographic information system (GIS)-based modeling as an alternative approach allowing a simple, yet spatially explicit description of chemical fate. We review the input data needed, and the options available for their retrieval at the European scale. We also discuss the importance of, and limitations in model evaluation. We observe that the high uncertainty in chemical emissions and physico-chemical behavior in the environment make realistic simulations difficult to obtain. Therefore we envisage a shift in model use from process simulation to hypothesis testing, in which explaining the discrepancies between observed and computed chemical concentrations in the environment takes importance over prediction per se. This shift may take advantage of using simple models in GIS with residual uses of complex models for detailed studies. It also calls for tighter joint interpretation of models and spatially distributed monitoring datasets, and more refined spatial representation of environmental drivers such as landscape and climate variables, and better emission estimates. In summary, we conclude that the problem is not "how to compute" (i.e. emphasis on numerical methods, spatial/temporal discretization, quantitative uncertainty and sensitivity analysis...) but "what to compute" (i.e. emphasis on spatial distribution of emissions, and the depiction of appropriate spatial patterns of environmental drivers).