Model comparison for risk assessment: a case study of contaminated groundwater

Many environmental multimedia risk assessment models have been developed and widely used along with increasing sophistication of the risk assessment method. Despite of the considerable improvement, uncertainty remains a primary threat to the credibility of and users' confidence in the model-based risk assessments. In particular, it has been indicated that scenario and model uncertainty may affect significantly the assessment outcome. Furthermore, the uncertainty resulting from choosing different models has been shown more important than that caused by parameter uncertainty. Based on the relationship between exposure pathways and estimated risk results, this study develops a screening procedure to compare the relative suitability between potential multimedia models, which would facilitate the reduction of uncertainty due to model selection. MEPAS, MMSOILS, and CalTOX models, combined with Monte Carlo simulation, are applied to a realistic groundwater-contaminated site to demonstrate the process. It is also shown that the identification of important parameters and exposure pathways, and implicitly, the subsequent design of uncertainty reduction and risk management measures, would be better-formed.     

Modeling traffic air pollution in street canyons in New York City for intra-urban exposure assessment in the US Multi-Ethnic Study of atherosclerosis and air pollution

We evaluated the Danish AirGIS air quality and exposure model system using air quality measurement data from New York City in the Multi-Ethnic Study of Atherosclerosis and Air Pollution (MESA Air). Measurements were used from three US EPA Air Quality System (AQS) monitoring stations and a comprehensive MESA Air measurement campaign including about 150 different locations and about 650 samples of about 2 week measurements of NO_x, NO₂ and PM_2.5. AirGIS is a deterministic exposure model system based on the dispersion models Operational Street Pollution Model (OSPM) and the Urban Background Model (UBM). The UBM model reproduced the annual levels within 1–26% depending on station and pollutant at the three urban background EPA monitor stations, and generally reproduced well the seasonal and diurnal variation. The full model with OSPM and UBM reproduced the MESA Air measurements with a correlation coefficient of r² = 0.51 for NO_x, r² = 0.28 for NO₂ and r² = 0.73 for PM_2.5.     

Regional-scale impacts of Phase I of the Clean Air Act Amendments in the USA: the relation between emissions and concentrations,both wet and dry

A comparison of data records in the 1990s, both before (1991–1994) and after (1995–1997) implementation of Phase I of the Clean Air Act Amendments (CAAA) of 1990 for the eastern US, shows a significant reduction in SO₂ emissions for most states, except for Texas, North Carolina, Illinois, Florida, and Alabama. However, of the major NO_x emitting states, only two eastern states (New York and Pennsylvania) show significant declines in NO_x. A pattern of large declines in SO₂ emissions (>20%) after CAAA implementation, and large declines in precipitation SO₄²⁻ and H⁺, as well as air concentrations of SO₂ and SO₄²⁻ (components of dry deposition), exists for most regions of the eastern US. In most cases, the emission/concentration relations are close to 1 : 1 when the source region based on 15-h back trajectories is used for the New England region, and source regions based on 9-h back trajectories are used for the six other eastern US regions that were studied. The southern Appalachian Mountain region, an acid-sensitive area receiving high levels of acidic deposition, has not seen an appreciable improvement in precipitation acidity. This area has also shown the least improvement in wet and dry sulfur concentrations, of the areas examined. Precipitation base cations (Ca²⁺ and Mg²⁺) show a pattern of either increasing or level concentrations when comparing 1990–1994 to 1995–1998 data, for six of the seven regions examined. Ammonium concentrations have generally changed <10%, except for the Illinois and southern Appalachian Mtn. regions, which increased >15%.     

Improving bioaerosol exposure assessments of composting facilities — Comparative modelling of emissions from different compost ages and processing activities

We present bioaerosol source term concentrations from passive and active composting sources and compare emissions from green waste compost aged 1, 2, 4, 6, 8, 12 and 16 weeks. Results reveal that the age of compost has little effect on the bioaerosol concentrations emitted for passive windrow sources. However emissions from turning compost during the early stages may be higher than during the later stages of the composting process. The bioaerosol emissions from passive sources were in the range of 10³–10⁴ cfu m⁻³, with releases from active sources typically 1-log higher. We propose improvements to current risk assessment methodologies by examining emission rates and the differences between two air dispersion models for the prediction of downwind bioaerosol concentrations at off-site points of exposure. The SCREEN3 model provides a more precautionary estimate of the source depletion curves of bioaerosol emissions in comparison to ADMS 3.3. The results from both models predict that bioaerosol concentrations decrease to below typical background concentrations before 250 m, the distance at which the regulator in England and Wales may require a risk assessment to be completed.     

Graphene—a promising material for removal of perchlorate (ClO4 −) from water

A batch adsorption process was applied to investigate the removal of perchlorate (ClO₄ ^?) from water by graphene. In doing so, the thermodynamic adsorption isotherm and kinetic studies were also carried out. Graphene was prepared by a facile liquid-phase exfoliation. Graphene was characterized by Raman spectroscopy, Fourier-transform infrared spectroscopy, powder X-ray diffraction, scanning electron microscope, and zeta potential measurements. A systematic study of the adsorption process was performed by varying pH, ionic strength, and temperature. The adsorption efficiency of graphene was 99.2 %, suggesting that graphene is an excellent adsorbent for ClO₄ ^? removal from water. The rate constants for all these kinetic models were calculated, and the results indicate that second-order kinetics model was well suitable to model the kinetic adsorption of ClO₄ ^?. Equilibrium data were well described by the typical Langmuir adsorption isotherm. The experimental results showed that graphene is an excellent perchlorate adsorbent with an adsorbent capacity of up to 0.024 mg/g at initial perchlorate concentration of 2 mg/L and temperature of 298 K. Thermodynamic studies revealed that the adsorption reaction was a spontaneous and endothermic process. Graphene removed the perchlorate present in the water and reduced it to a permissible level making it drinkable.     

Correlations between fine particulate matter (PM2.5) and meteorological variables in the United States: Implications for the sensitivity of PM2.5 to climate change

We applied a multiple linear regression (MLR) model to study the correlations of total PM_2.5 and its components with meteorological variables using an 11-year (1998–2008) observational record over the contiguous US. The data were deseasonalized and detrended to focus on synoptic-scale correlations. We find that daily variation in meteorology as described by the MLR can explain up to 50% of PM_2.5 variability with temperature, relative humidity (RH), precipitation, and circulation all being important predictors. Temperature is positively correlated with sulfate, organic carbon (OC) and elemental carbon (EC) almost everywhere. The correlation of nitrate with temperature is negative in the Southeast but positive in California and the Great Plains. RH is positively correlated with sulfate and nitrate, but negatively with OC and EC. Precipitation is strongly negatively correlated with all PM_2.5 components. We find that PM_2.5 concentrations are on average 2.6 μg m^?3 higher on stagnant vs. non-stagnant days. Our observed correlations provide a test for chemical transport models used to simulate the sensitivity of PM_2.5 to climate change. They point to the importance of adequately representing the temperature dependence of agricultural, biogenic and wildfire emissions in these models.     

Description and evaluation of a model of deposition velocities for routine estimates of air pollutant dry deposition over North America.: Part I: model development

This paper describes the development of a detailed dry deposition model for routine computation of dry deposition velocities of SO₂, O₃, HNO₃ and fine particle SO₄²⁻ across much of North America. Four different dry deposition/surface exchange sub-models have been combined with the current Canadian weather forecast model (Global Environmental Multiscale model) with a 3 h time resolution and a horizontal spatial resolution of 35 km. The present model uses the US Geological Survey North American Land Cover Characteristics data to obtain fourteen different land use and five seasonal categories. The four sub-models used are a multi-layer model for gaseous species over taller canopy land-use types, a big-leaf model for gaseous species over lower canopies (including bare soil and water) and for HNO₃ under all surface types and, two different models for SO₄²⁻, one for tall canopies and the other for short canopies. All necessary parameters for each sub-model, chemical species, land-use and seasonal categories have been selected from available data libraries or from the values reported in the literature. The purpose for developing this model (referred to as the Routine Deposition Model (RDM)), when coupled with air concentration data, is to provide estimates of seasonal dry deposition, which can be combined with wet deposition to produce total deposition estimates. Model theory is discussed in this paper and model sensitivity tests and results will be presented in a companion paper.     

Spatial patterns of mobile source particulate matter emissions-to-exposure relationships across the United States

Assessing the public health benefits from air pollution control measures is assisted by understanding the relationship between mobile source emissions and subsequent fine particulate matter (PM_2.5) exposure. Since this relationship varies by location, we characterized its magnitude and geographic distribution using the intake fraction (iF) concept. We considered emissions of primary PM_2.5 as well as particle precursors SO₂ and NO_x from each of 3080 counties in the US. We modeled the relationship between these emissions and total US population exposure to PM_2.5, making use of a source–receptor matrix developed for health risk assessment. For primary PM_2.5, we found a median iF of 1.2 per million, with a range of 0.12–25. Half of the total exposure was reached by a median distance of 150 km from the county where mobile source emissions originated, though this spatial extent varied across counties from within the county borders to 1800 km away. For secondary ammonium sulfate from SO₂ emissions, the median iF was 0.41 per million (range: 0.050–10), versus 0.068 per million for secondary ammonium nitrate from NO_x emissions (range: 0.00092–1.3). The median distance to half of the total exposure was greater for secondary PM_2.5 (450 km for sulfate, 390 km for nitrate). Regression analyses using exhaustive population predictors explained much of the variation in primary PM_2.5 iF (R²=0.83) as well as secondary sulfate and nitrate iF (R²=0.74 and 0.60), with greater near-source contribution for primary than for secondary PM_2.5. We conclude that long-range dispersion models with coarse geographic resolution are appropriate for risk assessments of secondary PM_2.5 or primary PM_2.5 emitted from mobile sources in rural areas, but that more resolved dispersion models are warranted for primary PM_2.5 in urban areas due to the substantial contribution of near-source populations.     

Measurements show that marginal wells are a disproportionate source of methane relative to production

ABSTRACT

Oil and natural gas wells are a prominent source of the greenhouse gas methane (CH₄), but most measurements are from newer, high producing wells. There are nearly 700,000 marginal “stripper” wells in the US, which produce less than 15 barrels of oil equivalent (BOE) d^?1. We made direct measurements of CH₄ and volatile organic carbon (VOC) emissions from marginal oil and gas wells in the Appalachian Basin of southeastern Ohio, all producing < 1 BOE d^?1. Methane and VOC emissions followed a skewed distribution, with many wells having zero or low emissions and a few wells responsible for the majority of emissions. The average CH₄ emission rate from marginal wells was 128 g h^?1 (median: 18 g h^?1; range: 0– 907 g h^?1). Follow-up measurements at five wells indicated high emissions were not episodic. Some wells were emitting all or more of the reported gas produced at each well, or venting gas from wells with no reported gas production. Measurements were made from wellheads only, not tanks, so our estimates may be conservative. Stochastic processes such as maintenance may be the main driver of emissions. Marginal wells are a disproportionate source of CH₄ and VOCs relative to oil and gas production. We estimate that oil and gas wells in this lowest production category emit approximately 11% of total annual CH₄ from oil and gas production in the EPA greenhouse gas inventory, although they produce about 0.2% of oil and 0.4% of gas in the US per year.

Implications: Low producing marginal wells are the most abundant type of oil and gas well in the United States, and a surprising number of them are venting all or more of their reported produced gas to the atmosphere. This makes marginal wells a disproportionate greenhouse gas emissions source compared to their energy return, and a good target for environmental mitigation.     

Optimized variable source-profile approach for source apportionment

An expanded chemical mass balance (CMB) approach for PM_2.5 source apportionment is presented in which both the local source compositions and corresponding contributions are determined from ambient measurements and initial estimates of source compositions using a global-optimization mechanism. Such an approach can serve as an alternative to using predetermined (measured) source profiles, as traditionally used in CMB applications, which are not always representative of the region and/or time period of interest. Constraints based on ranges of typical source profiles are used to ensure that the compositions identified are representative of sources and are less ambiguous than the factors/sources identified by typical factor analysis (FA) techniques. Gas-phase data (SO₂, CO and NO_y) are also used, as these data can assist in identifying sources. Impacts of identified sources are then quantified by minimizing the weighted-error between apportioned and measured levels of the fitting species. This technique was applied to a dataset of PM_2.5 measurements at the former Atlanta Supersite (Jefferson Street site), to apportion PM_2.5 mass into nine source categories. Good agreement is found when these source impacts are compared with those derived based on measured source profiles as well as those derived using a current FA technique, Positive Matrix Factorization. The proposed method can be used to assess the representativeness of measured source-profiles and to help identify those profiles that may be in significant error, as well as to quantify uncertainties in source-impact estimates, due in part to uncertainties in source compositions.     

Mesoscale modeling of wintertime particulate matter episodes in eastern Washington, USA

The Spokane, Washington area is classified as a non-attainment area for the 24-h PM₁₀ standard due to a history of high particulate matter concentrations. A Eulerian regional air quality model (CALMET/CALGRID) has been used to characterize the emission, transport and dispersion of PM₁₀ and PM_2.5 in Spokane. Observations from a residential site (Rockwood, RW) and an industrial site (Crown Zellerbach, CZ), spanning July 1994–August 1996 were used to evaluate the current emission inventory. Two major tasks were devised to conduct the objectives of this investigation. First, a simple and efficient urban dispersion model (WYNDValley) was used to simulate important episodes characterized by the highest PM₁₀ and PM_2.5 concentrations. The selected episodes included four days with wet conditions for which no roads would have been emitting and seven days with dry conditions for which roads would emit. In the second step, a single road-emitting event was selected from the previous predicted results for further analysis using the Eulerian regional air quality model to examine the emission inventory. The urban and regional models predicted the observed concentration distributions reasonably well for the source emissions inventoried in Spokane. The mass concentrations of PM₁₀ were well predicted for the roads emitting case examined by both models indicating that the emission inventory based primarily upon area sources including roads is reasonably well characterized, at least at the RW site. The area sources around CZ are less well characterized, so that the PM₁₀ concentrations are underpredicted at CZ. The models appear unable to reach an equilibrium mass balance status at the beginning of the simulation, and the urban model seems unable to properly resolve the nocturnal boundary layer.     

Wet deposition mercury fluxes in the Canadian sub-Arctic and southern Alberta,measured using an automated precipitation collector adapted to cold regions

This paper reports mercury (Hg) concentrations and fluxes in precipitation that was collected from 2006 to 2008 at three sites in Canada: sub-Arctic boreal forest, sub-Arctic coast, and southern Alberta, using cold-adapted precipitation collectors which operated reliably at temperatures below ?30 °C during the study. The southern Alberta site (Crossfield) may be influenced by Calgary urban air, whereas the sub-Arctic coastal (Churchill, Manitoba) and boreal forest (Fort Vermilion, Alberta) sites are in more remote northern areas. Annual mean Hg concentrations in precipitation (5.0–9.2 ng L^?1) at the study sites were in the lower half of the range reported for southern Canada and the USA by the Mercury Deposition Network (MDN). But owing to typically low precipitation rates, gross wet Hg fluxes (0.54–2.0 μg m^?2 yr^?1) were among the lowest reported by MDN, with Crossfield having about twice the flux in 2007 of the other two sites. Flux was significantly correlated with precipitation, and thus was highest in summer (June–August) and lowest during winter, a pattern typical of other temperate continental locations. There was no evidence of higher wet Hg fluxes or concentrations in springtime at Churchill where atmospheric mercury depletion events (AMDEs) occur. Measured gross deposition fluxes at the study locations were ～2–8 times lower than estimated by GEOS-Chem and GRAHM atmospheric models. The largest discrepancy occurred for Churchill, which raises the question of how well Hg deposition from AMDEs is described by current models. Better agreement between measurements and models was obtained from MDN stations in Alberta and Alaska, where wet Hg fluxes were 2–10 times higher than the study sites either because of power plant emissions (Alberta), or because of high precipitation rates (Alaska).     

超声波／零价铁降解水中双酚A

研究了超声波／零价铁（US／Fe^0）工艺对水中双酚A（BPA）的降解效果及影响因素。结果表明,US／Fe^0工艺可以有效降解水中的BPA,具有协同作用,且降解过程符合一级动力学规律;Fe^0具有促进和抑制的双重作用;超声功率越大,越有利于BPA的降解;初始浓度较低时,BPA的降解效果较好;弱酸性条件有利于BPA的降解;加入一定量的自由基捕获剂（正丁醇）可以抑制BPA的降解;联合作用2h后,TOC的去除率较低,说明其矿化程度不完全。     

Measurement and modeling of urban atmospheric PCB concentrations on a small (8 km) spatial scale

Atmospheric transport and deposition of polychlorinated biphenyls (PCBs) is an important problem for ecosystems around the world. Data from several monitoring networks demonstrate that atmospheric PCB concentrations are dramatically elevated in urban areas compared to rural or background regions, such that these urban emissions of PCBs support the regional and global transport and deposition of PCBs to more remote areas. Identifying and controlling the sources of urban atmospheric PCBs is thus essential in minimizing the regional and global transport and deposition of these compounds. From December 1999 to November 2000, gas-phase PCB concentrations were measured at two monitoring locations, 8 km apart, within the New York City metropolitan area, at Jersey City and Bayonne, NJ. Concentrations, congener patterns, and temporal patterns of PCBs differ dramatically at the two sites, suggesting that a significant source of atmospheric PCBs exists within 8 km of the Bayonne site, resulting in spikes in gas-phase PCB concentration at Bayonne that are not observed at Jersey City. The Regional Atmospheric Model System (RAMS) coupled with the Hybrid Particle and Concentration Transport model (HYPACT) was used to estimate that the PCB source near Bayonne emits a flux of ΣPCBs on the order of 100 g d⁻¹. Extrapolation of this source magnitude to the area of New York City suggests that this urban area emits at least 300 kg yr⁻¹ ΣPCBs to the regional atmosphere, similar in magnitude to the flow of ΣPCB out of the Upper Hudson River into the New York/New Jersey Harbor.     

Development of a modified factor analysis/multiple regression model to apportion suspended particulate matter in a complex urban airshed

A modified factor analysis/multiple regression (FA/MR) receptor-oriented source apportionment model has been developed which permits application of FA/MR statistical methods when some of the tracers are not unique to an individual source type. The new method uses factor and regression analyses to apportion non-unique tracer ambient concentrations in situations where there are unique tracers for all sources contributing to the non-unique tracer except one, and ascribes the residual concentration to that source. This value is then used as the source tracer in the final FA/MR apportionment model for ambient paniculate matter. In addition, factor analyses results are complemented with examination of regression residuals in order to optimize the number of identifiable sources.The new method has been applied to identify and apportion the sources of inhalable particulate matter (IPM; D₅₀15 μm), Pb and Fe at a site in Newark, NJ. The model indicated that sulfate/secondary aerosol contributed an average of 25.8 μ⁻³ (48%) to IPM concentrations, followed by soil resuspension (8.2 μ⁻³ or 15%), paint spraying/paint pigment (6.7/gmm⁻³or 13%), fuel oil burning/space heating (4.3 μ⁻³ or 8 %), industrial emissions (3.6 μm⁻³ or 7 %) and motor vehicle exhaust (2.7 μ⁻³ or 15 %). Contributions to ambient Pb concentrations were: motor vehicle exhaust (0.16μm⁻³or 36%), soil resuspension (0.10μm⁻³ or 24%), fuel oil burning/space heating (0.08μm⁻³or 18%), industrial emissions (0.07 μ⁻³ or 17 %), paint spraying/paint pigment (0.036 μm⁻³or 9 %) and zinc related sources (0.022 μ⁻³ or 5 %). Contributions to ambient Fe concentrations were: soil resuspension (0.43μ⁻³or 51%), paint spraying/paint pigment (0.28 μm⁻³or 33 %) and industrial emissions (0.15 μ⁻³or 18 %). The models were validated by comparing partial source profiles calculated from modeling results with the corresponding published source emissions composition.     

Irradiation of sorbents by ions of polymorphic metals for modeling <Superscript>90</Superscript>strontium sedimentation

Background, Aims and Scope Advances in radioecology can support improvements in environmental remediation technologies, especially by illuminating interaction processes between polymorphic metal radionuclides and various materials and their ions in aqueous solutions. This study modeled interaction processes of ⁹⁰Sr with transitive metals to delineate the behavior of polymorphic metal radionuclides. Experimental and modeling results confirmed Sr sedimentation was sensitive to the physical impact of radionuclides on various sorbents and possible chemical reactions occurring between the radionuclides and sorbents. Methods Models were developed to simulate ⁹⁰Sr sedimentation process, and the potential physical and chemical reactions accompanying the process. Models were verified, inorganic salts were used as sorbents to absorb metal cations, activity levels were recorded before and after mixing the inorganic salts while the efficiency of sedimentation using the heavy metals composites was quantified. Results and Discussion This research demonstrates that the process of the sedimentation is complex and occurs in several stages. Micro-structural analysis shows that zones of interaction between the sorbent and source metal are formed during the irradiation of the target’s metal surface. Electrical-microscopic analysis indicates that the composition of the formed zones of interaction of Ti (Sr) with target metals has various structures. Roentgenophase analysis indicates that the interaction of the ions of a precipitable source and a target occurs according to constitution diagrams of equilibrium systems. The results indicate that application of inorganic salts composites based on modeling increases the efficiency of the deactivation of aqueous solutions when compared to standard aluminum sulfate composite. Conclusions Experimental and modeling results confirm ⁹⁰Sr sedimentation is sensitive to the physical impact of radionuclides on various sorbents and possible chemical reactions occurring between the radionuclides and sorbents. The models support estimation of the physical impact of polymorphic metal radionuclides on various components of sorbents and possible chemical reactions occurring between the radionuclides and sorbents during the interaction. Inorganic salt composites deactivate and clear ⁹⁰Sr and Sr⁺² from water. This paper is openly accessible!     

Ultrasound enhanced activation of peroxydisulfate by activated carbon fiber for decolorization of azo dye

Activated carbon fiber (ACF) has become an emerging activator for peroxydisulfate (PDS) to generate sulfate radical (SO₄^??). However, the relative low activation efficiency and poor contaminant mineralization limited its widespread application. Herein, ultrasound (US) was introduced to the ACF activated PDS system, and the synergistic effect of US and ACF in PDS activation and the enhancement of contaminant mineralization were investigated. The synergistic effect of US and ACF was observed in the PDS activation to decolorize orange G (OG). The decolorization efficiency increased with increasing ACF loading and US power, and PDS/OG ratio from 1 to 40. The activation energy was determined to be 24.065 kJ/mol. The radical-induced decolorization of OG took place on the surface of ACF, and both SO₄^?? and hydroxyl radical (^?OH) contributed to OG decolorization. The azo bond and naphthalene ring on OG were destructed to other aromatic intermediates and finally mineralized to CO₂ and H₂O. The introduction of US in the ACF/PDS system significantly enhanced the mineralization of OG. The combination of US and PDS was highly efficient to activate PDS to decolorize azo dyes. Moreover, the introduction of US remarkably improved the contaminant mineralization.     

Blood mercury levels among fish consumers residing in areas with high environmental burden

Mercury is a ubiquitous, persistent toxicant found in the environment. In water, mercury bioaccumulates up the food chain and leads to high concentrations in fish. Consumption of contaminated fish is the major source of exposure to mercury in the US. The objective of this study was to enroll persons living in areas selected by the Environmental Protection Agency (EPA) to have high mercury concentrations and who consume at least 6 oz of locally caught fish per week to determine the feasibility of monitoring future trends among a population identified as highly exposed. Blood samples were collected at time of interview and analyzed for mercury. Participants (n = 287) were enrolled from North Carolina, Maryland, and South Dakota. Participants reported eating an average of five servings of fish per week. The overall geometric mean for total mercury was 0.75 μg L⁻¹, with North Carolina having the highest mean level (2.02 μg L⁻¹). Overall, 42% of the study population had levels greater than the US geometric mean 0.83 μg L⁻¹. The number of servings of fish consumed was not found to be associated with blood mercury levels. We were able to identify some persons with elevated mercury concentrations living in areas identified by EPA; however, identifying and monitoring a highly exposed population over time would be challenging.