Advancing exposure assessment approaches to improve wildlife risk assessment

The exposure assessment component of a Wildlife Ecological Risk Assessment aims to estimate the magnitude, frequency, and duration of exposure to a chemical or environmental contaminant, along with characteristics of the exposed population. This can be challenging in wildlife as there is often high uncertainty and error caused by broad-based, interspecific extrapolation and assumptions often because of a lack of data. Both the US Environmental Protection Agency (USEPA) and European Food Safety Authority (EFSA) have broadly directed exposure assessments to include estimates of the quantity (dose or concentration), frequency, and duration of exposure to a contaminant of interest while considering "all relevant factors." This ambiguity in the inclusion or exclusion of specific factors (e.g., individual and species-specific biology, diet, or proportion time in treated or contaminated area) can significantly influence the overall risk characterization. In this review, we identify four discrete categories of complexity that should be considered in an exposure assessment-chemical, environmental, organismal, and ecological. These may require more data, but a degree of inclusion at all stages of the risk assessment is critical to moving beyond screening-level methods that have a high degree of uncertainty and suffer from conservatism and a lack of realism. We demonstrate that there are many existing and emerging scientific tools and cross-cutting solutions for tackling exposure complexity. To foster greater application of these methods in wildlife exposure assessments, we present a new framework for risk assessors to construct an "exposure matrix." Using three case studies, we illustrate how the matrix can better inform, integrate, and more transparently communicate the important elements of complexity and realism in exposure assessments for wildlife. Modernizing wildlife exposure assessments is long overdue and will require improved collaboration, data sharing, application of standardized exposure scenarios, better communication of assumptions and uncertainty, and postregulatory tracking. Integr Environ Assess Manag 2024;20:674-698. © 2023 SETAC.

Knowledge-guided mixture density network for chlorophyll-a retrieval and associated pixel-by-pixel uncertainty assessment in optically variable inland waters

Machine learning has been increasingly used to retrieve chlorophyll-a (Chl-a) in optically variable waters. However, without the guidance of physical principles or expert knowledge, machine learning may produce biased mapping relationships, or waste considerable time searching for physically infeasible hyperparameter domains. In addition, most Chl-a retrieval models cannot evaluate retrieval uncertainty when ground observations are not available, and the retrieval uncertainty is crucial for understanding the model limitations and evaluating the reliability of retrieval results. In this study, we developed a novel knowledge-guided mixture density network to retrieve Chl-a in optically variable inland waters based on Sentinel-3 Ocean and Land Color Instrument (OLCI) imagery. The proposed method embedded prior knowledge derived from spectral shape classification into the mixture density network. Compared to another deterministic model, the knowledge-guided mixture density network outputted the conditional distribution of Chl-a given an input spectrum, enabling us to estimate the optimal retrieval and the associated uncertainty. The proposed method showed favorable correspondence with the field Chl-a, with root mean square error (RMSE) of 6.56 μg/L, and mean absolute percentage error (MAPE) of 43.64 %. Calibrated against Sentinel-3 OLCI spectrum, the proposed method also performed well when applied to field spectrum (RMSE = 4.58 μg/L, MAPE = 72.70 %), suggesting its effectiveness and good generalization. The proposed method provided the standard deviation of each estimated Chl-a, which enabled us to inspect the reliability of the estimated results and understand the model limitations. Overall, the proposed method improved the Chl-a retrieval in terms of model accuracy and uncertainty evaluation, providing a more comprehensive Chl-a observation of inland waters.

Mapping specific groundwater nitrate concentrations from spatial data using machine learning: A case study of chongqing, China

Groundwater resources is not only important essential water resources but also imperative connectors within the intricate framework of the ecological environment. High nitrate concentrations in groundwater can exerting adverse impacts on human health. It is imperative to accurately delineate the distribution characteristics of groundwater nitrate concentrations. Four different machine learning models (Gradient Boosting Regression (GB), Random Forest Regression (RF), Extreme Gradient Boosting Regression (XG) and Adaptive Boosting Regression (AD)) which combine spatial environmental data and different radius contributing area was developed to predict the distribution of nitrate concentration in groundwater. The models use 595 groundwater samples and included topography, remote sensing, hydrogeological and hydrological, climate, nitrate input, and socio-economic predictor. Gradient Boosting Regression model outperforms the other models (R2 = 0.627, MAE = 0.529, RMSE = 0.705, PICP = 0.924 for test dataset) under 500 m radius contributing area. A high-resolution (1 km) groundwater nitrate concentration distribution map reveal in the majority of the study area, groundwater nitrate concentrations are below 1 mg/L and high nitrate concentration (>10 mg/L) proportion in southeast, northeast and central main urban area karst valley regions is 1.89%, 0.91%, and 0.38% respectively. In study area, hydrogeological conditions, soil parameters, nitrogen input factors, and percentage of arable land are among the most influential explanatory factors. This work, serving as the inaugural application of utilizing effective spatial methods for predicting groundwater nitrate concentrations in Chongqing city, furnish decision-making support for the prevention and control of groundwater pollution, particularly in areas primarily dependent on groundwater for water supply and holds profound significance as a milestone achievement.

An uncertainty approach for Electric Submersible Pump modeling through Deep Neural Network

This work proposes a new methodology to identify and validate deep learning models for artificial oil lift systems that use submersible electric pumps. The proposed methodology allows for obtaining the models and evaluating the prediction's uncertainty jointly and systematically. The methodology employs a nonlinear model to generate training and validation data and the Markov Chain Monte Carlo algorithm to assess the neural network's epistemic uncertainty. The nonlinear model was used to overcome the limitations of the need for big datasets for training deep learning models. However, the developed models are validated against experimental data after training and validation with synthetic data. The validation is also performed through the models' uncertainty assessment and experimental data. From the implementation point of view, the method was coded in Python with Tensorflow and Keras libraries used to build the neural Networks and find the hyperparameters. The results show that the proposed methodology obtained models representing both the nonlinear model's dynamic behavior and the experimental data. It provides a most probable value close to the experimental data, and the uncertainty of the generated deep learning models has the same order of magnitude as that of the nonlinear model. This uncertainty assessment shows that the built models were adequately validated. The proposed deep learning models can be applied in several applications requiring a reliable and computationally lighter model. Hence, the obtained AI dynamic models can be employed for digital twin construction, control, and optimization.

Mixing apples and oranges: Assessing ecological status and its confidence from multiple and diverse indicators

Ecosystem responses to increasing human pressures are complex and diverse, affecting organisms across all trophic levels. This has prompted the development of methods that integrate information across many indicators for environmental management. Legislative frameworks such as the European Water Framework Directive (WFD), specifically prescribe that integrated assessme nt (IA) of ecological status must consider indicators representing various biological and supporting quality elements. We present a general approach for an IA system based on a piece-wise linear transformation of indicator distributions to a standardized scale, allowing for integrating information from multiple and diverse indicators through a policy-dependent aggregation scheme. Uncertainties associated with monitoring data used for calculating indicators and their propagation throughout the integration scheme allow for confidence assessment at all levels of the hierarchical integration. Specific pressures leading to ecological impact can be identified through the most impaired indicators in the hierarchical and transparent aggregation scheme. The IA and its confidence are facilitated though the development of an online tool that accesses information from monitoring databases and presents the outcome at all levels of the assessment, ensuring consistency and transparency in the calculations for all potential stakeholders. We demonstrate the versality and applicability of the approach using indicators and aggregation principles from the Swedish national guidelines for assessing ecological status of rivers, lakes and coastal waters according to the WFD. Although the approach and the tool were developed specifically for the WFD ecological status assessment in Sweden, the generality of the approach implies that it can easily be adapted to the WFD assessment methods of other countries as well as other policies, where an integrated assessment is required.

The fusion of multiple scale data indicates that the carbon sink function of the Qinghai-Tibet Plateau is substantial

BACKGROUND

The Qinghai-Tibet Plateau is the "sensitive area" of climate change, and also the "driver" and "amplifier" of global change. The response and feedback of its carbon dynamics to climate change will significantly affect the content of greenhouse gases in the atmosphere. However, due to the unique geographical environment characteristics of the Qinghai-Tibet Plateau, there is still much controversy about its carbon source and sink estimation results. This study designed a new algorithm based on machine learning to improve the accuracy of carbon source and sink estimation by integrating multiple scale carbon input (net primary productivity, NPP) and output (soil heterotrophic respiration, Rh) information from remote sensing and ground observations. Then, we compared spatial patterns of NPP and Rh derived from the fusion of multiple scale data with other widely used products and tried to quantify the differences and uncertainties of carbon sink simulation at a regional scale.

RESULTS

Our results indicate that although global warming has potentially increased the Rh of the Qinghai-Tibet Plateau, it will also increase its NPP, and its current performance is a net carbon sink area (carbon sink amount is 22.3 Tg C/year). Comparative analysis with other data products shows that CASA, GLOPEM, and MODIS products based on remote sensing underestimate the carbon input of the Qinghai-Tibet Plateau (30-70%), which is the main reason for the severe underestimation of the carbon sink level of the Qinghai-Tibet Plateau (even considered as a carbon source).

CONCLUSIONS

The estimation of the carbon sink in the Qinghai-Tibet Plateau is of great significance for ensuring its ecological barrier function. It can deepen the community's understanding of the response to climate change in sensitive areas of the plateau. This study can provide an essential basis for assessing the uncertainty of carbon sources and sinks in the Qinghai-Tibet Plateau, and also provide a scientific reference for helping China achieve "carbon neutrality" by 2060.

Uncertainty assessment for three-dimensional hydrodynamic and fecal coliform modeling in the Danshuei River estuarine system: The influence of first-order parametric decay reaction

Modeling fecal contamination in water bodies is of importance for microbiological risk assessment and management. This study investigated the transport of fecal coliform (e.g., up to 2.1 × 10⁶ CFU/100 ml at the Zhongshan Bridge due to the main point source from the Xinhai Bridge) in the Danshuei River estuarine system, Taiwan with the main focus on assessing model uncertainty due to three relevant parameters for the microbial decay process. First, a 3D hydrodynamic-fecal coliform model (i.e., SCHISM-FC) was developed and rigorously validated against the available data of water level, velocity, salinity, suspended sediment and fecal coliform measured in 2019. Subsequently, the variation ranges of decay reaction parameters were considered from several previous studies and properly determined using the Monte Carlo simulations. Our analysis showed that the constant ratio of solar radiation (α) as well as the settling velocity (v_s) had the normally-distributed variations while the attachment fraction of fecal coliform bacteria (F_p) was best fitted by the Weibull distribution. The modeled fecal coliform concentrations near the upstream (or downstream) stations were less sensitive to those parameter variations (see the smallest width of confidence interval about 1660 CFU/100 ml at the Zhongzheng Bridge station) due to the dominant effects of inflow discharge (or tides). On the other hand, for the middle parts of Danshuei River where complicated hydrodynamic circulation and decay reaction occurred, the variations of parameters led to much larger uncertainty in modeled fecal coliform concentration (see a wider confidence interval about 117,000 CFU/100 ml at the Bailing Bridge station). Overall, more detailed information revealed in this study would be helpful while the environmental authority needs to develop a proper strategy for water quality assessment and management. Owing to the uncertain decay parameters, for instance, the modeled fecal coliform impacts at Bailing Bridge over the study period showed a 25 % difference between the lowest and highest concentrations at several moments. For the detection of pollution occurrence, the highest to lowest probabilities for a required fecal coliform concentration (e.g., 260,000 CFU/100 ml over the environmental regulation) at Bailing Bridge was possibly greater than three.

Experimental uncertainty assessment of meso- and microplastic concentrations in rivers based on net sampling

Meso- and microplastics have been collected via net sampling in marine and freshwater environments, but the effect of net clogging on evaluations of their concentrations (mPC) remains uncertain. We experimentally investigated the mPC uncertainties resulting from net clogging in the Ohori and Tone-unga Rivers, typical urban rivers in Japan, throughout 16 samplings with five filtration durations in one day. The weighted mean concentration in the Ohori River was significantly lower than that in the Tone-unga River, allowing us to examine the effect of clogging in rivers with different contamination levels. The variances in both rivers consistently tended to increase with increasing filtration duration, which can be expressed by applying the integral form of the Weibull reliability function (WRF). Furthermore, application of the WRF successfully revealed the optimal filtration durations in the Ohori and Tone-unga Rivers, which depended on the plastic abundance and sample volume. Since it could be difficult to obtain the plastic contamination level in advance, our suggestion is to predict the time sustained above 85 % filtration efficiency by applying a WRF-based model. In actuality, the sustained time in the Ohori (Tone-unga) River varied between 2.6 and 6.2 min (3.2 and 7.1 min) throughout the experiment, which permitted low mPC uncertainties of 12 % and 9.5 %, respectively. If notable uncertainty exists due to a low contamination level, a net with a high open area ratio should be used to increase the filtration duration. Hence, our results emphasize the importance of considering the open area ratio of nets used for sampling in studies. Our study provides insights into the occurrence of uncertainty due to net clogging to establish a standardized methodology for meso- and microplastic monitoring in aquatic environments via net sampling and consequently contributes to improving the sampling accuracy.

Prediction of the concentration of cadmium in agricultural soil in the Czech Republic using legacy data, preferential sampling, Sentinel-2, Landsat-8, and ensemble models

The current study assesses and predicts cadmium (Cd) concentration in agricultural soil using two Cd datasets, namely legacy data (LD) and preferential sampling-legacy data (PS-LD), along with four streams of auxiliary datasets extracted from Sentinel-2 (S2) and Landsat-8 (L8) bands. The study was divided into two contexts: Cd prediction in agricultural soil using LD, ensemble models, 10 and 20 m spatial resolution of S2 and L8 (context 1), and Cd prediction in agricultural soil using PS-LD, ensemble models and 10 and 20 m spatial resolution of S2 and L8 (context 2). In context 1, ensemble 1, L8 with PS-LD was the cumulative optimal approach that predicted Cd in agricultural soil with a higher R² value of 0.76, root mean square error (RMSE) of 0.66, mean absolute error (MAE) of 0.35, and median absolute error (MdAE) of 0.13. However, with R² = 0.78, RMSE = 0.63, MAE = 0.34, and MdAE = 0.15, ensemble 1, S2 of PS-LD was the best prediction approach in predicting Cd concentration in agricultural soil in context 2. Overall, the predictions from both contexts indicated that ensemble 1 of S2 combined with PS-LD was the most appropriate and best model for Cd prediction in agricultural soil. The modeling approaches' uncertainty in both contexts was assessed using ensemble-sequential gaussian simulation (EnSGS), which revealed that the degree of uncertainty propagated in the study area was within 5% in both contexts. The combination of the PS dataset and the LD along with ensemble models and the remote sensing dataset, produced promising results. Nonetheless, the results demonstrated that the 20 m spatial resolution band dataset used in the prediction of Cd in agricultural soil outperformed the 10 m spatial resolution. When PS is combined with LD, an appropriate modeling approach, and a well-correlated remote sensing dataset are used, good results are obtained.

Optimal zinc level and uncertainty quantification in agricultural soils via visible near-infrared reflectance and soil chemical properties

Zinc (Zn) is a vital element required by all living creatures for optimal health and ecosystem functioning. Therefore, several researchers have modeled and mapped its occurrence and distribution in soils. Nonetheless, leveraging model predictive performances while coupling information derived from visible near-infrared (Vis-NIR) and soils (i.e. chemical properties) to estimate potential toxic elements (PTEs) like Zn in agricultural soils is largely untapped. This study applies two methods to rapidly monitor Zn concentration in agricultural soil. Firstly, employing Vis-NIR and machine learning algorithms (MLAs) (Context 1) and secondly, applying Vis-NIR, soil chemical properties (SCP), and MLAs (Context 2). For the Vis-NIR information, single and combined pretreatment methods were applied. The following MLAs were used: conditional inference forest (CIF), partial least squares regression (PLSR), M5 tree model (M5), extreme gradient boosting (EGB), and support vector machine regression (SVMR) respectively. For context 1, the results indicated that M5-MSC (M5 tree model-multiplicative scatter correction) with coefficient of determination (R²) = 0.72, root mean square error (RMSE) = 21.08 (mg/kg), median absolute error (MdAE) = 13.69 and ratio of performance to interquartile range (RPIQ) = 1.63 was promising. Regarding context 2, CIF with spectral pretreatment and soil properties [CIF-DWTLOGMSC + SCP (conditional inference forest-discrete wavelet transformation-logarithmic transformation-multiplicative scatter correction-soil chemical properties)] yielded the best performance of R² = 0.86, RMSE = 14.52 (mg/kg), MdAE = 6.25 and RPIQ = 1.78. Altogether, for contexts 1 and 2, the CIF-DWTLOGMSC + SCP approach (context 2) was the best Zn model outcome for the agricultural soil. The uncertainty map revealed a low to high error distribution in context 1, and a low to moderate distribution in context 2 for all models except CIF, which had some patches with high uncertainty. We conclude that a multiple optimization approach for modeling Zn levels in agricultural soils is invaluable and may provide fast and reliable information needed for area-specific decision-making.

An approach to regionalise the life cycle inventories of Spanish agriculture: Monitoring the environmental impacts of orange and tomato crops

Agricultural life cycle assessment (LCA) at the sub-national regional level may be a valuable input for the decision-makers. Obtaining representative and sufficient data to develop life cycle inventories (LCIs) at that level is a relevant challenge. This study aims to contribute to the development of LCIs representative Spanish crops based on economic and operational information available in official sources to assess the average environmental impacts of these crops in the main producing regions. A comprehensive approach is proposed considering both the temporal variability and uncertainty of input data by using different methods (e.g. linear programming, weighted averages, Monte Carlo simulation, forecasted irrigation, etc.) to estimate the inventory data of reference holdings. From these inventories, the environmental assessment of those reference holdings is carried out. Two case studies are developed, on orange and tomato crops in the main producing regions, where climate change (CC), freshwater scarcity (WS), human toxicity non-cancer (HTnc), and freshwater ecotoxicity (ET) are evaluated. The environmental scores obtained differ significantly from region to region. The highest environmental scores of orange reference holdings correspond to Comunidad Valenciana for CC (1.94·10^-1 kg CO₂ eq.) HTnc (4.16·10^-11 CTUh) and ET (7.45·10^-3 CTUe), and to Andalucia in WS (17.4 m³ world eq.). As to greenhouse tomatoes, the highest scores correspond to Comunidad Valenciana in the four categories analysed (CC = 3.18 kg CO₂ eq., HTnc = 3.6·10^-9 CTUh, ET = 1.5 CTUe and WS = 13.3 m³ world eq.). The environmental scores estimated in this study are consistent with the literature, showing that the approach is useful to obtain a representative description of the environmental profile of crops from official statistical data and other information sources. Widening the data gathered in ECREA-FADN, and also that from other data sources used, would increase the quality of the environmental impact estimation.

Assessing spatial connectivity effects on daily streamflow forecasting using Bayesian-based graph neural network

Data-driven models have been widely developed and achieved impressive results in streamflow prediction. However, the existing data-driven models mostly focus on the selection of input features and the adjustment of model structure, and less on the impact of spatial connectivity on daily streamflow prediction. In this paper, a basin network based on graph-structured data is constructed by considering the spatial connectivity of different stations in the real basin. Furthermore, a novel graph neural network model, variational Bayesian edge-conditioned graph convolution model, which consists of edge-conditioned convolution networks and variational Bayesian inference, is proposed to assess the spatial connectivity effects on daily streamflow forecasting. The proposed graph neural network model is applied to forecast the next-day streamflow of a hydrological station in the Yangtze River Basin, China. Six comparative models and three comparative experimental groups are used to validate model performance. The results show that the proposed model has excellent performance in terms of deterministic prediction accuracy (NSE ≈ 0.980, RMSE≈1362.7 and MAE ≈ 745.8) and probabilistic prediction reliability (ICPC≈0.984 and CRPS≈574.1), which demonstrates that establishing appropriate connectivity and reasonably identifying connection relationships in the basin network can effectively improve the deterministic and probabilistic forecasting performance of the graph convolutional model.

Evaluation of uncertain signals' impact on deep reinforcement learning-based real-time control strategy of urban drainage systems

Real-time control (RTC) is a recognized technology to enhance the efficiency of urban drainage systems (UDS). Deep reinforcement learning (DRL) has recently provided a new solution for RTC. However, the practice of DRL-based RTC has been impeded by different sources of uncertainties. The present study aimed to evaluate the impact caused by the uncertainties on DRL-based RTC to promote its application. The impact of uncertainties in the measurement of water level signals was evaluated through large-scale simulation experiments and quantified using measures of statistical dispersion of control performance distribution and relative change of control performance compared to the baseline scenario with no uncertainty. Results show that the statistical dispersion of DRL-based RTC was reduced by 15.48%-81.93% concerning random and systematic uncertainties compared to the conventional rule-based control (RBC) strategy. The findings indicated that DRL-based RTC is robust and could be reliably applied to safety-critical real-world UDS.

Semi-automated harmonization and selection of chemical data for risk and impact assessment

Chemical data for thousands of substances are available for safety, risk, life cycle and substitution assessments, as submitted for example under the European Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) Regulation. However, to widely disseminate reported physicochemical properties as well as human and ecological exposure and toxicological data for use in various science and policy fields, systematic methods for data harmonization and selection are necessary. In response to this need, we developed a semi-automated method for deriving appropriate substance property values as input for various assessment frameworks with different requirements for resolution and data quality. Starting with data reported for a given substance and property, we propose a set of aligned data selection and harmonization criteria to obtain a representative mean value and related confidence intervals per chemical-property combination. The proposed method was tested on a set of octanol-water partition coefficients (K_ow) for an illustrative set of 20 substances, reported under the REACH regulation as example data source. Our method is generally applicable to any set of substances, and can assess specific distributions in quality and variability across reported data. Further research can likely extend our method for mining information from text fields and adapt it to available data reported or collected from other sources and other substance properties to improve the reliability of input data for risk and impact assessments.

Discussion on the existing methodology of entropy-weights in water quality indexing and proposal for a modification of the expected conflicts

The present research focuses on addressing various ambiguities in the existing method of integrating information entropy and water quality, thereby presenting a novel approach for an entropy-weighted water quality index. A three-dimensional water quality dataset is considered in the proposed method, the third dimension being the sampling frequency factor. The probability of observed values adhering to desirable limits prescribed by a standard code is estimated, leading to the computation of information entropy and, eventually, entropy weights. These weights are then used for the computation of the Modified Entropy-weight Water Quality Index (MEWQI) values. To verify the proposed method's applicability, the water quality dataset of Deepor Beel, India, was considered. IS 10500: 2012 was used for estimating MEWQI values. Results showed an excellent correlation with the observed dataset and their uncertainties of occurrence. The reliability and correctness of the proposed methodology were finally confirmed through both cluster analysis and sensitivity analysis. The cluster analysis showed remarkable associations with the computed MEWQI values, while the sensitivity analysis proved that no particular parameter was accountable for the contribution of MEWQI values; instead, all parameters exhibited equal contributions. The proposed methodology was thus found to be the most reasonable and reliable as it considered both factors, i.e., measured values concerning standard limits and the uncertainty, necessary for a consistent water quality monitoring program.

A novel approach for estimation of aboveground biomass of a carbon-rich mangrove site in India

Mangroves can play a crucial part in climate change mitigation policies due to their high carbon-storing capacity. However, the carbon sequestration potential of Indian mangroves generally remained unexplored to date. In this study, multi-temporal Sentinel-1 and 2 data-derived variables were used to estimate the AGB of a tropical carbon-rich mangrove forest of India. Ensemble prediction of multiple machine learning algorithms, including Random Forest (RF), Gradient Boosted Model (GBM), and Extreme Gradient Boosting (XGB), were used for AGB prediction. The multi-temporal dataset was used in two different ways to find the most suitable method of using them. The results of the analysis showed that the modeling field measured AGB with individual date data values results in estimates with root mean square errors (RMSE) ranging from 149.242 t/ha for XGB to 151.149 t/ha for the RF. Modeling AGB with the average and percentile metrics of the multi-temporal image stack improves the prediction accuracy of AGB, with RMSE ranging from 81.882 t/ha for the XGB to 74.493 t/ha for the RF. The AGB modeling using ensemble prediction showed further improvement in accuracy with an RMSE of 72.864 t/ha and normalized RMSE of 11.38%. In this study, the intra-seasonal variation of Sentinel-1 and 2 data for mangrove ecosystems was explored for the first time. The variations in remotely sensed variables could be attributed mainly to soil moisture availability and rainfall in the mangrove ecosystem. The efficiency of Sentinel-1 and 2 data-derived variables and ensemble prediction of machine learning models for Indian mangroves were also explored for the first time. The methodologies established in this study can be used in the future for accurate prediction and repeated monitoring of AGB for mangrove ecosystems.

Polycyclic aromatic hydrocarbons, pesticides, and metals in olive: analysis and probabilistic risk assessment

In the present study, levels of 22 pesticides, eight metals, and 16 polycyclic aromatic hydrocarbons (PAHs) in 1800 Iranian olive samples (20 cultivars from six different cultivation zones), were determined; then, health risk posed by oral consumption of the olive samples to Iranian consumers was assessed. Quantification of PAHs and pesticides was done by chromatography-mass spectrometry (GC-MS), and metal levels were determined using inductively coupled plasma-optical emission spectrometry (ICP-OES). There were no significant differences among the cultivars and zones in terms of the levels of the tested compounds. Target hazard quotients (THQ) were <1.0 for all pesticides, and total hazard indices (HI) indicated di minimis risk. At the 25th or 95th centiles, Incremental Life Time Cancer Risks (ILCRs) for carcinogenic elements, arsenic, and lead and noncarcinogenic metals did not exhibit a significant hazard (HI <1.0 for both cases). At the 25th or 95th centiles, ILCR and margins of exposure (MoE) for PAHs indicated di minimis risk. Sensitivity analysis showed that concentrations of contaminants had the most significant effect on carcinogenic and noncarcinogenic risks.

Inter-Subject Analysis: A Partial Gaussian Graphical Model Approach

Different from traditional intra-subject analysis, the goal of inter-subject analysis (ISA) is to explore the dependency structure between different subjects with the intra-subject dependency as nuisance. ISA has important applications in neuroscience to study the functional connectivity between brain regions under natural stimuli. We propose a modeling framework for ISA that is based on Gaussian graphical models, under which ISA can be converted to the problem of estimation and inference of a partial Gaussian graphical model. The main statistical challenge is that we do not impose sparsity constraints on the whole precision matrix and we only assume the inter-subject part is sparse. For estimation, we propose to estimate an alternative parameter to get around the nonsparse issue and it can achieve asymptotic consistency even if the intra-subject dependency is dense. For inference, we propose an "untangle and chord" procedure to de-bias our estimator. It is valid without the sparsity assumption on the inverse Hessian of the log-likelihood function. This inferential method is general and can be applied to many other statistical problems, thus it is of independent theoretical interest. Numerical experiments on both simulated and brain imaging data validate our methods and theory. Supplementary materials for this article are available online.

Dual isotope ratio normalization of nitrous oxide by bacterial denitrification of USGS reference materials

We measured the δ values of N₂O using gas chromatography isotope ratio mass spectrometry with a preconcentrator (precon-GC-IRMS). The instrumental precision of the mass spectrometer was restricted to below the shot noise limit, which agreed with the theoretical and experimental results of 0.02‰ (δ¹⁵N) and 0.04‰ (δ¹⁸O), respectively. The precision of the measured δ values was significantly improved by the temperature regulation protocol of the LN₂ preconcentrator, which was monitored by various temperature sensors placed along the U-trap. The reproducibility of the He-diluted N₂O gas measurements resulted in 0.063‰ (δ¹⁵N) and 0.075‰ (δ¹⁸O) due to additional sources of uncertainty in the vials used for autosampling and in the general preconcentration process. Multipoint normalization of the dual δ values of the measured N₂O samples was conducted using United States Geological Survey reference materials denitrified by Pseudomonas aureofaciens. Kaiser's ion correction method, based on International Atomic Energy Agency parameters, exhibited low bias for the atomic isotope ratio reduction of the nitrate reference material, for which the oxygen anomaly was considerably high. Dedicated corrections for net isotope fractionation and water exchange were important in improving uncertainties in the procedure for normalizing the oxygen isotope ratio. Blank measurements for correcting biases in isotope ratios caused by pre-dissolved nitrate and nitrite ions in the water solvent led to further improvements, i.e. beyond unevenly controlled net isotope fractionation, throughout the bacterial denitrification process. The uncertainty evaluation revealed that three-point normalization can significantly improve the normalization accuracy compared with two-point normalization. In addition, an alternative strategy was suggested for assigning δ¹⁸O using a CO₂ lab tank, allowing its use as a reference material for N₂O gas tanks.

Predicting uncertainty of machine learning models for modelling nitrate pollution of groundwater using quantile regression and UNEEC methods

Although estimating the uncertainty of models used for modelling nitrate contamination of groundwater is essential in groundwater management, it has been generally ignored. This issue motivates this research to explore the predictive uncertainty of machine-learning (ML) models in this field of study using two different residuals uncertainty methods: quantile regression (QR) and uncertainty estimation based on local errors and clustering (UNEEC). Prediction-interval coverage probability (PICP), the most important of the statistical measures of uncertainty, was used to evaluate uncertainty. Additionally, three state-of-the-art ML models including support vector machine (SVM), random forest (RF), and k-nearest neighbor (kNN) were selected to spatially model groundwater nitrate concentrations. The models were calibrated with nitrate concentrations from 80 wells (70% of the data) and then validated with nitrate concentrations from 34 wells (30% of the data). Both uncertainty and predictive performance criteria should be considered when comparing and selecting the best model. Results highlight that the kNN model is the best model because not only did it have the lowest uncertainty based on the PICP statistic in both the QR (0.94) and the UNEEC (in all clusters, 0.85-0.91) methods, but it also had predictive performance statistics (RMSE = 10.63, R² = 0.71) that were relatively similar to RF (RMSE = 10.41, R² = 0.72) and higher than SVM (RMSE = 13.28, R² = 0.58). Determining the uncertainty of ML models used for spatially modelling groundwater-nitrate pollution enables managers to achieve better risk-based decision making and consequently increases the reliability and credibility of groundwater-nitrate predictions.

Multipoint normalization of δ18O of water against the VSMOW2-SLAP2 scale with an uncertainty assessment

In this study, we measured the oxygen stable isotope ratio of drinking water using gas chromatography isotope ratio mass spectrometry. The δ¹⁸O value of drinking water was normalized based on the Vienna Standard Mean Ocean Water 2 (VSMOW2), Standard Light Antarctic Precipitation 2 (SLAP2), and Greenland Ice Sheet Precipitation (GISP) scale by CO₂ equilibrium for 24 h. The isotope ratio responses of a dummy sample drifted as much as 0.145‰ due to a significant decrease in the amount of injected sample. The autodilution technique improved measurement precision of the δ¹⁸O of dummy sample two-fold compared to that without autodilution to 0.025‰. The autodilution of an injected concentration of equilibrated CO₂ also helped improve the measurement precision of the isotope ratio response. The drift of the ratio responses was tested using linear model regression to validate linearity within the sample concentration and isotope ratio ranges. Measurement reliability was assessed using various statistical approaches. One-way analysis of variance verified non-reproducible results of individual measurements. Normalization uncertainties were then assessed by various normalization schemes including two-point and three-point values consisting of the VSMOW2, SLAP2, and GISP standards, showing equivalent results associated with similar extent of normalization uncertainties among various normalization methods. In particular, the uncertainty of the GISP (0.09‰) contributed to one-third of the total normalization uncertainty, implying that the three-point normalization can be improved by a potential standard of which uncertainty is equivalent to the bracketing standards, VSMOW2 (0.02‰) and SLAP2 (0.02‰).

A multi-time-scale power prediction model of hydropower station considering multiple uncertainties

Hydropower, as one of renewable energies, has been widely used all over the world. The uncertainties such as reservoir inflows and electricity price cause random changes in the output power and the hydropower generation benefit. Thus, it is important to research on the power prediction of hydropower station considering the uncertainties. This study proposes a multi-time-scale power prediction model of hydropower station based on dynamic Bayesian network theory, considering the uncertainties of reservoir inflow, electricity price, and hydropower consumption rate. The proposed model consists of three components: a multi-time-scale coupling operation (MCO) model, a dynamic Bayesian network (DBN) model, and a probability-based prediction (PBP) model for decision making. The MCO model provides training data inputs for the DBN model, which is established based on expert knowledge and the relationships among the uncertainties. The PBP model performs power prediction of the hydropower station for decision making using the trained DBN. We apply the proposed model to the Tankeng hydropower station in China. The results show that the model not only quantitatively predicts the multi-time-scale output power and benefit of the hydropower station considering the uncertainties, but also provides the risks of power generation deficiency and power output deficiency.

Anthropogenic atmospheric toxic metals emission inventory and its spatial characteristics in Guangdong province, China

Atmospheric toxic metals (TMs) may cause adverse effects on the environment and human health due to their bioavailability and toxicity. High-resolution TMs emission inventory is important input data for assessing human exposure risks, especially synergistic toxicity of multiple toxic metals. By using the latest city- and enterprise-level environment statistical data, an emission inventory of five TMs (Hg, As, Pb, Cd, Cr) in Guangdong province for the year of 2014 was developed using a bottom-up approach. The total emissions of Hg, As, Pb, Cd and Cr in Guangdong were estimated as 17.70, 32.59, 411.34, 13.13, and 84.16 t, respectively. Major emission sources for each TM were different. Hg emissions were dominated by coal combustion (33%), fluorescent lamp (18%) and cement (17%). 78% of Hg emissions were in the form of Hg⁰, 19% of Hg²⁺, and only 3% of Hg^p due to strict particulate matter control policies. Coal combustion (48%), nonferrous metal smelting (25%) and iron and steel industry (24%) were the major sources of As. Pb emissions primarily came from battery production (42%), iron and steel industry (21%) and gasoline combustion (17%). Cd and Cr emissions were dominated by nonferrous metal smelting (71%) and iron and steel industry (82%), respectively. Most of these TMs were emitted in the non-Pearl River Delta region, where the newly-built iron and steel industry, nonferrous metal smelting and cement production factories were intense. The uncertainties in the five TM emissions were high, due much to high uncertainties in TM emission factors and limited activity data. Thus, to improve the accuracy of these estimates, we recommend more field tests of TM emissions, especially for the industrial process sector. This study provides scientific support for formulating robust TMs control policies to alleviate the high risk of TMs exposure in Guangdong.

Reducing uncertainty of estimated nitrogen load reductions to aquatic systems through spatially targeting agricultural mitigation measures using groundwater nitrogen reduction

The need to further abate agricultural nitrate (N)-loadings to coastal waters in Denmark represents the main driver for development of a new spatially targeted regulation that focus on locating N-mitigation measures in agricultural areas with high N-load. This targeting makes use of the spatial variation across the landscape in natural N-reduction (denitrification) of leached nitrate in the groundwater and surface water systems. A critical basis for including spatial targeting in regulation of N-load in Denmark is the uncertainty associated with the effect of spatially targeting measures, since the effect will be critically affected by uncertainty in the quantification of the spatial variation in N-reduction. In this study, we used 30 equally plausible N-reduction maps, at 100 m grid and sub-catchment resolutions, for the 85-km² groundwater dominated Norsminde catchment in Denmark, applying set-aside as the measure on high N-load areas to reach a N-load reduction target of 20%. The uncertainty on these N-reduction maps resulted in uncertainty on the estimated N-load and on the required set-aside area. We tested several methods for spatially targeting set-aside that took into account the uncertainty on set-aside area and developed methods to reduce uncertainty on the estimated N-load reductions. These methods includes application of set-aside based on each individual N-reduction map compared to a mean N-reduction map, using spatial frequency of high N-load and using spatial frequency of low N-reduction. The results revealed that increasing the ensemble size for averaging the N-reduction maps would decrease the uncertainty on the estimated set-aside area with a stable effect when using an ensemble of 15 or more maps. The spatial resolution of the groundwater N-reduction map is essential for the effectiveness of set-aside, but uncertainty of the finer spatial resolution of N-reduction is greater compared to sub-catchment scale, and application of a spatially targeted strategy with uncertain N-reduction maps will result in incorrect set-aside area and uncertain estimations of N-load reductions. To reduce the uncertainty on estimated N-load reductions, this study finds the method of set-aside application based on spatial frequency of high N-load to be more effective than other methods tested.

LCIA framework and cross-cutting issues guidance within the UNEP-SETAC Life Cycle Initiative

Increasing needs for decision support and advances in scientific knowledge within life cycle assessment (LCA) led to substantial efforts to provide global guidance on environmental life cycle impact assessment (LCIA) indicators under the auspices of the UNEP-SETAC Life Cycle Initiative. As part of these efforts, a dedicated task force focused on addressing several LCIA cross-cutting issues as aspects spanning several impact categories, including spatiotemporal aspects, reference states, normalization and weighting, and uncertainty assessment. Here, findings of the cross-cutting issues task force are presented along with an update of the existing UNEP-SETAC LCIA emission-to-damage framework. Specific recommendations are provided with respect to metrics for human health (Disability Adjusted Life Years, DALY) and ecosystem quality (Potentially Disappeared Fraction of species, PDF). Additionally, we stress the importance of transparent reporting of characterization models, reference states, and assumptions, in order to facilitate cross-comparison between chosen methods and indicators. We recommend developing spatially regionalized characterization models, whenever the nature of impacts shows spatial variability and related spatial data are available. Standard formats should be used for reporting spatially differentiated models, and choices regarding spatiotemporal scales should be clearly communicated. For normalization, we recommend using external normalization references. Over the next two years, the task force will continue its effort with a focus on providing guidance for LCA practitioners on how to use the UNEP-SETAC LCIA framework as well as for method developers on how to consistently extend and further improve this framework.

The impacts of decision uncertainty on municipal solid waste management

Municipal solid waste treatment options are not necessarily pragmatic if the stakeholders in the system don't mutually agree on their shares of liabilities. Stakeholders will select an option if their benefits are maximized and costs are minimized. A decision support framework is required to assess various waste treatment options and predict the optimal decision, considering multiple criteria and conflicting preferences of multiple stakeholders. Because of the inherent complexity, uncertainty is unavoidable and should be acknowledged to enhance the reliability in the decision-making process. Uncertainties in the cost and benefit estimates, and stakeholders' ability in verbalizing their preferences and their knowledge about each other's priorities can impact the outcome of such environmental management problem. In this study, uncertainty assessment methods such as sensitivity analysis, fuzzy Analytical Hierarchy Process, and Bayesian games have been explored. A case study in Vancouver (BC, Canada) has been used as a proof of concept.

Assimilation of optical and radar remote sensing data in 3D mapping of soil properties over large areas

Soil is very important for many land functions. To achieve sustainability it is important to understand how soils vary over space in the landscape. Remote sensing data can be instrumental in mapping and spatial modelling of soil properties, resources and their variability. The aims of this study were to compare satellite sensors (MODIS, Landsat, Sentinel-1 and Sentinel-2) with varying spatial, temporal and spectral resolutions for Digital Soil Mapping (DSM) of a set of soil properties in Scotland, evaluate the potential benefits of adding Sentinel-1 data to DSM models, select the most suited mix of sensors for DSM to map the considered set of soil properties and validate the results of topsoil (2D) and whole profile (3D) models. The results showed that the use of a mixture of sensors proved more effective to model and map soil properties than single sensors. The use of radar Sentinel-1 data proved useful for all soil properties, improving the prediction capability of models with only optical bands. The use of MODIS time series provided stronger relationships than the use of temporal snapshots. The results showed good validation statistics with a RMSE below 20% of the range for all considered soil properties. The RMSE improved from previous studies including only MODIS sensor and using a coarser prediction grid. The performance of the models was similar to previous studies at regional, national or continental scale. A mix of optical and radar data proved useful to map soil properties along the profile. The produced maps of soil properties describing both lateral and vertical variability, with associated uncertainty, are important for further modelling and management of soil resources and ecosystem services. Coupled with further data the soil properties maps could be used to assess soil functions and therefore conditions and suitability of soils for a range of purposes.

Utilization of health insurance data in an environmental epidemiology

OBJECTIVES

In South Korea, health insurance data are used as material for the health insurance of national whole subject. In general, health insurance data could be useful for estimating prevalence or incidence rate that is representative of the actual value in a population. The purpose of this study was to apply the concept of episode of care (EoC) in the utilization of health insurance data in the field of environmental epidemiology and to propose an improved methodology through an uncertainty assessment of disease course and outcome.

METHODS

In this study, we introduced the concept of EoC as a methodology to utilize health insurance data in the field of environmental epidemiology. The characterization analysis of the course and outcome of applying the EoC concept to health insurance data was performed through an uncertainty assessment.

RESULTS

The EoC concept in this study was applied to heat stroke (International Classification of Disease, 10th revision, code T67). In the comparison of results between before and after applying the EoC concept, we observed a reduction in the deviation of daily claims after applying the EoC concept. After that, we categorized context, model, and input uncertainty and characterized these uncertainties in three dimensions by using uncertainty typology.

CONCLUSIONS

This study is the first to show the process of constructing episode data for environmental epidemiological studies by using health insurance data. Our results will help in obtaining representative results for the processing of health insurance data in environmental epidemiological research. Furthermore, these results could be used in the processing of health insurance data in the future.

Probabilistic parameter estimation of activated sludge processes using Markov Chain Monte Carlo

One of the most important challenges in making activated sludge models (ASMs) applicable to design problems is identifying the values of its many stoichiometric and kinetic parameters. When wastewater characteristics data from full-scale biological treatment systems are used for parameter estimation, several sources of uncertainty, including uncertainty in measured data, external forcing (e.g. influent characteristics), and model structural errors influence the value of the estimated parameters. This paper presents a Bayesian hierarchical modeling framework for the probabilistic estimation of activated sludge process parameters. The method provides the joint probability density functions (JPDFs) of stoichiometric and kinetic parameters by updating prior information regarding the parameters obtained from expert knowledge and literature. The method also provides the posterior correlations between the parameters, as well as a measure of sensitivity of the different constituents with respect to the parameters. This information can be used to design experiments to provide higher information content regarding certain parameters. The method is illustrated using the ASM1 model to describe synthetically generated data from a hypothetical biological treatment system. The results indicate that data from full-scale systems can narrow down the ranges of some parameters substantially whereas the amount of information they provide regarding other parameters is small, due to either large correlations between some of the parameters or a lack of sensitivity with respect to the parameters.

Selection of appropriate tumour data sets for Benchmark Dose Modelling (BMD) and derivation of a Margin of Exposure (MoE) for substances that are genotoxic and carcinogenic: considerations of biological relevance of tumour type, data quality and uncertainty assessment

This article addresses a number of concepts related to the selection and modelling of carcinogenicity data for the calculation of a Margin of Exposure. It follows up on the recommendations put forward by the International Life Sciences Institute - European branch in 2010 on the application of the Margin of Exposure (MoE) approach to substances in food that are genotoxic and carcinogenic. The aims are to provide practical guidance on the relevance of animal tumour data for human carcinogenic hazard assessment, appropriate selection of tumour data for Benchmark Dose Modelling, and approaches for dealing with the uncertainty associated with the selection of data for modelling and, consequently, the derived Point of Departure (PoD) used to calculate the MoE. Although the concepts outlined in this article are interrelated, the background expertise needed to address each topic varies. For instance, the expertise needed to make a judgement on biological relevance of a specific tumour type is clearly different to that needed to determine the statistical uncertainty around the data used for modelling a benchmark dose. As such, each topic is dealt with separately to allow those with specialised knowledge to target key areas of guidance and provide a more in-depth discussion on each subject for those new to the concept of the Margin of Exposure approach.