Browsing by Author "Cipullo, Sabrina"
Now showing 1 - 17 of 17
Results Per Page
Sort Options
Item Open Access Analytical progress and challenges for the detection of oxygenated polycyclic aromatic hydrocarbon transformation products in aqueous and soil environmental matrices: A review(Taylor and Francis, 2019-01-11) Pulleyblank, Coren; Cipullo, Sabrina; Campo, Pablo; Kelleher, Brian; Coulon, FredericOver the past 20 years, a growing body of research has raised concerns about the toxicity, fate, and transport of oxygenated transformation products of polycyclic aromatic hydrocarbons. Research targeting these diverse compounds in soil and water systems has been challenged by a lack of standard analytical techniques and suitable reference materials. However, recent efforts towards the consolidation of traditional analytical techniques as well as the development of novel approaches to improve sample preparation and hyphenated instrumental techniques show promise. This review discusses progress and challenges for both trends in analytical method development and makes recommendations for supporting oxygenated PAH research.Item Open Access The applicability of spectroscopy methods for estimating potentially toxic elements in soils: state-of-the art and future trends(Taylor and Francis, 2019-05-08) Nawar, S.; Cipullo, Sabrina; Douglas, Reward; Coulon, Frederic; Mouazen, A. M.Potentially toxic elements (PTEs) in soils pose severe threats to the environment and human health. It is therefore imperative to have access to simple, rapid, portable, and accurate methods for their detection in soils. In this regard, the review introduces recent progresses made in the development and applications of spectroscopic methods for in situ semi-quantitative and quantitative detection of PTEs in soil and critically compares them to standard analytical methods. The advantages and limitations of these methods are discussed together with recent advances in chemometrics and data mining techniques allowing to extract useful information based on spectral data. Furthermore, the factors influencing soil spectra and data analysis are discussed and recommendations on how to reduce or eliminate their influences are provided. Future research and development needs for spectroscopy techniques are emphasized, and an analytical framework based on technology integration and data fusion is proposed to improve the measurement accuracy of PTEs in soil.Item Open Access Assessing bioavailability of complex mixtures in contaminated soils: progress made and research needs(Elsevier, 2017-10-17) Cipullo, Sabrina; Prpich, George; Campo, Pablo; Coulon, FredericUnderstanding the distribution, behaviour and interactions of complex chemical mixtures is key for providing the evidence necessary to make informed decisions and implement robust remediation strategies. Much of the current risk assessment frameworks applied to manage land contamination are based on total contaminant concentrations and the exposure assessments embedded within them do not explicitly address the partitioning and bioavailability of chemical mixtures. These oversights may contribute to an overestimation of both the eco-toxicological effects of the fractions and the mobility of contaminants. In turn, this may limit the efficacy of risk frameworks to inform targeted and proportionate remediation strategies. In this review we analyse the science surrounding bioavailability, its regulatory inclusion and the challenges of incorporating bioavailability in decision making process. While a number of physical and chemical techniques have proven to be valuable tools for estimating bioavailability of organic and inorganic contaminants in soils, doubts have been cast on its implementation into risk management soil frameworks mainly due to a general disagreement on the interchangeable use of bioavailability and bioaccessibility, and the associated methods which are still not standardised. This review focuses on the role of biotic and abiotic factors affecting bioavailability along with soil physicochemical properties and contaminant composition. We also included advantages and disadvantages of different extraction techniques and their implications for bioavailability quantitative estimation. In order to move forward the integration of bioavailability into site-specific risk assessments we should (1) account for soil and contaminant physicochemical characteristics and their effect on bioavailability; (2) evaluate receptor's potential exposure and uptake based on mild-extraction; (3) adopt a combined approach where chemical-techniques are used along with biological methods; (4) consider a simplified and cost-effective methodology to apply at regulatory and industry setting; (5) use single-contaminant exposure assessments to inform and predict complex chemical mixture behaviour and bioavailability.Item Open Access Compositional and physicochemical changes in waste materials and biogas production across 7 landfill sites in UK(Elsevier, 2016-08-28) Frank, R. R.; Cipullo, Sabrina; Garcia, J.; Davies, S.; Wagland, Stuart Thomas; Villa, Raffaella; Trois, C.; Coulon, FredericThe aim of this study was to evaluate the spatial distribution of the paper and fines across seven landfill sites (LFS) and assess the relationship between waste physicochemical properties and biogas production. Physicochemical analysis of the waste samples demonstrated that there were no clear trends in the spatial distribution of total solids (TS), moisture content (MC) and waste organic strength (VS) across all LFS. There was however noticeable difference between samples from the same landfill site. The effect of landfill age on waste physicochemical properties showed no clear relationship, thus, providing evidence that waste remains dormant and non-degraded for long periods of time. Landfill age was however directly correlated with the biochemical methane potential (BMP) of waste; with the highest BMP obtained from the most recent LFS. BMP was also correlated with depth as the average methane production decreased linearly with increasing depth. There was also a high degree of correlation between the Enzymatic Hydrolysis Test (EHT) and BMP test results, which motivates its potential use as an alternative to the BMP test method. Further to this, there were also positive correlations between MC and VS, VS and biogas volume and biogas volume and CH4 content. Outcomes of this work can be used to inform waste degradation and methane enhancement strategies for improving recovery of methane from landfills.Item Open Access Diagnostic strategy and risk assessment framework for complex chemical mixtures(2018-10) Cipullo, Sabrina; Coulon, Frederic; Campo Moreno, PabloEnvironmental contamination comprises a complex mixture of both organic and inorganic contaminants. Understanding their distribution, behaviour and chemical interactions provides the evidence necessary to make informed decision and implement robust remediation strategies. However most of the current risk assessment frameworks, used to manage land contamination, are based on the total contaminant concentration rather than the concentration likely to pose significant risk, the bioavailable concentration. Further to this, the exposure assessments embedded within the frameworks do not explicitly address the partitioning and bioavailability of chemical mixtures. This inability may contribute to an overestimation of both the eco-toxicological effects of the fractions and their mobility in air and water; leading to an overestimation of health and environmental effects. In turn, this may limit the efficacy of the risk assessment frameworks to inform targeted and proportionate remediation strategies. The aim of this PhD study was to address this gap by delivering an integrated risk assessment framework for sites contaminated with complex chemical mixtures. Specifically, this PhD study investigated the fate and behaviour of complex mixtures of petroleum hydrocarbons, metals and metalloids in soils and its implication for partitioning, bioavailability and risk assessment through a 12 month mesocosms study. Further to this, an integrated approach, where contaminants bioavailability and distribution changes along with a range of microbiological indicators and ecotoxicological bioassays, was used to provide multiple lines of evidence to support the risk characterisation and assess the remediation end-point over a 6 month study. From the empirical data obtained from the two mesocosm studies, two Machine Leaning (ML) approaches have been developed to provide a quick and reliable tool to assess multi-contaminated sites with Visible and Near-Infrared Spectroscopy (Vis-NIRS), and to predict bioavailability and toxicity changes occurring during bioremediation. Overall this PhD study shed light on the behaviour of bioavailability, and toxicity of complex chemical mixtures in soils genuinely contaminated. This was supported through a comprehensive and integrated analytical framework providing the necessary lines of evidence to evaluate the implications for risk assessment and identify the end point remediation. The developed framework can significantly help to identify optimal remediation strategies and contribute to change the over-conservative nature of the current risk assessments.Item Open Access Enhanced pilot bioremediation of oily sludge from petroleum refinery disposal under hot-summer Mediterranean climate(Elsevier, 2021-10-23) Said, Olfa Ben; Cravo-Laureau, Cristiana; Armougom, Fabrice; Cipullo, Sabrina; Khelil, Meriem Ben; Yahiya, Marouen Ben Haj; Douihech, Abdeljabar; Beyrem, Hamouda; Coulon, Frederic; Duran, RobertLarge pilot scale bioremediation approaches were implemented for the treatments of oily sludge (OS) characterised by alkaline pH (pH > 9), high concentration of metals (3% dry weight) and high total petroleum hydrocarbons content (TPH) rangingbetween 22,000 and 67,300 mg kg −1 from a Tunisian petroleum refinery. The treatments included bioaugmentation and biostimulation approaches with autochthonous isolated bacterial strains and consortia. Chemical, microbial, and ecotoxicological analyses were performed over a period of 180 days incubation. The bioremediation treatments favoured the development of Proteobacteria, Firmicutes and Bacteroidetes following an ecological succession of specialist bacterial groups, first associated to hydrocarbon degradation (e.g. Marinobacter and Alcanivorax) that resulted in a greater extent of TPH-degradation (up to 80%), and the selection of metal resistant bacteria including Hyphomonas, Phaeobacter, and Desulfuromusa. The best performances were obtained when bioaugmentation and biostimulation were combined. Over 90% of the TPH initial concentration was degraded over 180 days, which was accompanied with a 3-fold reduction of ecotoxicity. Our study demonstrates the efficacy of large pilot scale bioremediation of highly contaminated oily sludge, providing the evidence that the management of autochthonous microbial communities is of paramount importance for the success of the bioremediation process.Item Open Access Evaluation of vis-NIR reflectance spectroscopy sensitivity to weathering for enhanced assessment of oil contaminated soils(Elsevier, 2018-02-19) Douglas, Reward K.; Newar, S; Cipullo, Sabrina; Alamar, M. Carmen; Coulon, Frederic; Mouazen, A. M.This study investigated the sensitivity of visible near-infrared spectroscopy (vis-NIR) to discriminate between fresh and weathered oil contaminated soils. The performance of random forest (RF) and partial least squares regression (PLSR) for the estimation of total petroleum hydrocarbon (TPH) throughout the time was also explored. Soil samples (n = 13) with 5 different textures of sandy loam, sandy clay loam, clay loam, sandy clay and clay were collected from 10 different locations across the Cranfield University's Research Farm (UK). A series of soil mesocosms was then set up where each soil sample was spiked with 10 ml of Alaskan crude oil (equivalent to 8450 mg/kg), allowed to equilibrate for 48 h (T2 d) and further kept at room temperature (21 °C). Soils scanning was carried out before spiking (control TC) and then after 2 days (T2 d) and months 4 (T4 m), 8 (T8 m), 12 (T12 m), 16 (T16 m), 20 (T20 m), 24 (T24 m), whereas gas chromatography mass spectroscopy (GC–MS) analysis was performed on T2 d, T4 m, T12 m, T16 m, T20 m, and T24 m. Soil scanning was done simultaneously using an AgroSpec spectrometer (305 to 2200 nm) (tec5 Technology for Spectroscopy, Germany) and Analytical Spectral Device (ASD) spectrometer (350 to 2500 nm) (ASDI, USA) to assess and compare their sensitivity and response against GC–MS data. Principle component analysis (PCA) showed that ASD performed better than tec5 for discriminating weathered versus fresh oil contaminated soil samples. The prediction results proved that RF models outperformed PLSR and resulted in coefficient of determination (R2) of 0.92, ratio of prediction deviation (RPD) of 3.79, and root mean square error of prediction (RMSEP) of 108.56 mg/kg. Overall, the results demonstrate that vis–NIR is a promising tool for rapid site investigation of weathered oil contamination in soils and for TPH monitoring without the need of collecting soil samples and lengthy hydrocarbon extraction for further quantification analysis.Item Open Access Factors governing the solid phase distribution of Cr, Cu and As in contaminated soil after 40 years of ageing(Elsevier, 2018-10-18) Tardif, Stacie; Cipullo, Sabrina; Sø, Helle U.; Wragg, Joanna; Holm, Peter E.; Coulon, Frederic; Brandt, Kristian K.; Cave, MarkThe physico-chemical factors affecting the distribution, behavior and speciation of chromium (Cr), copper (Cu) and arsenic (As) was investigated at a former wood impregnation site (Fredensborg, Denmark). Forty soil samples were collected and extracted using a sequential extraction technique known as the Chemometric Identification of Substrates and Element Distributions (CISED) and a multivariate statistical tool (redundancy analysis) was applied. CISED data was linked to water-extractable Cr, Cu and As and bioavailable Cu as determined by a whole-cell bacterial bioreporter assay. Results showed that soil pH significantly affected the solid phase distribution of all three elements on site. Additionally, elements competing for binding sites, Ca, Mg and Mn in the case of Cu, and P, in the case of As, played a major role in the distribution of these elements in soil. Element-specific distributions were observed amongst the six identified soil phases including residual pore salts, exchangeable, carbonates (tentative designation), Mn-Al oxide, amorphous Fe oxide, and crystalline Fe oxide. While Cr was strongly bound to non-extractable crystalline Fe oxide in the oxic top soil, Cu and notably, As were associated with readily extractable phases, suggesting that Cu and As, and not Cr, constitute the highest risk to environmental and human health. However, bioavailable Cu did not significantly correlate with CISED identified soil phases, suggesting that sequential extraction schemes such as CISED may not be ideally suited for inferring bioavailability to microorganisms in soil and supports the integration of receptor-specific bioavailability tests into risk assessments as a complement to chemical methods.Item Open Access Fingerprinting ambient air to understand bioaerosol profiles in three different environments in the South East of England(Cranfield University, 2020-02-24 08:10) Coulon, Frederic; Garcia Alcega, Sonia; Tyrrel, Sean; Nasar, Zaheer; Drew, Gill; Cipullo, Sabrina; colbeck, ian; ferguson, Robert; Whitby, Corinne; J. Dumbrell, Alex; Yan, ChengRaw data used and supporting the data and results presented in: "Fingerprinting ambient air to understand bioaerosol profiles in three different environments in the South East of England" Science of the Total EnvironmentItem Open Access Fingerprinting ambient air to understand bioaerosol profiles in three different environments in the South East of England(Elsevier, 2020-02-24) Garcia Alcega, Sonia; Nasir, Zaheer A.; Cipullo, Sabrina; Ferguson, Robert M. W.; Yan, Cheng; Whitby, Corinne; Dumbrell, Alex J.; Drew, Gillian; Colbeck, Ian; Tyrrel, Sean F.Molecular and chemical fingerprints from 10 contrasting outdoor air environments, including three agricultural farms, three urban parks and four industrial sites were investigated to advance our understanding of bioaerosol distribution and emissions. Both phospholipid fatty acids (PLFA) and microbial volatile organic compounds (MVOC) profiles showed a different distribution in summer compared to winter. Further to this, a strong positive correlation was found between the total concentration of MVOCs and PLFAs (r = 0.670, p = 0.004 in winter and r = 0.767, p = 0.001 in summer) demonstrating that either chemical or molecular fingerprints of outdoor environments can provide good insights into the sources and distribution of bioaerosols. Environment specific variables and most representative MVOCs were identified and linked to microbial species emissions via a MVOC database and PLFAs taxonomical classification. While similar MVOCs and PLFAs were identified across all the environments suggesting common microbial communities, specific MVOCs were identified for each contrasting environment. Specifically, 3,4-dimethylpent-1-yn-3-ol, ethoxyethane and propanal were identified as key MVOCs for the industrial areas (and were correlated to fungi, Staphylococcus aureus (Gram positive bacteria) and Gram negative bacteria, R = 0.863, R = 0.618 and R = 0.676, respectively) while phthalic acid, propene and isobutane were key for urban environments (correlated to Gram negative bacteria, fungi and bacteria, R = 0.874, R = 0.962 and R = 0.969 respectively); and ethanol, 2-methyl-2-propanol, 2-methyl-1-pentene, butane, isoprene and methyl acetate were key for farms (correlated to fungi, Gram positive bacteria and bacteria, R = 0.690 and 0.783, R = 0.706 and R = 0.790, 0.761 and 0.768). The combination of MVOCs and PLFAs markers can assist in rapid microbial fingerprinting of distinct environmental influences on ambient air quality.Item Open Access Incorporating oral bioaccessibility into human health risk assessment due to potentially toxic elements in extractive waste and contaminated soils from an abandoned mine site(Elsevier, 2020-04-30) Mehta, Neha; Cipullo, Sabrina; Cocerva, Tatiana; Coulon, Frederic; Dino, Giovanna Antonella; Ajmone-Marsan, Franco; Padoan, Elio; Cox, Siobhan Fiona; Cave, Mark R.; De Luca, Domenico AntonioThe waste rock, tailings and soil around an abandoned mine site in Gorno (northwest Italy) contain elevated concentrations of potentially toxic elements (PTE) exceeding the permissible limits for residential uses. Specifically, the maximum concentrations of As, Cd, Pb, and Zn were 107 mg/kg, 340 mg/kg, 1064 mg/kg, and 148 433 mg/kg, respectively. A site-specific human health risk assessment (HHRA) was conducted for residential and recreational exposure scenarios, using an approach based on Risk Based Corrective Action (RBCA) method, refined by incorporating oral bioaccessibility data. Oral bioaccessibility analyses were performed by simulating the human digestion process in vitro (Unified BARGE Method). Detailed analysis of oral bioaccessible fraction (BAF i.e. ratio of bioaccessible concentrations to total concentrations on <250 μm fraction) indicated BAF of As (5-33%), Cd (72-98%), Co (24-42%), Cr (3-11%), Cu (25-90%), Ni (17-60%), Pb (16-88%) and Zn (73-94%). The solid phase distribution and mineralogical analyses showed that the variation of BAF is attributed to presence of alkaline calcareous rocks and association of PTE with a variety of minerals. The HHRA for ingestion pathway, suggested that bioaccessibility-corrected cancer risk reached up to 2.7 × 10−5 and 0.55 × 10−5 for residential and recreational senarios respectively (acceptable level is 1 × 10−5). The hazard index (HI) recalculated after incorporation of oral bioaccessible concentrations for a residential scenario ranged from 0.02 to 17.9. This was above the acceptable level (>1) for 50% of the samples, indicating potential human health risks. This study provides information for site-specific risk assessments and planning future research.Item Open Access Linking bioavailability and toxicity changes of complex chemicals mixture to support decision making for remediation endpoint of contaminated soils(Elsevier, 2018-09-27) Cipullo, Sabrina; Negrin, I.; Claveau, Leila; Snapir, Boris; Tardif, S.; Pulleyblank, C.; Prpich, George; Campo Moreno, Pablo; Coulon, FredericA six-month laboratory scale study was carried out to investigate the effect of biochar and compost amendments on complex chemical mixtures of tar, heavy metals and metalloids in two genuine contaminated soils. An integrated approach, where organic and inorganic contaminants bioavailability and distribution changes, along with a range of microbiological indicators and ecotoxicological bioassays, was used to provide multiple lines of evidence to support the risk characterisation and assess the remediation end-point. Both compost and biochar amendment (p = 0.005) as well as incubation time (p = 0.001) significantly affected the total and bioavailable concentrations of the total petroleum hydrocarbons (TPH) in the two soils. Specifically, TPH concentration decreased by 46% and 30% in Soil 1 and Soil 2 amended with compost. These decreases were accompanied by a reduction of 78% (Soil 1) and 6% (Soil 2) of the bioavailable hydrocarbons and the most significant decrease was observed for the medium to long chain aliphatic compounds (EC16–35) and medium molecular weight aromatic compounds (EC16–21). Compost amendment enhanced the degradation of both the aliphatic and aromatic fractions in the two soils, while biochar contributed to lock the hydrocarbons in the contaminated soils. Neither compost nor biochar affected the distribution and behaviour of the heavy metals (HM) and metalloids in the different soil phases, suggesting that the co-presence of heavy metals and metalloids posed a low risk. Strong negative correlations were observed between the bioavailable hydrocarbon fractions and the ecotoxicological assays suggesting that when bioavailable concentrations decreased, the toxicity also decreased. This study showed that adopting a combined diagnostic approach can significantly help to identify optimal remediation strategies and contribute to change the over-conservative nature of the current risk assessments thus reducing the costs associated with remediation endpoint.Item Open Access Linking oral bioaccessibility and solid phase distribution of potentially toxic elements in extractive waste and soil from an abandoned mine site: Case study in Campello Monti, NW Italy(Elsevier, 2018-10-11) Mehta, Neha; Cocerva, Tatiana; Cipullo, Sabrina; Padoan, Elio; Dino, Giovanna Antonella; Ajmone-Marsan, Franco; Cox, Siobhan Fiona; Coulon, Frederic; De Luca, Domenico AntonioOral bioaccessibility and solid phase distribution of potentially toxic elements (PTE) from extractive waste streams were investigated to assess the potential human health risk posed by abandoned mines. The solid phase distribution along with micro-X-ray fluorescence (micro-XRF) and scanning electron microscopy (SEM) analysis were also performed. The results showed that the total concentrations of PTE were higher in <250 μm size fractions of waste rock and soil samples in comparison to the <2 mm size fractions. Mean value of total concentrations of chromium(Cr), copper (Cu), and nickel (Ni) in waste rocks (size fractions <250 μm) were found to be 1299, 1570, and 4010 mg/kg respectively due to the parent material. However, only 11% of Ni in this sample was orally bioaccessible. Detailed analysis of the oral bioaccessible fraction (BAF, reported as the ratio of highest bioaccessible concentration compared with the total concentration from the 250 μm fraction) across all samples showed that Cr, Cu, and Ni varied from 1 to 6%, 14 to 47%, and 5 to 21%, respectively. The variation can be attributed to the difference in pH, organic matter content and mineralogical composition of the samples. Non-specific sequential extraction showed that the non-mobile forms of PTE were associated with the clay and Fe oxide components of the environmental matrices. The present study demonstrates how oral bioaccessibility, solid phase distribution and mineralogical analysis can provide insights into the distribution, fate and behaviour of PTE in waste streams from abandoned mine sites and inform human health risk posed by such sites.Item Open Access A novel machine learning-based approach for the risk assessment of nitrate groundwater contamination(Elsevier, 2018-07-11) Sajedi-Hosseini, Farzaneh; Malekian, Arash; Choubin, Bahram; Rahmati, Omid; Cipullo, Sabrina; Coulon, Frederic; Pradhan, BiswajeetThis study aimed to develop a novel framework for risk assessment of nitrate groundwater contamination by integrating chemical and statistical analysis for an arid region. A standard method was applied for assessing the vulnerability of groundwater to nitrate pollution in Lenjanat plain, Iran. Nitrate concentration were collected from 102 wells of the plain and used to provide pollution occurrence and probability maps. Three machine learning models including boosted regression trees (BRT), multivariate discriminant analysis (MDA), and support vector machine (SVM) were used for the probability of groundwater pollution occurrence. Afterwards, an ensemble modeling approach was applied for production of the groundwater pollution occurrence probability map. Validation of the models was carried out using area under the receiver operating characteristic curve method (AUC); values above 80% were selected to contribute in ensembling process. Results indicated that accuracy for the three models ranged from 0.81 to 0.87, therefore all models were considered for ensemble modeling process. The resultant groundwater pollution risk (produced by vulnerability, pollution, and probability maps) indicated that the central regions of the plain have high and very high risk of nitrate pollution further confirmed by the exiting landuse map. The findings may provide very helpful information in decision making for groundwater pollution risk management especially in semi-arid regions.Item Open Access Predicting bioavailability change of complex chemical mixtures in contaminated soils using visible and near-infrared spectroscopy and random forest regression em(Cranfield University, 2018-12-04 12:08) Coulon, Frederic; Cipullo, Sabrina; Campo Moreno, Pablo; Mouazen, Abdul; Nawar, SaidRaw data for total and bioavailable concentrations of petroleum hydrocarbons compounds, heavy metals and metalloids in the five soils.Spectra raw data are also providedItem Open Access Predicting uncertainty of machine learning models for modelling nitrate pollution of groundwater using quantile regression and UNNEC methods(Elsevier, 2019-06-21) Rahmati, Omid; Choubin, Bahram; Fathabadi, Abolhasan; Coulon, Frederic; Soltani, Elinaz; Shahabi, Himan; Mollaefar, Eisa; Tiefenbacher, John; Cipullo, Sabrina; Bin Ahmad, Baharin; Tien Bui, DieuAlthough 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.Item Open Access Prediction of bioavailability and toxicity of complex chemical mixtures through machine learning models(Elsevier, 2018-09-11) Cipullo, Sabrina; Snapir, Boris; Prpich, George; Campo Moreno, Pablo; Coulon, FredericEmpirical data from a 6-month mesocosms experiment were used to assess the ability and performance of two machine learning (ML) models, including artificial neural network (NN) and random forest (RF), to predict temporal bioavailability changes of complex chemical mixtures in contaminated soils amended with compost or biochar. From the predicted bioavailability data, toxicity response for relevant ecological receptors was then forecasted to establish environmental risk implications and determine acceptable end-point remediation. The dataset corresponds to replicate samples collected over 180 days and analysed for total and bioavailable petroleum hydrocarbons and heavy metals/metalloids content. Further to this, a range of biological indicators including bacteria count, soil respiration, microbial community fingerprint, seeds germination, earthworm's lethality, and bioluminescent bacteria were evaluated to inform the environmental risk assessment. Parameters such as soil type, amendment (biochar and compost), initial concentration of individual compounds, and incubation time were used as inputs of the ML models. The relative importance of the input variables was also analysed to better understand the drivers of temporal changes in bioavailability and toxicity. It showed that toxicity changes can be driven by multiple factors (combined effects), which may not be accounted for in classical linear regression analysis (correlation). The use of ML models could improve our understanding of rate-limiting processes affecting the freely available fraction (bioavailable) of contaminants in soil, therefore contributing to mitigate potential risks and to inform appropriate response and recovery methods.