Assessing the relevance of environmental exposure data sets.

Environmental exposure data are used by decision-makers to assess environmental risks and implement actions to mitigate risks from contaminants. The first article in this series summarized the available evaluation schemes for environmental exposure data, of which there are few compared to those available for environmental hazard data. The second article covered the assessment of the reliability of environmental exposure data sets under the Criteria for the Reporting and Evaluation of Exposure Data (CREED). The aim of this article is to provide an overview and practical guidance on the relevance assessment in the context of the CREED approach for evaluating exposure monitoring data sets. Systematically considering relevance is critical for both evaluating existing data sets and for optimizing the design of new monitoring studies. Relevance is defined here as the degree of suitability or appropriateness of a data set to address a specific purpose or to answer the questions that have been defined by the assessor or for those generating exposure data. The purpose definition will be the foundation for the relevance assessment, to clarify how the assessor should rate the assessment criteria (fully met, partly met, not met/inappropriate, not reported, not applicable). This will provide transparency for anyone reviewing the outcomes. An explicit gap analysis (i.e., an articulation of the data set limitations for the stated purpose) is an important outcome of the relevance assessment. The relevance evaluation approach is demonstrated with three case studies, all relating to the freshwater aquatic environment, where the data sets are scored as relevant with or without restrictions, not relevant, or not assignable. The case studies represent both organic and inorganic constituents, and have different data characteristics (e.g., percentage of censored data, sampling frequencies, relation to supporting parameters). Integr Environ Assess Manag 2024;00:1-15. © 2023 SETAC.

Value and limitations of machine learning in high-frequency nutrient data for gap-filling, forecasting, and transport process interpretation.

High-frequency monitoring of water quality in catchments brings along the challenge of post-processing large amounts of data. Moreover, monitoring stations are often remote and technical issues resulting in data gaps are common. Machine learning algorithms can be applied to fill these gaps, and to a certain extent, for predictions and interpretation. The objectives of this study were (1) to evaluate six different machine learning models for gap-filling in a high-frequency nitrate and total phosphorus concentration time series, (2) to showcase the potential added value (and limitations) of machine learning to interpret underlying processes, and (3) to study the limits of machine learning algorithms for predictions outside the training period. We used a 4-year high-frequency dataset from a ditch draining one intensive dairy farm in the east of The Netherlands. Continuous time series of precipitation, evapotranspiration, groundwater levels, discharge, turbidity, and nitrate or total phosphorus were used as predictors for total phosphorus and nitrate concentrations respectively. Our results showed that the random forest algorithm had the best performance to fill in data-gaps, with R2 higher than 0.92 and short computation times. The feature importance helped understanding the changes in transport processes linked to water conservation measures and rain variability. Applying the machine learning model outside the training period resulted in a low performance, largely due to system changes (manure surplus and water conservation) which were not included as predictors. This study offers a valuable and novel example of how to use and interpret machine learning models for post-processing high-frequency water quality data.

Transport-limited kinetics of phosphate retention on iron-coated sand and practical implications.

Iron-coated sand (ICS) is a by-product from drinking water treatment made of sand coated with ferric iron (hydr)oxides. It is considered a suitable material for large-scale measures for phosphate removal from natural and agricultural waters to prevent eutrophication. Previous studies demonstrated that the residence time of water must be very long to reach equilibrium partitioning between phosphate and ICS but specifics for application are missing. First, SEM-EDX images were used to support the conceptual assumption that P adsorption inside the coating is a transport-limited process. Second, a conceptual model of phosphate adsorption was proposed considering two types of sites: one type with fast adsorption kinetics and reaching equilibrium with the percolating solution, and another type for which adsorption is also reversible but described by pseudo-first-order kinetics. The latter is conceived to account for transport-limited adsorption in the interior of the coating while the former fraction of sites is assumed to be easily accessible and located close to the grain surface. Third, the kinetics of phosphate adsorption on ICS were quantitatively determined to describe and predict phosphate retention in filters under various flow conditions. The model was calibrated and validated with long-term column experiments, which lasted for 3500 h to approach equilibrium on the slowly reacting sites. The model reproduced the outflowing phosphate concentrations: the pronounced increase after a few pore volumes and the slow increase over the remaining part of the experiment. The parameterized model was also able to predict the time evolution of phosphate concentrations in the outflow of column experiments with different flow velocities, flow interruption, and in desorption experiments. The equilibrium partition coefficient for the experimental conditions was identified as 28.1 L/g-Fe at pH 6.8 and a phosphate concentration of 1.7 mg-P / L. The optimized first-order mass transfer coefficient for the slow adsorption process was 1.56 10-4 h-1, implying that the slow adsorption process has a time scale of several months. However, based on the parameterized model, the slow adsorption process accounted for 95.5% of the equilibrium adsorption capacity, emphasizing the potential relevance of this process for practical applications. The implications for the design, operation, and lifespan of ICS filters are exemplarily illustrated for different scenarios.

Forms of phosphorus in suspended particulate matter in agriculture-dominated lowland catchments: Iron as phosphorus carrier.

The fate and environmental effects of phosphorus (P) in natural waters depend on its chemical forms. The particulate P (PP) concentration is dominant over the dissolved P concentration in agriculture-dominated headwaters in the Netherlands. Routine water quality monitoring programmes do not include the chemical fractionation of PP. To quantify the chemical forms of PP under various conditions in six agriculture-dominated lowland catchments in the Netherlands, a sequential chemical extraction method was applied to suspended particulate matter (SPM) samples collected by centrifugation or filtration. Centrifuge samples had lower values for the sum of the PP fractions compared with the filtration samples due to lower contents from PP fractions other than the Fe-P pool. With an average value of 8.8mgg-1, internationally high P contents of the SPM were found. Ferric iron-bound P was the most important PP fraction in SPM samples (38-95%; median 74%), followed by organic P (2-38%; median 15%). Exchangeable P ranged from 0.2 to 27%, with a median of 4.4%, Ca-P ranged from 0.1 to 11% with a median of 3.9% and detrital P was present in only a small fraction (0-6%; median 1.1%). Ferric iron-bound P was the dominant PP pool throughout the entire range of watercourses (from headwater ditches to catchment outlets) and in samples taken during winter months as well as those taken during summer months. Furthermore, the PP fraction distribution did not change markedly when flow conditions were altered from low to high discharge. The dominance of the Fe-P pool denotes the presence of Fe(III) precipitates in SPM that originate from exfiltration of anoxic Fe-bearing groundwater. These Fe(III) precipitates are a major fraction of the total SPM concentration (4 to 67% as Fe(OH)3; median 18%). Although not measured directly, our results suggest that formation of authigenic Fe(III) precipitates causes a rapid transformation of dissolved P in groundwater to PP in surface water. We advise including sequential chemical extraction of SPM monitoring programmes because the composition of particles is critical for P bioavailability, which is a key driving factor for eutrophication.

Metal immobilization in soils using synthetic zeolites.

In situ immobilization of heavy metals in contaminated soils is a technique to improve soil quality. Synthetic zeolites are potentially useful additives to bind heavy metals. This study selected the most effective zeolite in cadmium and zinc binding out of six synthetic zeolites (mordenite-type, faujasite-type, zeolite X, zeolite P, and two zeolites A) and one natural zeolite (clinoptilolite). Zeolite A appeared to have the highest binding capacity between pH 5 and 6.5 and was stable above pH 5.5. The second objective of this study was to investigate the effects of zeolite addition on the dissolved organic matter (DOM) concentration. Since zeolites increase soil pH and bind Ca, their application might lead to dispersion of organic matter. In a batch experiment, the DOM concentration increased by a factor of 5 when the pH increased from 6 to 8 as a result of zeolite A addition. A strong increase in DOM was also found in the leachate of soil columns, particularly in the beginning of the experiment. This resulted in higher metal leaching caused by metal-DOM complexes. In contrast, the free ionic concentration of Cd and Zn strongly decreased after the addition of zeolites, which might explain the reduction in metal uptake observed in plant growth experiments. Pretreatment of zeolites with acid (to prevent a pH increase) or Ca (to coagulate organic matter) suppressed the dispersion of organic matter, but also decreased the metal binding capacity of the zeolites due to competition of protons or Ca.

Measuring and modeling zinc and cadmium binding by humic acid.

Free metal ions in aqueous and terrestrial systems strongly influence bioavailability and toxicity. Most analytical techniques determine the total metal concentration, including the metal ions bound by dissolved organic matter. Ion activity can be measured with ion-specific electrodes (ISEs) for some metals, but an electrode for Zn is not commercially available. As a result, very few data are available on Zn binding by natural dissolved organic matter. The aim of this study is to determine free Zn concentrations in purified humic acid solutions using the recently developed Donnan membrane technique. However, several analytical aspects of the Donnan membrane technique had to be clarified before reliable data could be composed. Cd was chosen for validation. This study shows that free Cd concentrations as measured by the Donnan membrane technique agreed well with Cd ISE measurements. It is also shown that the Donnan membrane technique could be used at high pH. The Donnan membrane technique provided consistent results in a range of p[Cd2+] = 3-9 and p[Zn2+] = 3-8 at pH 4, 6, and 8. Metal speciation in humic acid solutions was also calculated with the consistent NICA-Donnan model using generic parameters. The model could excellently describe the experimental data without adjusting any of the parameters (R2Cd = 0.971, R2Zn = 0.988).