Implementation of Bayesian networks and Bayesian inference using a Cu0.1Te0.9/HfO2/Pt threshold switching memristor.

Bayesian networks and Bayesian inference, which forecast uncertain causal relationships within a stochastic framework, are used in various artificial intelligence applications. However, implementing hardware circuits for the Bayesian inference has shortcomings regarding device performance and circuit complexity. This work proposed a Bayesian network and inference circuit using a Cu0.1Te0.9/HfO2/Pt volatile memristor, a probabilistic bit neuron that can control the probability of being 'true' or 'false.' Nodal probabilities within the network are feasibly sampled with low errors, even with the device's cycle-to-cycle variations. Furthermore, Bayesian inference of all conditional probabilities within the network is implemented with low power (<186 nW) and energy consumption (441.4 fJ), and a normalized mean squared error of ∼7.5 × 10-4 through division feedback logic with a variational learning rate to suppress the inherent variation of the memristor. The suggested memristor-based Bayesian network shows the potential to replace the conventional complementary metal oxide semiconductor-based Bayesian estimation method with power efficiency using a stochastic computing method.

Vertically Stackable Ovonic Threshold Switch Oscillator Using Atomic Layer Deposited Ge0.6Se0.4 Film for High-Density Artificial Neural Networks.

Nanodevice oscillators (nano-oscillators) have received considerable attention to implement in neuromorphic computing as hardware because they can significantly improve the device integration density and energy efficiency compared to complementary metal oxide semiconductor circuit-based oscillators. This work demonstrates vertically stackable nano-oscillators using an ovonic threshold switch (OTS) for high-density neuromorphic hardware. A vertically stackable Ge0.6Se0.4 OTS-oscillator (VOTS-OSC) is fabricated with a vertical crossbar array structure by growing Ge0.6Se0.4 film conformally on a contact hole structure using atomic layer deposition. The VOTS-OSC can be vertically integrated onto peripheral circuits without causing thermal damage because the fabrication temperature is <400 °C. The fabricated device exhibits oscillation characteristics, which can serve as leaky integrate-and-fire neurons in spiking neural networks (SNNs) and coupled oscillators in oscillatory neural networks (ONNs). For practical applications, pattern recognition and vertex coloring are demonstrated with SNNs and ONNs, respectively, using semiempirical simulations. This structure increases the oscillator integration density significantly, enabling complex tasks with a large number of oscillators. Moreover, it can enhance the computational speed of neural networks due to its rapid switching speed.

2Memristor-1Capacitor Integrated Temporal Kernel for High-Dimensional Data Mapping.

Compact but precise feature-extracting ability is core to processing complex computational tasks in neuromorphic hardware. Physical reservoir computing (RC) offers a robust framework to map temporal data into a high-dimensional space using the time dynamics of a material system, such as a volatile memristor. However, conventional physical RC systems have limited dynamics for the given material properties, restricting the methods to increase their dimensionality. This study proposes an integrated temporal kernel composed of a 2-memristor and 1-capacitor (2M1C) using a W/HfO2 /TiN memristor and TiN/ZrO2 /Al2 O3 /ZrO2 /TiN capacitor to achieve higher dimensionality and tunable dynamics. The kernel elements are carefully designed and fabricated into an integrated array, of which performances are evaluated under diverse conditions. By optimizing the time dynamics of the 2M1C kernel, each memristor simultaneously extracts complementary information from input signals. The MNIST benchmark digit classification task achieves a high accuracy of 94.3% with a (196×10) single-layer network. Analog input mapping ability is tested with a Mackey-Glass time series prediction, and the system records a normalized root mean square error of 0.04 with a 20×1 readout network, the smallest readout network ever used for Mackey-Glass prediction in RC. These performances demonstrate its high potential for efficient temporal data analysis.

A high-dimensional in-sensor reservoir computing system with optoelectronic memristors for high-performance neuromorphic machine vision.

In-sensor reservoir computing (RC) is a promising technology to reduce power consumption and training costs of machine vision systems by processing optical signals temporally. This study demonstrates a high-dimensional in-sensor RC system with optoelectronic memristors to enhance the performance of the in-sensor RC system. Because optoelectronic memristors can respond to both optical and electrical stimuli, optical and electrical masks are proposed to improve the dimensionality and performance of the in-sensor RC system. An optical mask is employed to regulate the wavelength of light, while an electrical mask is used to control the initial conductance of zinc oxide optoelectronic memristors. The distinct characteristics of these two masks contribute to the representation of various distinguishable reservoir states, making it possible to implement diverse reservoir configurations with minimal correlation and to increase the dimensionality of the in-sensor RC system. Using the high-dimensional in-sensor RC system, handwritten digits are successfully classified with an accuracy of 94.1%. Furthermore, human action pattern recognition is achieved with a high accuracy of 99.4%. These high accuracies are achieved with the use of a single-layer readout network, which can significantly reduce the network size and training costs.

Spatiotemporal Data Processing with Memristor Crossbar-Array-Based Graph Reservoir.

Memristor-based physical reservoir computing (RC) is a robust framework for processing complex spatiotemporal data parallelly. However, conventional memristor-based reservoirs cannot capture the spatial relationship between the time-varying inputs due to the specific mapping scheme assigning one input signal to one memristor conductance. Here, a physical "graph reservoir" is introduced using a metal cell at the diagonal-crossbar array (mCBA) with dynamic self-rectifying memristors. Input and inverted input signals are applied to the word and bit lines of the mCBA, respectively, storing the correlation information between input signals in the memristors. In this way, the mCBA graph reservoirs can map the spatiotemporal correlation of the input data in a high-dimensional feature space. The high-dimensional mapping characteristics of the graph reservoir achieve notable results, including a normalized root-mean-square error of 0.09 in Mackey-Glass time series prediction, a 97.21% accuracy in MNIST recognition, and an 80.0% diagnostic accuracy in human connectome classification.

Cross-Wired Memristive Crossbar Array for Effective Graph Data Analysis.

Graphs adequately represent the enormous interconnections among numerous entities in big data, incurring high computational costs in analyzing them with conventional hardware. Physical graph representation (PGR) is an approach that replicates the graph within a physical system, allowing for efficient analysis. This study introduces a cross-wired crossbar array (cwCBA), uniquely connecting diagonal and non-diagonal components in a CBA by a cross-wiring process. The cross-wired diagonal cells enable cwCBA to achieve precise PGR and dynamic node state control. For this purpose, a cwCBA is fabricated using Pt/Ta2O5/HfO2/TiN (PTHT) memristor with high on/off and self-rectifying characteristics. The structural and device benefits of PTHT cwCBA for enhanced PGR precision are highlighted, and the practical efficacy is demonstrated for two applications. First, it executes a dynamic path-finding algorithm, identifying the shortest paths in a dynamic graph. PTHT cwCBA shows a more accurate inferred distance and ≈1/3800 lower processing complexity than the conventional method. Second, it analyzes the protein-protein interaction (PPI) networks containing self-interacting proteins, which possess intricate characteristics compared to typical graphs. The PPI prediction results exhibit an average of 30.5% and 21.3% improvement in area under the curve and F1-score, respectively, compared to existing algorithms.

Emergency contraceptive provision in the emergency department and risk analysis for delayed utilization in South Korea.

In certain nations, the emergency department (ED) has been designated as the primary center to provide emergency contraception (EC). However, analyses of ED visits for EC are limited. Moreover, ED-based research that focuses on time is limited to only a few surveys. The aims of this study were to examine the characteristics of ED visitors for EC and the interval between the coitus and arrival at the ED, and to analyze the factors associated with delays in visiting the ED. This retrospective cohort study involved patients at 2 urban tertiary academic hospitals in South Korea. All patients who presented to the ED for EC between January 2019 and December 2021 were analyzed. The median age of the participants was 26 years. The most common variables were age of 20 to 29 years (42.0%), evening visits (34.9%), weekends or public holidays (62.6%), single status (89.2%), and visits after contraceptive failure (79.1%). The mean time interval was 7.49 hours, and 77.4% of all patients visited the ED within 12 hours. Patients who received public sex education presented earlier (P < .001). ED visits after nonconsensual sexual incidents represented significantly delayed presentations (P < .001). Regression analysis revealed that both the lack of public education and the occurrence of nonconsensual coitus were associated with incident-to-ED visit intervals of >12 hours. Most patients received emergency contraceptive pill (ECP) within the recommended timeframe. In particular, nationwide school-based public sex education positively affected early ECP access. In contrast, ECP provision was delayed for patients who experienced nonconsensual coitus. Strategies for timely ECP access should account for possible concerns about stigmatization and privacy.

Characterization of seasonal behaviors of naturally occurring radioactive materials in a groundwater borehole using an in situ monitoring system.

Understanding the temporal behaviors of naturally occurring radioactive materials is important for safeguarding groundwater as a secure water resource for drinking, agriculture, and industry usage. This study reports the vertical profiles of 238U concentration and 222Rn activity and the management of in situ monitoring systems during intensive field sampling of a national groundwater-monitoring borehole for seven years (2015-2021). The aim was to capture the seasonal characteristics of the 238U concentrations and 222Rn activity. Both factors were low in the rainy season and high in the winter season, reflecting the dilution effect of rainfall recharge. The 238U and 222Rn behaviors were associated with water-rock interactions of calcite dissolution in fracture zones filled with carbonate minerals. Furthermore, multilayer perceptron models estimated the 238U concentration and 222Rn activity with reasonable regression and classification accuracy. Hydrometeorological indicators (temperature and groundwater-level fluctuations) were more important estimators of 238U concentration and 222Rn activity than geochemical process indicators. The regression accuracy performance was higher at deeper sampling depths, where seasonality in the 238U and 222Rn behaviors dominated. From the predicted distributions of 238U concentrations and 222Rn activities, we could estimate the ranges of 238U concentrations and 222Rn activities emerging from groundwater boreholes. High exposure threats from 238U and 222Rn during groundwater usage were found in the winter season. When the multilayer perceptron models use the entire in situ monitoring data at refined temporal resolution, we can quickly determine the naturally occurring radioactive materials and further develop the national groundwater-monitoring borehole equipped with the in-situ monitoring system, supplementing the occasionally obtained field-measurement data.

Detection of microplastic traces in four different types of municipal wastewater treatment plants through FT-IR and TED-GC-MS.

Large amounts of microplastics are discharged into wastewater treatment plants (WWTPs), from where some of them are released into natural waterbodies on account of their not being fully eliminated by WWTPs. To investigate the behavior and emission of microplastics from WWTPs, we selected four WWTPs with different treatment technologies, including anaerobic-anoxic-aerobic (A2O), sequence batch reactor (SBR), media, and membrane bioreactor (MBR). The number of microplastics detected using Fourier transform infrared (FT-IR) spectroscopy ranged from 520 to 1820 particles/L in influent and from 0.56 to 2.34 particles/L in effluent. The microplastic removal efficiencies of four WWTPs were over 99%, indicating that the type of treatment technologies did not significantly affect the removal rate of microplastics. In the unit process for each WWTP, the major stages relating to microplastic removal were the secondary clarifier and tertiary treatment processes. Most microplastics detected were categorized as fragments and fibers, while other types were hardly detected. The size of more than 80% of microplastic particles detected in WWTPs ranged between 20 and 300 µm, indicating that they were significantly smaller than the size threshold defined for microplastics. Therefore, we used thermal extraction-desorption coupled with gas chromatography-mass spectroscopy (TED-GC-MS) to evaluate the microplastic mass content in all four WWTPs, and the results were compared with those of the FT-IR analysis. In this method, only four components, namely polyethylene, polypropylene, polystyrene, and polyethylene terephthalate, were analyzed because of the analysis limitation, and the total microplastic concentration represented the sum of four components concentrations. The influent and effluent microplastic concentrations estimated by TED-GC-MS ranged from not detectable to 160 µg/L and 0.04-1.07 µg/L, respectively, indicating a correlation coefficient of 0.861 (p < 0.05) between the TED-GC-MS and FT-IR results, when compared to the combined abundance of the four microplastic components by FT-IR analysis.

Exposure to an accidental trichlorosilane spill: three case reports.

Chlorosilane is a hazardous chemical compound which is used as a raw material in the production of silicone. Despite strict restrictions, accidental spillage of chlorosilane is often reported. However, human exposure was rarely reported in the past decades and the effect on humans is barely known. We report cases of human exposure to an accidental trichlorosilane spill. Three middle aged male industrial workers visited our emergency department after exposure to trichlorosilane. They presented with shortness of breath and burns on multiple sites. Chest radiograph and laboratory studies were performed. None of the reports showed serious results and were discharged after conservative management.

Development of diagnostic systems for wide range and highly sensitive detection of two waterborne hepatitis viruses from groundwater using the conventional reverse transcription nested PCR assay.

Waterborne epidemics of human hepatitis virus A and E (HAV and HEV) have been reported worldwide. Molecular biology techniques, such as reverse transcription polymerase chain reaction (RT-PCR), have been widely used to detect the two hepatitis viruses. However, comparative studies of various types of samples are needed, and different environmental factors, including the low copy pathogens, presence of PCR inhibitors in the sample, unknown non-specific reaction with template, and sequence diversity leading to new variants in viruses, should be considered. In addition, standard positive material is required to determine the accuracy of the PCR and should be able to distinguish between false and real positives. In this study, we developed RT-PCR primer sets and optimised standard templates for HAV and HEV detection to address the above concerns associated with test sensitivity and possible PCR inhibition. Finally, previously reported diagnostic methods of HAV and HEV were compared and an applicability test using groundwater was performed. The nested RT-PCR developed in this study is expected to contribute to assess water safety by monitoring HAV and HEV in non-disinfected water, like groundwater.

Local Behavior of Lap-Spliced Deformed Rebars in Reinforced Concrete Beams.

Twelve full-scale reinforced concrete beams with two tension lap splices were constructed and tested under a four-point loading test. Half of these beams had shorter lap splices than that recommended by American Concrete Institute Building Code ACI 318-19; they failed by bond loss between steel and concrete at the lap splice region before rebar yielding. The other half of the beams were designed with a lap splice length slightly exceeding that recommended by ACI 318-19; they failed by rebar yielding and exhibited a ductile behavior. Several strain gauges were attached to the longitudinal bars in the lap splice region to study the local behavior of deformed bars during loading. The strain in a rebar was maximum at the loaded end of the lap splice and progressively decreased toward the unloaded end because the rebar at this end could not sustain any load. Stress flow discontinuity occurred at the loaded end and caused stress concentration. The effect of this concentration was investigated based on test results. The comparison of bond strengths calculated by existing equations and those of tested specimens indicated that the results agreed well.

High Mobility Group Box Chromosomal Protein-1 Induces Myofibroblast Differentiation and Extracellular Matrix Production via RAGE, p38, JNK and AP-1 Signaling Pathways in Nasal Fibroblasts.

BACKGROUND: Chronic rhinosinusitis is involved in myofibroblast differentiation and extracellular matrix (ECM) accumulation. High mobility group box chromosomal protein 1 (HMGB-1) is known to stimulate lung fibroblast to produce ECM in lung fibrosis. The aim of this study was to investigate whether HMGB-1 induces myofibroblast differentiation and ECM production in nasal fibroblasts and to identify the signal pathway. METHODS: Human nasal fibroblasts were cultured. After stimulation with HMGB-1, expressions of α-smooth muscle actin (α-SMA) and fibronectin were determined by real-time PCR and western blot. Total collagen was measured by Sircol assay. To investigate signal pathway, various signal inhibitors and RAGE siRNA were used. RESULTS: HMGB-1 increased α-SMA and fibronectin in mRNA and protein levels. It also increased collagen production. RAGE siRNA inhibited HMGB-1-induced α-SMA and fibronectin, and production of collagen. Furthermore, the inhibitors of RAGE downstream molecules such as p38, JNK and AP-1 also blocked the HMGB-1-induced effects. CONCLUSIONS: HMGB-1 induces myofibroblast differentiation and ECM production in nasal fibroblast, which is mediated by RAGE, p38, JNK and AP-1 signal pathway. These results suggest that HMGB-1 may play an important role in tissue remodeling during chronic rhinosinusitis progression.

Spatial variation of pharmaceuticals in the unit processes of full-scale municipal wastewater treatment plants in Korea.

Several reports have elucidated the removal of pharmaceutical residues in municipal wastewater treatment plants (WWTPs). However, there remains a need to determine the spatial distribution of pharmaceuticals in the unit processes of full-scale municipal WWTPs. Herein, spatial variations of fifteen pharmaceuticals in the unit processes of four full-scale municipal WWTPs were assessed by analyzing both solid and liquid samples. Furthermore, different pathways of each pharmaceutical such as biodegradation, adsorption, deconjugation, and electrostatic interaction were investigated. Pharmaceutical mass loading were measured at various points for the different unit process and evaluated using liquid chromatography-tandem mass spectrometry. The average mass loading of acetaminophen and caffeine decreased tremendously in the first biological treatment process regardless of the process configuration. In contrast, a temporary increase was observed in the mass loading of ibuprofen in the anaerobic and/or anoxic processes, which was presumably caused by deconjugation. Additionally, the adverse effect of coagulation on ibuprofen removal was validated. The major removal mechanism for the selected antibiotics, except for sulfamethoxazole, was the adsorption by biosolids due to electrostatic interaction. Subsequently, a drastic decrease was observed in their mass loadings in the solid-liquid separation process of the WWTPs. The membrane bioreactor (MBR) shows excellent capability for mitigation of pharmaceuticals in municipal wastewater because it comprises a high concentration of biosolids that act as adsorbents. The evaluation of the spatial variations of the selected pharmaceuticals in different unit processes provides valuable information on their behavior and removal mechanisms.

Occurrence of microplastics in the Han River and riverine fish in South Korea.

Microplastic pollution has been paid attention due to the possibly global threat to human health and ecosystem in recent years. In this study, we investigated the distribution of microplastics in the Han River and its tributaries, South Korea, and in six species of inhabiting fish, namely carp (C. carpio), crucian carp (C. cuvieri), bluegill (L. macrochirus), bass (M. salmoides), catfish (S. asotus), and snakehead (C. argus). We found that the concentration of microplastics in the surface waters (0 m) was 0-42.9 particles/m3 (mean: 7.0 ± 12.9 particles/m3) compared to 20.0-180.0 particles/m3 (mean: 102.0 ± 50.3 particles/m3) at a depth of 2 m. Concentrations in the river tributaries ranged from 1.2 to 234.5 particles/m3 (mean: 91.1 ± 72.3 particles/m3). The most common types the plastic identified were polyethylene (PE), silicone, and polystyrene, while polytetrafluoroethylene (PTFE), polyethylene, and polyester dominated in the tributaries. With respect to shape, >73% of the recovered microplastics were fragments and the rest were fibers in the water. We also measured the concentration of microplastics in the intestines of fish, which ranged from 4 to 48 particles/fish (mean: 22.0 ± 16.0 particles/fish). The most common types of plastic found in the sampled fish were polytetrafluoroethylene (PFTE), polyethylene (PE), and rayon, and >94% of all the microplastic found in fish was in the form of fragments with the remainder being fibers. The concentrations of microplastic in the gills of fish ranged from 1 to 16 particles/fish (mean: 8.3 ± 6.0 particles/fish). In contrast, no microplastic was found in the flesh of the sampled fish. Our results imply that the ingestion of microplastics by fish is more closely related to habitat rather than feeding habits.

Development of water quality criteria for arsenic to protect aquatic life based on species sensitivity distribution.

Arsenic is a hazardous environmental pollutant widely distributed globally. Arsenic toxicity is well known and it is regulated by many countries in terms of managing water quality and protecting aquatic organisms. Unfortunately, water quality criterion (WQC) to protect aquatic organisms has not been introduced in Korea yet. Thus, it is of great importance and necessity to introduce WQC to protect aquatic organisms from arsenic, as WQC play a significant role in protecting aquatic ecosystems from pollutants. Therefore, the purpose of this study is to derive arsenic water quality criterion for aquatic life in Korea. Arsenic acute toxicity tests were performed with 10 Korean native aquatic species, which belong to 7 different taxonomic groups. Based on the results of acute toxicity test and additional toxicity data from literature, the species sensitivity distribution (SSD) method was used in ecological risk assessment. The arsenic concentration of 95% protection level for aquatic life was 0.229 mg L-1 in this study. An assessment factor 3 and a background concentration 0.0004 mg L-1 were applied to the concentration value in consideration of the uncertainty of the data and the amount of arsenic natural generation. Consequently, the WQC value derived for arsenic was found to be 0.077 mg L-1. These results will serve as reference values to establish water quality criterion for the protection of aquatic life in Korea.

Effect of MeCP2 on TGF- ß1-induced Extracellular Matrix Production in Nasal Polyp-derived Fibroblasts.

Purpose Methyl-CpG-binding protein 2 (MeCP2), known as a transcriptional regulator, has been suggested to play an important role in myofibroblast differentiation in the lung. The purpose of this study was to investigate the role of MeCP2 in transforming growth factor (TGF)- ß1-induced myofibroblast differentiation and extracellular matrix (ECM) production in nasal polyp-derived fibroblasts (NPDFs). Methods To identify the expression of MeCP2 in nasal polyp tissues, immunohistochemistry staining and Western blot were performed. TGF- ß1-induced NPDFs were treated with 5-azacytidine, a DNA methylation inhibitor, and the expression levels of α-SMA and fibronectin were determined by semiquantitative reverse transcription polymerase chain reaction, immunofluorescent staining, and Western blotting. The total soluble collagen was analyzed by the Sircol collagen assay. MeCP2 silenced by MeCP2-specific small interference ( si) RNA was verified by Western blot. Results The expression levels of MeCP2 increased in nasal polyp tissues compared to normal inferior turbinate tissues. 5-Azacytidine significantly inhibited the expression of α-SMA and fibronectin mRNA in a dose-dependent manner. In addition, 5-azacytidine suppressed collagen production and the expression of MeCP2 in the same manner. The expression levels of a-SMA and collagen production were significantly blocked by MeCP2 silencing in TGF- ß1-induced NPDFs. Conclusions Our data suggest that MeCP2 plays an essential role in TGF- ß1-induced myofibroblast differentiation and ECM production in NPDFs.

Development of water quality criteria of ammonia for protecting aquatic life in freshwater using species sensitivity distribution method.

Ammonia is deemed one of the most important pollutants in the freshwater environment because of its highly toxic nature and ubiquity in surface water. This study thus aims to derive the criteria for ammonia in freshwater to protect aquatic life because there are no water quality criteria for ammonia in Korea. Short-term lethal tests were conducted to perform the species sensitivity distribution (SSD) method. This method is widely used in ecological risk assessment to determine the chemical concentrations to protect aquatic species. Based on the species sensitivity distribution method using Korean indigenous aquatic biota, the hazardous concentration for 5% of biological species (HC5) value calculated in this study was 44mg/L as total ammonia nitrogen (TAN). The value of the assessment factor was set at 2. Consequently, the criteria for ammonia were derived as 22mg/L at pH7 and 20°C. When the derived value was applied to the monitoring data nationwide, 0.51%, 0.09%, 0.18%, 0.20%, and 0.35% of the monitoring sites in Han River, Nakdong River, Geum River, Youngsan River, and lakes throughout the nation, respectively, exceeded this criteria. The Ministry of Environment in Korea has been considering introducing water quality standard of ammonia for protecting aquatic life. Therefore, our results can provide the basis for introducing the ammonia standard in Korea.

Spectroscopic and molecular characterization of humic substances (HS) from soils and sediments in a watershed: comparative study of HS chemical fractions and the origins.

Optical properties and molecular composition of humic substances (HS) can provide valuable information on the sources and the history of the associated biogeochemical processes. In this study, many well-known spectral and molecular characteristics were examined in eight different HS samples, which were extracted from soils and sediments located in a forested watershed, via two advanced tools including fluorescence excitation emission matrix-parallel factor analysis (EEM-PARAFAC) and high-resolution Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS). Two humic-like (C1 and C2) and one protein-like (C3) components were identified from EEM-PARAFAC. Irrespective of the origins, humic acid (HA) fractions were distinguished from fulvic acid (FA) fractions by the HS characteristics of specific UV absorbance (SUVA), the number of formulas, maximum fluorescence intensities of C1 and C2, condensed aromatics, tannins, and CHON, CHOS, and CHONS classes. In contrast, only five HS indices, including C3 intensity, H%, modified aromatic index (AImod), the percentages of carbohydrates, and unsaturated hydrocarbons, were found to be significant factors in discriminating between the two HS origins (i.e., soils and sediments). The ordination of the Bray-Curtis dissimilarity matrix further confirmed that the HS chemical fraction (i.e., HA or FA) was the more important factor to determine the measured HS characteristics than the HS origin. Our results provided an in-depth insight into the chemical and structural heterogeneity of bulk HS, which could be even beyond the differences observed along the two HS origins. This study also delivers a cautious message that the two operationally defined HS chemical fractions should be carefully considered in tracking the origins of different HS samples.

Seasonal and storm-driven changes in chemical composition of dissolved organic matter: a case study of a reservoir and its forested tributaries.

Many drinking water supplies are located in forested watersheds, which operate as an important source of dissolved organic matter (DOM). In this study, monthly sampling campaigns were conducted from a reservoir (Daecheong Reservoir, South Korea) and its forested tributaries for five consecutive months (June to October) to examine the variations of DOM composition. Excitation-emission matrix fluorescence spectroscopy combined with parallel factor analysis (EEM-PARAFAC) was applied to track the variations of different fluorescent components within bulk DOM. Selected samples were further separated into hydrophobic (Ho) and hydrophilic (Hi) fractions. Water quality and DOM composition varied greatly with the sampling locations including the upstream and the downstream tributary sites, and the reservoir. Non-metric multidimensional scaling (NMDS) provided the information on the DOM sources and the potential processes leading to the observed DOM changes. Four of the five fluorescent components, identified by EEM-PARAFAC, were well correlated with the flow rates of the tributaries, suggesting hydrological control on DOM composition. The greatest effects were found on two terrestrial humic-like components (C1 and C2). The Ho fraction of DOM was more abundant for the post-storm samples versus the non-storm samples, supporting the important roles of hydrology on the changes in chemical composition of DOM. The amounts of the DOM resin fractions, either Ho or Hi, showed strong relationships with C1 and C2, suggesting that DOM fluorescence could be successfully applied to estimate different DOM chemical constituents in forested watersheds.