Prediction of effluent concentration in a wastewater treatment plant using machine learning models.

Of growing amount of food waste, the integrated food waste and waste water treatment was regarded as one of the efficient modeling method. However, the load of food waste to the conventional waste treatment process might lead to the high concentration of total nitrogen (T-N) impact on the effluent water quality. The objective of this study is to establish two machine learning models-artificial neural networks (ANNs) and support vector machines (SVMs), in order to predict 1-day interval T-N concentration of effluent from a wastewater treatment plant in Ulsan, Korea. Daily water quality data and meteorological data were used and the performance of both models was evaluated in terms of the coefficient of determination (R2), Nash-Sutcliff efficiency (NSE), relative efficiency criteria (drel). Additionally, Latin-Hypercube one-factor-at-a-time (LH-OAT) and a pattern search algorithm were applied to sensitivity analysis and model parameter optimization, respectively. Results showed that both models could be effectively applied to the 1-day interval prediction of T-N concentration of effluent. SVM model showed a higher prediction accuracy in the training stage and similar result in the validation stage. However, the sensitivity analysis demonstrated that the ANN model was a superior model for 1-day interval T-N concentration prediction in terms of the cause-and-effect relationship between T-N concentration and modeling input values to integrated food waste and waste water treatment. This study suggested the efficient and robust nonlinear time-series modeling method for an early prediction of the water quality of integrated food waste and waste water treatment process.

Multi-level stressor analysis from the DNA/biochemical level to community levels in an urban stream and integrative health response (IHR) assessments.

The objectives of this study were to identify multi-level stressors at the DNA/biochemical level to the community level in fish in an urban stream and to develop an integrative health response (IHR) model for ecological health diagnosis. A pristine control site (S (c) ) and an impacted site (S (i) ) were selected from among seven pre-screened sites studied over seven years. Various chemical analyses indicated that nutrient enrichment (Nitrogen, Phosphorus) and organic pollution were significantly greater (t > 8.783, p < 0.01) at the S (i) site compared to the S (c) site. Single-cell gel electrophoresis (comet assays) of DNA-level impairment indicated significantly (t = 5.678, p < 0.01) greater tail intensity, expressed as % tail-DNA, at the S (i) site and genotoxic responses were detected in the downstream reach. Ethoxyresorufin-O-deethylase (EROD) assays, as a physiological bioindicator, were 2.8-fold higher (p < 0.05, NK-test after ANOVA) at the S (i) site. Tissue analysis using a necropsy-based health assessment index (NHAI) showed distinct internal organ disorders in three tissues, i.e., liver, kidney, and gill, at the S (i) site. Population-level analysis using the sentinel species Zacco platypus showed that the regression coefficient (b) was 3.012 for the S (i) site and 2.915 for the S (c) site, indicating population skewness in the downstream reach. Community-level health was impaired at the S (i) site based on an index of biological integrity (IBI), and physical habitat modifications were identified by a qualitative habitat evaluation index (QHEI). Overall, the model values for the integrative health response (IHR), developed using the star plot approach, were 3.22 (80.5%) at the S (c) site and 0.74 (18.5%) at the S (i) site, indicating that, overall, ecological health impairments were evident in the urban reach. Our study was based on multi-level approaches using biological organization and the results suggest that there is a pivotal point of linkage between mechanistic understanding and real ecological consequences of environmental stressors.

Optimization of the Design Configuration and Operation Strategy of Single-Pass Seawater Reverse Osmosis.

The numerical study was conducted to compare process performance depending on the pump type and process configuration. The daily monitoring data of seawater temperature and salinity offshore from Daesan, Republic of Korea was used to reflect the site-specific seawater conditions. An algorithm for reverse osmosis in constant permeate mode was developed to simulate the process in time-variant conditions. Two types of pumps with different maximum leachable efficiencies were employed to organize pump-train configuration: separated feed lines and common pressure center design. The results showed pump type and design configuration did not have a significant effect on process performance. The annual means of specific energy consumption (SEC) for every design configuration were under 2 kWh/m3, except for a worst-case. The worst-case was decided when the pump was operated out of the best operation range. The two operation strategies were evaluated to determine the optimal configuration. The permeate flow rate was reduced to 80% of the designed permeate flow rate with two approaches: feed flow rate reduction in every train and pump shutdown in a specific train. The operation mode with feed flow rate reduction was more efficient than the other. The operating pressure reduction led to a decrease in SEC.

Prevalence of season-specific Escherichia coli strains in the Yeongsan River Basin of South Korea.

Seasonal and spatial variation in the genotypic richness of 3480 Escherichia coli isolates obtained from the Yeongsan River basin in South Korea was investigated by using the horizontal fluorophore-enhanced rep-PCR (HFERP) DNA fingerprinting technique. The relationship between 60 E. coli isolates from each of 58 freshwater samples was determined by using multidimensional scaling (MDS) analysis and self-organized maps (SOMs). The MDS analysis, done based on HFERP DNA fingerprints, showed that E. coli isolates obtained in October through December clustered tightly, while those obtained in other sampling periods were more genetically diverse. However, site-specific E. coli genotypes were not observed. SOMs analysis, done using the 10 most frequently isolated E. coli genotypes, showed the occurrence of season-specific E. coli genotypes and the main SOMs clusters were most influenced by temperature, strain diversity and biochemical oxygen demand. Diversity among E. coli genotypes tended to decrease as water temperature decreased, and the numbers of E. coli genotypes observed in urban area were greater, more diverse and less dependent on water temperature than those obtained from agricultural areas. Taken together, our findings indicate that that an ecological approach needs to be considered in order to obtain a better understanding of E. coli community dynamics in the environment and that SOMs analysis is useful to visualize the multidimensional dependent variables that are influencing the types and dynamics of specific E. coli genotypes in the environment.

Theoretical Analysis of a Mathematical Relation between Driving Pressures in Membrane-Based Desalting Processes.

Osmotic and hydraulic pressures are both indispensable for operating membrane-based desalting processes, such as forward osmosis (FO), pressure-retarded osmosis (PRO), and reverse osmosis (RO). However, a clear relation between these driving pressures has not thus far been identified; hence, the effect of change in driving pressures on systems has not yet been sufficiently analyzed. In this context, this study formulates an actual mathematical relation between the driving pressures of membrane-based desalting processes by taking into consideration the presence of energy loss in each driving pressure. To do so, this study defines the pseudo-driving pressures representing the water transport direction of a system and the similarity coefficients that quantify the energy conservation rule. Consequently, this study finds three other theoretical constraints that are required to operate membrane-based desalting processes. Furthermore, along with the features of the similarity coefficients, this study diagnoses the commercial advantage of RO over FO/PRO and suggests desirable optimization sequences applicable to each process. Since this study provides researchers with guidelines regarding optimization sequences between membrane parameters and operational parameters for membrane-based desalting processes, it is expected that detailed optimization strategies for the processes could be established.

Use of barcoded pyrosequencing and shared OTUs to determine sources of fecal bacteria in watersheds.

While many current microbial source tracking (MST) methods rely on the use of specific molecular marker genes to identify sources of fecal contamination, these methods often fail to determine all point and nonpoint contributors of fecal inputs into waterways. In this study, we developed a new library-dependent MST method that uses pyrosequencing-derived shared operational taxonomy units (OTUs) to define sources of fecal contamination in waterways. A total 56,841 pyrosequencing reads of 16S rDNA obtained from the feces of humans and animals were evaluated and used to compare fecal microbial diversity in three freshwater samples obtained from the Yeongsan river basin in Jeonnam Province, South Korea. Sites included an urbanized agricultural area (Y1) (Escherichia coli counts ≥ 1600 CFU/100 mL), an open area (Y2) with no major industrial activities (940 CFU/100 mL), and a typical agricultural area (Y3) (≥ 1600 CFU/100 mL). Data analyses indicated that the majority of bacteria in the feces of humans and domesticated animals were comprised of members of the phyla Bacteroidetes or Firmicutes, whereas the majority of bacteria in wild goose feces and freshwater samples were classified to the phylum Proteobacteria. Analysis of OTUs shared between the fecal and environmental samples suggested that the potential sources of the fecal contamination at the sites were of human and swine origin. Quantification of fecal contamination was also examined by comparing the density of pyrosequencing reads in each fecal sample within shared OTUs. Taken together, our results indicated that analysis of shared OTUs derived from barcoded pyrosequencing reads provide the necessary resolution and discrimination to be useful as a next generation platform for microbial source tracking studies.

Factors dominating stratification cycle and seasonal water quality variation in a Korean estuarine reservoir.

A comprehensive monitoring program was conducted during 2005-2007 to investigate seasonal variations of hydrologic stability and water quality in the Yeongsan Reservoir (YSR), located at the downstream end of the Yeongsan River, Korea. A principal component analysis (PCA) was performed to identify factors dominating the seasonal water quality variation from a large suite of measured data--11 physico-chemical parameters from 48 sampling sites. The results showed that three principal components explained approximately 62% of spatio-seasonal water quality variation, which are related to stratifications, pollutant loadings and resultant eutrophication, and the advective mixing process during the episodic rainfall-runoff events. A comparison was then made between YSR and an upstream freshwater reservoir (Damyang Reservoir, DYR) in the same river basin during an autumn season. It was found that the saline stratification and pollutant input from the upstream contributed to greater concentrations of nutrients and organic matter in YSR compared to DYR. In YSR, saline stratification in combination with thermal stratification was a dominant cause of the longer period (for two consecutive seasons) of hypoxic conditions at the reservoir bottom. The results presented here will help better understand the season- and geography-dependent characteristics of reservoir water quality in Asian Monsoon climate regions such as Korea.

Spatial and temporal variability of fecal indicator bacteria in an urban stream under different meteorological regimes.

As a representative urban stream in Korea, the Gwangju (GJ) stream suffers from chronic fecal contamination. In this study, to characterize levels of fecal pollution in the GJ stream, the monthly monitoring data for seven years (from 2001 to 2007) and the hourly monitoring data from two field experiments were examined with respect to seasonal/daily variations and spatial distribution under wet and dry weather conditions. This research revealed that concentrations of fecal indicator bacteria strongly varied depending on the prevalent meteorological conditions. That is, during the dry daytime, fecal indicator bacteria concentrations decreased due to inactivation from solar irradiation, but rapidly increased in the absence of sunlight, suggesting external source inputs. In addition, bacterial concentrations substantially increased during rainfall events, due probably to a major contribution from combined sewer overflow. The observations in this study can be useful for implementing fecal pollution management strategies and for predicting fecal contamination as a function of meteorological conditions.

Evaluation of the relationship between two different methods for enumeration fecal indicator bacteria: colony-forming unit and most probable number.

Most probable number (MPN) and colony-forming unit (CFU) estimates of fecal indicator bacteria (FIB) concentration are common measures of water quality in aquatic environments. Thus, FIB intensively monitored in Yeongsan Watershed in an attempt to compare two different methods and to develop a statistical model to convert from CFU to MPN estimates or vice versa. As a result, the significant difference was found in the MPN and CFU estimates. The enumerated Escherichia coli concentrations in MPN are greater than those in CFU, except for the measurement in winter. Especially in fall, E. coli concentrations in MPN are one order of magnitude greater than that in CFU. Contrarily, enterococci bacteria in MPN are lower than those in CFU. However, in general, a strongly positive relationship are found between MPN and CFU estimates. Therefore, the statistical models were developed, and showed the reasonable converting FIB concentrations from CFU estimates to MPN estimates. We expect this study will provide preliminary information towards future research on whether different analysis methods may result in different water quality standard violation frequencies for the same water sample.

Absence of Escherichia coli phylogenetic group B2 strains in humans and domesticated animals from Jeonnam Province, Republic of Korea.

Multiplex PCR analyses of DNAs from genotypically unique Escherichia coli strains isolated from the feces of 138 humans and 376 domesticated animals from Jeonnam Province, South Korea, performed using primers specific for the chuA and yjaA genes and an unknown DNA fragment, TSPE4.C2, indicated that none of the strains belonged to E. coli phylogenetic group B2. In contrast, phylogenetic group B2 strains were detected in about 17% (8 of 48) of isolates from feces of 24 wild geese and in 3% (3 of 96) of isolates obtained from the Yeongsan River in Jeonnam Province, South Korea. The distribution of E. coli strains in phylogenetic groups A, B1, and D varied depending on the host examined, and there was no apparent seasonal variation in the distribution of strains in phylogenetic groups among the Yeongsan River isolates. The distribution of four virulence genes (eaeA, hlyA, stx(1), and stx(2)) in isolates was also examined by using multiplex PCR. Virulence genes were detected in about 5% (38 of 707) of the total group of unique strains examined, with 24, 13, 13, and 9 strains containing hlyA, eaeA, stx(2), and stx(1), respectively. The virulence genes were most frequently present in phylogenetic group B1 strains isolated from beef cattle. Taken together, results of these studies indicate that E. coli strains in phylogenetic group B2 were rarely found in humans and domesticated animals in Jeonnam Province, South Korea, and that the majority of strains containing virulence genes belonged to phylogenetic group B1 and were isolated from beef cattle. Results of this study also suggest that the relationship between the presence and types of virulence genes and phylogenetic groupings may differ among geographically distinct E. coli populations.

Determination of the optimal parameters in regression models for the prediction of chlorophyll-a: a case study of the Yeongsan Reservoir, Korea.

Statistical regression models involve linear equations, which often lead to significant prediction errors due to poor statistical stability and accuracy. This concern arises from multicollinearity in the models, which may drastically affect model performance in terms of a trade-off scenario for effective water resource management logistics. In this paper, we propose a new methodology for improving the statistical stability and accuracy of regression models, and then show how to cope with pitfalls in the models and determine optimal parameters with a decreased number of predictive variables. Here, a comparison of the predictive performance was made using four types of multiple linear regression (MLR) and principal component regression (PCR) models in the prediction of chlorophyll-a (chl-a) concentration in the Yeongsan (YS) Reservoir, Korea, an estuarine reservoir that historically suffers from high levels of nutrient input. During a 3-year water quality monitoring period, results showed that PCRs could be a compact solution for improving the accuracy of the models, as in each case MLR could not accurately produce reliable predictions due to a persistent collinearity problem. Furthermore, based on R(2) (goodness of fit) and F-overall number (confidence of regression), and the number of explanatory variables (R-F-N) curve, it was revealed that PCR-F(7) was the best model among the four regression models in predicting chl-a, having the fewest explanatory variables (seven) and the lowest uncertainty. Seven PCs were identified as significant variables, related to eight water quality parameters: pH, 5-day biochemical oxygen demand, total coliform, fecal indicator bacteria, chemical oxygen demand, ammonia-nitrogen, total nitrogen, and dissolved oxygen. Overall, the results not only demonstrated that the models employed successfully simulated chl-a in a reservoir in both the test and validation periods, but also suggested that the optimal parameters should cautiously be considered in the design of regression models.

Factors affecting metal exchange between sediment and water in an estuarine reservoir: a spatial and seasonal observation.

Water quality response in a reservoir has often been assessed using relatively restricted datasets that cannot provide sufficient information, thereby giving rise to a dramatic over- or underestimate of actual figures. In this paper we discuss how the levels of metallic elements between the sediment and overlying water in an estuarine reservoir can be influenced by aquatic parameters in response to spatial and seasonal conditions. To better elucidate the interfacial exchange between sediment and water, statistical analyses are employed to intensive data sets collected from the Yeongsan Reservoir (YSR), Korea, which has undergone widespread deterioration in water quality due to the continuous growth of anthropogenic sources. During three seasonal sampling campaigns, we found that oxygen deficiency at the bottom water layer promotes Fe and Ni accumulation in sediment, likely due to the formation of sulfide and oxide complexes under anoxic and suboxic environments, respectively. In addition, salinity levels as high as 11 per thousand in the bottom water layer during autumn substantially increase the release of Mn, restricting the use of YSR as a primary source of agricultural irrigation water. Although most dissolved metals are at acceptable levels for sustaining aquatic life, it is recommended that for long-term planning the elevated Fe and Mn levels in sediment should be controlled with oxygen deficiency during dry weather to ensure a sustainable water supply or, at a minimum, better coordinated operation of YSR.

An analytical model for non-conservative pollutants mixing in the surf zone.

Accurate simulation of the surf zone is a prerequisite to improve beach management as well as to understand the fundamentals of fate and transport of contaminants. In the present study, a diagnostic model modified from a classic solute model is provided to illuminate non-conservative pollutants behavior in the surf zone. To readily understand controlling processes in the surf zone, a new dimensionless quantity is employed with index of kappa number (K, a ratio of inactivation rate to transport rate of microbial pollutant in the surf zone), which was then evaluated under different environmental frames during a week simulation period. The sensitivity analysis showed that hydrodynamics and concentration gradients in the surf zone mostly depend on n (number of rip currents), indicating that n should be carefully adjusted in the model. The simulation results reveal, furthermore, that large deviation typically occurs in the daytime, signifying inactivation of fecal indicator bacteria is the main process to control surf zone water quality during the day. Overall, the analytical model shows a good agreement between predicted and synthetic data (R(2) = 0.51 and 0.67 for FC and ENT, respectively) for the simulated period, amplifying its potential use in the surf zone modelling. It is recommended that when the dimensionless index is much larger than 0.5, the present modified model can predict better than the conventional model, but if index is smaller than 0.5, the conventional model is more efficient with respect to time and cost.

Effect of environmental flow management on river water quality: a case study at Yeongsan River, Korea.

This paper describes a management scheme to control river water quality using additional water discharges from upstream dams, which results in an increase environmental flow (EF) followed by an enhancement of water quality in a target river. To suggest a creditable management plan among a suite of ideal scenarios, the monthly averaged water quality monitoring data from 2001 to 2006 at the Yeongsan (YS) River, Korea were investigated with respect to seasonal variation and spatial distribution. From the analysis, it was found that while biochemical oxygen demand (BOD(5)) level in the YS River was extremely high during the dry/drought season (April, May, and June; AMJ), the level was subsequently decreased during the monsoon season (July, August, and September; JAS) due mainly to the dilution effect of rainfall. To improve the water quality in AMJ, we here suggested a scenario of increasing EF using surplus water discharges from upstream dams, which was examined by one dimensional riverine water quality model, QUAL2E model. Simulation result showed that additional discharge from the upstream dams could lead, on average, to a 36% of water quality improvement in mainstream with respect to BOD(5). Model coefficients were validated by comparing the six year monitoring data to minimize a sum of squares error, and showed a good agreement with the observed data. Overall, the methodology developed in this paper appears to be quite clear and straightforward, and thus, can be applied to a wide range of the flow managements or water quality controls in a stream with artificial structures.

Evaluation of pollutants removal efficiency to achieve successful urban river restoration.

Greater efforts to provide alternative scenarios are key to successful urban stream restoration planning. In this study, we discuss two different aspects of water quality management schemes, biodegradation and human health, which are incorporated in the restoration project of original, pristine condition of urban stream at the Gwangju (GJ) Stream, Korea. For this study, monthly monitoring of biochemical oxygen demand (BOD(5)) and fecal indicator bacteria (FIB) data were obtained from 2003 to 2008 and for 2008, respectively, and these were evaluated to explore pollutant magnitude and variation with respect to space and time window. Ideal scenarios to reduce target pollutants were determined based on their seasonal characteristics and correlations between the concentrations at a water intake and discharge point, where we suggested an increase of environmental flow and wetland as pollutants reduction drawing for BOD(5) and FIB, respectively. The scenarios were separately examined by the Qual2E model and hypothetically (but planned) constructed wetland, respectively. The results revealed that while controlling of the water quality at the intake point guaranteed the lower pollution level of BOD(5) in the GJ Stream, a wetland constructed at the discharge point may be a promising strategy to mitigate mass loads of FIB. Overall, this study suggests that a combination of the two can be plausible scenarios not only to support sustainable urban water resources management, but to enhance a quality of urban stream restoration assignment.

Statistical assessment for spatio-temporal water quality in Angkor, Cambodia.

Comprehensive water quality monitoring was conducted to assess the water quality conditions and to determine the impact of urban infrastructure on ambient water quality in Angkor, Cambodia. During this study, surface water, groundwater, and sediment samples were collected for two distinctive seasons in 2006-2007 at 58 monitoring sites along and near the Siem Reap River, in Tole Sap Lake (TSL), and West Baray, the primary water resources in this region. To assess the seasonal and spatial variability of 27 water quality parameters, multivariate analysis of variance, hierarchical cluster analysis, and the Kruskal-Wallis test were conducted using the obtained data. Differences and relationships between the surface water and groundwater were also investigated using t-test and correlation analysis, respectively. The results of these tests showed that the bacterial indicators need special attention as the urban infrastructure of the downtown area caused increased levels of these bacterial indicators in both surface water and groundwater. However, for most parameters, though surface water showed strong seasonal variations, groundwater presented relatively stable conditions between seasons (p > 0.05) with site-specific geochemical conditions. Sediment quality illustrated that pollution levels of 10 trace metals were the highest in TSL because of its unique characteristic (river with backward flow), but did not reflect any potential enrichment from urban development. Overall, the results reveal that while the urban infrastructure in this region has not significantly affected most of the water quality parameters, bacteria and coliphages are still a main concern due to their contributions in widespread waterborne diseases. Thus, careful mitigation plans for reducing each pollutant source are needed in the Angkor area.

A new methodology for determining dispersion coefficient using ordinary and partial differential transport equations.

The present study proposes a methodology for determining the effective dispersion coefficient based on the field measurements performed in Gwangju (GJ) Creek in South Korea which is environmentally degraded by the artificial interferences such as weirs and culverts. Many previous works determining the dispersion coefficient were limited in application due to the complexity and artificial interferences in natural stream. Therefore, the sequential combination of N-Tank-In-Series (NTIS) model and Advection-Dispersion-Reaction (ADR) model was proposed for evaluating dispersion process in complex stream channel in this study. The series of water quality data were intensively monitored in the field to determine the effective dispersion coefficient of E. coli in rainy day. As a result, the suggested methodology reasonably estimates the dispersion coefficient for GJ Creek with 1.25 m(2)/s. Also, the sequential combined method provided Number of tank-Velocity-Dispersion coefficient (NVD) curves for convenient evaluation of dispersion coefficient of other rivers or streams. Comparing the previous studies, the present methodology is quite general and simple for determining the effective dispersion coefficients which are applicable for other rivers and streams.

Interpretation of seasonal water quality variation in the Yeongsan Reservoir, Korea using multivariate statistical analyses.

The Yeongsan (YS) Reservoir is an estuarine reservoir which provides surrounding areas with public goods, such as water supply for agricultural and industrial areas and flood control. Beneficial uses of the YS Reservoir, however, are recently threatened by enriched non-point and point source inputs. A series of multivariate statistical approaches including principal component analysis (PCA) were applied to extract significant characteristics contained in a large suite of water quality data (18 variables monthly recorded for 5 years); thereby to provide the important phenomenal information for establishing effective water resource management plans for the YS Reservoir. The PCA results identified the most important five principal components (PCs), explaining 71% of total variance of the original data set. The five PCs were interpreted as hydro-meteorological effect, nitrogen loading, phosphorus loading, primary production of phytoplankton, and fecal indicator bacteria (FIB) loading. Furthermore, hydro-meteorological effect and nitrogen loading could be characterized by a yearly periodicity whereas FIB loading showed an increasing trend with respect to time. The study results presented here might be useful to establish preliminary strategies for abating water quality degradation in the YS Reservoir.

Development of a new biomonitoring method to detect the abnormal activity of Daphnia magna using automated Grid Counter device.

The aim of the present study was to develop a biological early warning system (BEWS), equipped with six monitoring channels to individually observe the activity of Daphnia magna, using a digital 'Grid Counter', which would trigger an alarm within an appropriate time, and examine the functional performance of the newly developed BEWS for detecting unusual water quality. In order to detect the changes in the activity of D. magna, six relative activity parameter values (Z(a)) were computed from the 6 individual monitoring channels; with the activity data for D. magna calculated every 5 min. The Student's t-test was used to verify the difference between the mean value of the system in a steady state, as a control, and the exposure values during a sudden pollution event. The test results illustrate that the threshold value for the alarm can be at p=0.0093 for 3 consecutive detections. The time period, defined as the average time taken from the detection of hyper to retarded activity of the organism, for Cu concentrations of 50, 100, 200 and 400 ppb were 7.17+/-1.75, 3.94+/-2.02, 1.85+/-0.49 and 1.00+/-0.18 h, respectively. Based on the results of this study, it is proposed that p values from the t-test, with Z(a), are more accurate, stable and predictable parameters for the detection of chemical exposures than the other values, such as the swimming speed and trajectory, etc. Consequently, it would be possible to reduce the number of false alarms and achieve confidence for a system, with the ability of highly accurate detection, such as with the six-channel monitoring system developed in this study.

Perchlorate assessment of the Nakdong and Yeongsan watersheds, Republic of Korea.

The objective of the present study was to conduct a preliminary assessment for perchlorate in surface water, drinking water, and wastewater treatment plant effluent samples obtained from the Nakdong and Yeongsan watersheds in the Republic of Korea. Samples were analyzed for perchlorate using ion chromatography with suppressed conductivity detection (IC-CD) and/or liquid chromatography with tandem mass spectrometry detection (LC-MS/MS). Method reporting limits were 5.0 microg/L for IC-CD and 0.05 microg/L for LC-MS/MS analysis. At perchlorate levels above 5.0 microg/L, IC-CD and LC-MS/MS provided comparable results. The levels of perchlorate detected in the samples procured from the Yeongsan watershed were < 5.0 microg/L in each case. However, Nakdong watershed samples contained perchlorate at levels up to 60 microg/L. The highest concentrations of perchlorate were found in surface water samples, although drinking water contained perchlorate at concentrations up to 35 microg/L. In a subset of samples analyzed by LC-MS/MS, chlorate and bromate also were detected at concentrations ranging from < 0.10 to 44 microg/L and < 0.10 to 2.6 microg/L, respectively. To the best of the authors' knowledge, this is the first perchlorate assessment reported for water sources in the Republic of Korea.