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(R~2), Nash–Sutcliff efficiency(NSE), relative efficiency criteria(d rel). 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.     

Modelling equilibrium adsorption of single,binary, and ternary combinations of Cu,Pb, and Zn onto granular activated carbon

Environmental Science and Pollution Research - Elevated concentrations of heavy metals in water can be toxic to humans, animals, and aquatic organisms. A study was conducted on the removal of Cu,...     

<Emphasis Type="Italic">Synechococcus</Emphasis> production and grazing loss rates in nearshore tropical waters

Temporal variation of Synechococcus, its production (μ) and grazing loss (g) rates were studied for 2 years at nearshore stations, i.e. Port Dickson and Port Klang along the Straits of Malacca. Synechococcus abundance at Port Dickson (0.3–2.3 × 10⁵ cell ml^?1) was always higher than at Port Klang (0.3–7.1 × 10⁴ cell ml^?1) (p < 0.001). μ ranged up to 0.98 day^?1 (0.51 ± 0.29 day^?1), while g ranged from 0.02 to 0.31 day^?1 (0.15 ± 0.07 day^?1) at Port Klang. At Port Dickson, μ and g averaged 0.47 ± 0.13 day^?1 (0.29–0.82 day^?1) and 0.31 ± 0.14 day^?1 (0.13–0.63 day^?1), respectively. Synechococcus abundance did not correlate with temperature (p > 0.25), but nutrient and light availability were important factors for their distribution. The relationship was modelled as log Synechococcus = 0.37Secchi ? 0.01DIN + 4.52 where light availability (as Secchi disc depth) was a more important determinant. From a two-factorial experiment, nutrients were not significant for Synechococcus growth as in situ nutrient concentrations exceeded the threshold for saturated growth. However, light availability was important and elevated Synechococcus growth rates especially at Port Dickson (F = 5.94, p < 0.05). As for grazing loss rates, they were independent of either nutrients or light intensity (p > 0.30). In nearshore tropical waters, an estimated 69 % of Synechococcus production could be grazed.     

Determination of secondhand smoke leakage from the smoking room of an Internet café

Although Internet cafes have been designated as nonsmoking areas in Korea, smoke-free legislation has allowed the installation of indoor smoking rooms. The purposes of this study were to determine secondhand smoke (SHS) leakage from an Internet café smoking room and to identify factors associated with SHS leakage. PM_2.5 (particulate matter with an aerodynamic diameter ≤2.5 μm) mass concentrations were measured simultaneously both inside and outside the door to the smoking room. During each measurement, a field technician observed how long the smoking room door was opened and closed, the direction of door opening, and the number of smokers. A multivariate linear regression model was used to identify the causality of SHS leakage from the smoking room. A time series of PM_2.5 concentrations both inside and outside the door to the smoking room showed a similar trend. SHS leakage was significantly increased because of factors associated with the direction of the smoking room door being opened, the duration of how long the smoking room door was opened until it was closed, and the average PM_2.5 concentration inside the smoking room when the door was opened. SHS leakage from inside the smoking room to outside the smoking room was evident especially when the smoking room door was opened. Since the smoking room is not effective in preventing SHS exposure, the smoking room should be removed from the facilities to protect citizens from SHS exposure through revision of the current legislation, which permits installation of a smoking room.

Implications: This paper concerns secondhand smoke (SHS) leakage from indoor smoking room. Unlike previous studies, the authors statistically analyzed the causality of PM_2.5 concentration leakage from a smoking room using time-series analysis. Since the authors selected the most common smoking room, the outcomes could be generalized. The study demonstrated that SHS leakage from smoking room and SHS leakage were clearly associated with door opening. The finding demonstrated ineffectiveness of smoking room to protect citizens and supports removal of indoor smoking room.     

A resilience-based approach for comparing expert preferences across two large-scale coastal management programs

This report proposes a method for assessing resilience-building components in coastal social–ecological systems. Using the proposed model, the preferences of experts in Masan Bay (South Korea) and Puget Sound (USA) are compared. A total of 30 management objectives were determined and used to build a hierarchic tree designed using the principles of the Analytic Hierarchy Process (AHP). Surveys were performed with 35 Puget Sound experts using face-to-face interviews and with 28 Masan Bay experts by mail. The results demonstrate that the legal objective, which enables legislation, was the highest preferred component in both regions. The knowledge translation variable was also given a high preference score in both regions. An analysis of variance (ANOVA) showed that the Puget Sound experts significantly favored attention to education, habitat restoration and species protection objectives in comparison to the Masan Bay experts. The Masan Bay experts placed greater emphasis on legislation and the type of institutional design than did the Puget Sound experts. Using cluster analysis, four distinct groups of respondents were independently identified in Puget Sound and three groups were identified in Masan Bay. One unique subgroup in the Puget Sound experts group, which was characterized by its high preferences for habitat restoration and species protection, was not observed in Masan Bay. Demographic variables (length of career and role in coastal issue) failed to account for the differences in groupings and preferences in either region, except for the variable ‘favoring information source’ in the Puget Sound group. This finding implies that the demographic information was not related to differences in group opinions in both regions. The analysis framework presented here was effective in identifying expert preferences regarding the overall structure and emphasis in coastal management programs. Thus, this framework can be applied towards coastal policy development.     

Development of a Local-Scale Spatially Refined Multimedia Fate Model (LSRMFM) for Urban-Scale Risk Assessment: Model Formulation, GIS-Based Preprocessing, and Case Study

A local-scale spatially refined multimedia fate model (LSRMFM) was developed to evaluate in detail the multimedia transport of organic compounds at a spatial level. The model was derived using a combination of an advection?Cdispersion?Creaction partial differential equation, a steady-state multimedia fugacity model, and a geographical information system. The model was applied to predicting four major volatile organic compounds that are produced as emissions (benzene, toluene, xylene, and styrene) in an urban and industrial area (the 50?×?50-km area was divided into 0.5?×?0.5-km segments) in Korea. To test the accuracy of the model, the LSRMFM was used to predict the extent of dispersion and the data compared with actual measured concentrations and the results of a generic multimedia fate model (GMFM). The results indicated that the method developed herein is appropriate for predicting long-term multimedia pollution. However, the comparison study also illustrated that the developed model has some limitations (e.g., steady-state assumption) in terms of explaining all the observed concentrations, and additional verification and study (e.g., validation using a large observed data set, integration with a more accurate runoff model) would be desirable. In comparing LSRMFM and GMFM, discrepancies between the LSRMFM and GMFM outputs were found, as the result of geographical effects, even though the environmental parameters were identical. The geographical variation for LSRMFM output indicated the existence of considerable local human and ecological risks, whereas the GMFM output indicated less average risk. These results demonstrate that the model has the potential for improving the management of pollutant levels under these refined spatial conditions.