Decision Tree-Based Classifier Combined with Neural-Based Predictor for Water-Stage Forecasts in a River Basin During Typhoons: A Case Study in Taiwan

To solve the complicated problem of water-stage predictions under the interaction of upstream flows and tidal effects during typhoon attacks, this article presents a novel approach to river-stage predictions. The proposed CART-ANN model combines both the decision trees (classification and regression trees [CART]) and the artificial neural network (ANN) techniques, which comprise the multilayer perceptron (MLP) and radial basis function (RBFNN). The combined CART-ANN model involves a two-step predicting process. First, the CART stage-level classifier can classify the river stages into higher, middle, and lower levels. Then, the ANN-based water-stage predictors are employed to predict the water stages. The proposed model was applied to the Tanshui River Basin in Taiwan. The Taipei Bridge, which is close to the estuary and affected by tidal effects, was taken as the study gauge. The mean square error and the mean absolute error were used for evaluating the variance and bias performances of the models. This study makes two contributions. First, the CART-MLP and CART-RBF were modeled to predict river stages under tidal effects during typhoons, and they were compared with three benchmark models, CART, back-propagation neural network, and RBFNN. Second, the CART-RBF successfully demonstrated that it achieved more accurate prediction than CART-MLP and three benchmark models.     

Co-firing of paper mill sludge and coal in an industrial circulating fluidized bed boiler

Co-firing of coal and paper mill sludge was conducted in a 103 MWth circulating fluidized bed boiler to investigate the effect of the sludge feeding rate on emissions of SOx, NOx, and CO. The preliminary results show that emissions of SOx and Nx decrease with increasing sludge feeding rate, but CO shows the reverse tendency due to the decrease in combustion temperature caused by a large amount of moisture in the sludge. All emissions met the local environmental requirements. The combustion ashes could be recycled as feed materials in the cement manufacturing process.     

Simulation modeling for nitrogen removal and experimental estimation of mass fractions of microbial groups in single-sludge system

Nitrification-denitrification in a single-sludge nitrogen removal system (SSNRS; with a sufficient carbon source for denitrification) was performed. With an increase in the mixed liquor recycle ratio (R(m)) from 1 to 2, the total nitrogen (TN) removal efficiency at a lower volumetric loading rate (VLR=0.21 NH(4)(+)-N m(-3) d(-1)) increased, but the TN removal efficiency at a higher VLR (0.35 kg NH(4)(+)-N m(-3) d(-1)) decreased. A kinetic model that accounts for the mass fractions of Nitrosomonas, Nitrobacter, nitrate reducer and nitrite reducer (f(n1), f(n2), f(dn1), and f(dn2)) in the SSNRS and an experimental approach for the estimation of the mass fractions of nitrogen-related microbial groups are also proposed. The estimated f(dn1) plus f(dn2) (0.65-0.83) was significantly larger than the f(n1) plus f(n2) (0.28-0.32); the f(n1) (0.21-0.26) was larger than the f(n2) (0.05-0.07); and the f(dn1) (0.32-0.45) varied slightly with the f(dn2) (0.33-0.38). At the lower VLR, the f(dn1) plus f(dn2) increased with increasing R(m); however at the higher VLR, the f(dn1) plus f(dn2) did not increase with increasing R(m). By using the kinetic model, the calculated residual NH(4)(+)-N and NO(2)(-)-N in the anoxic reactor and NO(2)(-)-N and NO(3)(-)-N in the aerobic reactor were in fairly good agreement with the experimental data; the calculated NO(3)(-)-N in the anoxic reactor was over-estimated and the calculated NH(4)(+)-N in the aerobic reactor was under-estimated.     

Estimation and projection of nitrous oxide (N2O) emissions from anthropogenic sources in Taiwan

Taiwan is a densely populated and developed country with more than 97% of energy consumption supplied by imported fuels. Greenhouse gas emissions are thus becoming significant environmental issues in the country. Using the Intergovernmental Panel on Climate Change (IPCC) recommended methodologies, anthropogenic emissions of nitrous oxide (N2O) in Taiwan during 2000-2003 were estimated to be around 41 thousand metric tons annually. About 87% of N2O emissions come from agriculture, 7% from the energy sector, 3% from industrial processes sector, 3% from waste sector. On the basis of N2O emissions in 2000, projections for the year 2010 show that emissions were estimated to decline by about 6% mainly due to agricultural changes in response to the entry of WTO in 2002. In contrast to projections for the year 2020, N2O emissions were projected to grow by about 17%. This is based on the reasonable scenario that a new adipic acid/nitric acid plant will be probably started after 2010.     

Genetic polymorphisms in CYP1A1 and GSTM1 predispose humans to PCBs/PCDFs-induced skin lesions

INTRODUCTION: Polychlorinated biphenyls (PCBs) and dibenzofurans (PCDFs) are ubiquitous persistent pollutants in humans. Whether people with different genotypes are with different susceptibility to these chemicals are unknown. In a group of people highly exposed to PCBs/PCDFs, we tested the hypothesis that genotypic polymorphisms affected susceptibility for development of skin manifestations. METHODS: In 1979, approximately 2000 people in central Taiwan ingested cooking oil contaminated with PCBs/PCDFs. Skin disorder such as chloracne, abnormal nail, hyperkeratosis and skin allergy were found in PCBs/PCDFs exposed group. We recruited exposed and community background exposure subjects for blood testing and telephone-interview. Single nucleotide polymorphisms, AhR Arg554Lys, CYP1A1 Ile462Val, CYP1A1 T6235C, and GSTM1/T1 deletion, were determined. Occurrence of skin manifestations was compared among people with different genotypes while stratified by PCB exposure levels by logistic regression. RESULTS: Data on exposure, medical history, and genotypes were obtained from 393 exposed and 181 background exposure groups. Skin manifestations including chloracne, allergy, abnormal nail, and hyperkeratosis were more prevalent in exposed people in a dose-related manner. Among highly exposed individuals, combined CYP1A1-MspI mutant genotype and GSTM1-null genotype were associated with increased risk of chloracne (odds ratio 2.8, 95% confidence interval 1.1-7.6). Among intermediately exposed individuals, GSTM1 null genotype was associated with skin allergy. AhR Arg554Lys genotype and GSTT1 null genotype were not related to susceptibility to skin manifestations in PCB/PCDF-exposed population. CONCLUSION: CYP1A1 and GSTM1 genotypic polymorphisms might be related to the susceptibility to PCB/PCDF-induced skin manifestations.     

Direct field observation of the relative humidity effect on the beta-gauge readings

The effect of ambient relative humidity (RH) on hourly particulate matter (PM10) readings of beta-gauge monitors has been studied using two collocated monitors in the field. The inlet air of monitor 1 was conditioned with water vapor to increase its RH, whereas monitor 2 operated normally in ambient conditions. Experimental data showed that PM10 readings of monitor 1 were nearly the same as monitor 2, as long as the RH of its conditioned incoming air did not exceed approximately 80-85%. However, when the RH exceeded approximately 80-85%, PM10 readings of monitor 1 became higher than monitor 2, and the difference increased with increasing RH. The measurement of pressure drop across the filter was also conducted, and the data revealed that the increase of pressure drop per unit of PM10 concentration decreased when RH was higher than approximately 80-85%, as compared with the case when RH was lower than 80-85%. This is perhaps because of more porous structure of deposited particles in the beta-gauge monitor when RH is greater than approximately 80-85%. The theoretical calculation using an evaporation model and a thermodynamic model has been conducted to simulate the beta-gauge readings. The results show that the theoretical PM10 concentrations using the evaporation model are in better agreement with the actual beta-gauge readings than those using the thermodynamic equilibrium model.     

Catalytic hydrodechlorination of 1,2-dichloroethane using copper nanoparticles under reduction conditions of sodium borohydride

1,2-Dichloroethane (1,2-DCA) is a raw material used for the manufacture of vinyl chloride monomer (VCM) and therefore has very often been detected in the groundwater nearby the VCM manufacturing plant. Zero-valent iron (ZVI) is capable of degrading a wide array of highly chlorinated contaminants; however, the reactivity of ZVI towards 1,2-DCA is very low. In this study, zero-valent copper nanoparticles have been synthesized for effective dechlorination of 1,2-DCA under reduction conditions of sodium borohydride. Copper nanoparticles consisted of mainly metallic copper (Cu(0)) with small amounts of cuprous oxide (Cu(2)O). They have surface areas of about 19.0 m(2) g(-1) and an average diameter of 15 nm. Batch experiments were conducted to test the effectiveness of copper nanoparticles for 1,2-DCA degradation using sodium borohydride as electron donors where the ORP was measured as -1100 mV. More than 80% of 1,2-DCA (30 mg L(-1)) was rapidly degraded within 2 h in the presence of both copper nanoparticles (2.5 g L(-1)) and borohydride (25 mM). No reduction of 1,2-DCA was observed when the system contained either copper nanoparticles alone or borohydride alone. The degradation intermediates included ethane and ethylene accounting for 79% and ～1.5% of the 1,2-DCA lost, respectively. Potential environmental applications can be achieved by immobilizing copper nanoparticles onto the surface of reducing metals to form a reactive bimetallic structure.