Machine-learning paradigms for selecting ecologically significant input variables

Harmful algal blooms, which are considered a serious environmental problem nowadays, occur in coastal waters in many parts of the world. They cause acute ecological damage and ensuing economic losses, due to fish kills and shellfish poisoning as well as public health threats posed by toxic blooms. Recently, data-driven models including machine-learning (ML) techniques have been employed to mimic dynamics of algal blooms. One of the most important steps in the application of a ML technique is the selection of significant model input variables. In the present paper, we use two extensively used ML techniques, artificial neural networks (ANN) and genetic programming (GP) for selecting the significant input variables. The efficacy of these techniques is first demonstrated on a test problem with known dependence and then they are applied to a real-world case study of water quality data from Tolo Harbour, Hong Kong. These ML techniques overcome some of the limitations of the currently used techniques for input variable selection, a review of which is also presented. The interpretation of the weights of the trained ANN and the GP evolved equations demonstrate their ability to identify the ecologically significant variables precisely. The significant variables suggested by the ML techniques also indicate chlorophyll-a (Chl-a) itself to be the most significant input in predicting the algal blooms, suggesting an auto-regressive nature or persistence in the algal bloom dynamics, which may be related to the long flushing time in the semi-enclosed coastal waters. The study also confirms the previous understanding that the algal blooms in coastal waters of Hong Kong often occur with a life cycle of the order of 1–2 weeks.     

饮用水中天然有机物的分析与表征方法

天然有机物(NOM)是一类分布广泛、成分复杂的有机混合物,可严重影响饮用水的处理效率,并可能对出水水质产生严重的不利影响。因此,为了改进和优化饮用水的处理过程,在研究和实际应用阶段对NOM进行表征和定量以了解其结构特性和反应性具有重要意义。文中先对NOM的富集方法进行了介绍,接着梳理了不同分离方法的依据及作用,最后总结了各类表征方法的原理及应用,以期有助于对NOM和饮用水处理的深入研究。     

Volatilisation and catalytic effects of alkali and alkaline earth metallic species during the pyrolysis and gasification of Victorian brown coal. Part VII. Raman spectroscopic study on the changes in char structure during the catalytic gasification in air

FT-IR/Raman spectroscopies have been used to identify the structural features of Victorian brown coal chars during the gasification in air at 400 °C. The deconvolution of the Raman spectra has allowed us to identify the main structural sites in char where preferential reaction with O₂ takes place. The presence of Na and Ca catalysts is shown to alter the reaction pathways between char and O₂. In the absence of a catalyst, the O-containing functional groups formed in char during gasification were closely associated with the aromatic structure and thus tended to loosen the aromatic structure. The non-catalysed gasification was slow and took place on some specific (especially sp³-rich or sp²–sp³ mixture) sites distributed throughout the char. In the presence of a catalyst (Na or Ca), the O-containing functional groups were not closely associated with the main aromatic structure throughout the char. The catalytic gasification reactions were localised on the sites associated with the catalysts. The preferential removal of smaller aromatic ring systems and the persistence of cross-linking structures in the presence of a catalyst mean that the large aromatic ring systems were increasingly concentrated with little flexibility, affecting the dispersion of catalyst.     

The interrelation between the electrochemical behaviour of a polymer-coated electrode and the ion exchange properties of the coating

The Fe(CN)^3−/4−₆ charge-transfer reaction was studied with anion-exchange polymer-coated electrodes SO²⁻₄ and ClO⁻₄ were mainly used as the anions of the background electrolyte. The electrochemical results are correlated with the equilibrium distribution of the ions between the electrolyte and the coating. Since the considered anions are ir-active their concentration in the film could be determined by ir-reflection-absorption spectroscopy, after the coated electrodes had been withdrawn from the electrolyte. The strong influence of the films' ion-exchange selectivity on the electrode behaviour is demonstrated.     

正渗透膜过程中临界通量的影响因素

临界通量是膜过程中一种重要的污染特性指标.采用阶梯汲取液浓度递增法测定不同污染物、架桥离子浓度及膜面流速对正渗透(FO)膜过程临界通量的影响.结果表明,海藻酸钠(SA)、腐殖酸(HA)及二氧化硅(SiO2)污染时FO膜临界通量值分别为29.32,46.35和32.17 L/(m2·h);随 Ca2+浓度由0 mmol/...     

Spectroscopic and electrochemical investigations of lead–lead dioxide glasses and vitroceramics with applications for rechargeable lead–acid batteries

A new class of glass and vitroceramic electrodes with applications for rechargeable batteries was obtained by a melt quenching method. The structural characterization of the samples having the xPb·(100−x)PbO₂ composition, where x=0, 10, 20, 30, 40 and 50 mol% Pb, was performed by UV–vis and FTIR spectroscopies investigations.UV–vis and FTIR data reveal that the excess of lead content in the host matrix generates the transformation and/or disintegration of [PbO₆] octahedral structural units into [PbO₄] structural units or Pb²⁺ ions and non-bonding oxygen ions centers.The electrochemical performances of the glass and vitroceramics electrodes were investigated by cyclic voltammetry. The shapes of the cyclic voltammograms and redox peaks depend on the electrolyte solution concentration and the lead content in host matrix. Differences between these waves are determined by the type of electrochemically active species existing in the glass or glass ceramics. The improved performance of the vitroceramic electrodes is attributed to the presence of the lead metallic phase that seems to offer an easier route for the charge process of the electrodes. Thus, the presence of these phases generates more electrochemically active species, inhibits the secondary reactions implying PbO, takes up the ionic conduction of the larger electrolyte solution and increases the charge/discharge rate of the electrochemical processes.     

Novel direct Z-scheme AgI/N–TiO2 photocatalyst for removal of polluted tetracycline under visible irradiation

In the study, we doped N into TiO₂ lattice to narrow its band gap energy. Then, the synthesized N doped TiO₂ material was combined with AgI to form AgI/N–TiO₂ (ANT) direct Z scheme materials. The synthesized materials were utilized for photocatalytic removal of tetracycline (TC) using visible irradiation as an excitation source. We also conducted radical scavenging experiments to determine photocatalytic degradation mechanism. We investigated that these photo-excited electrons (e⁻) in N–TiO₂ conduction band tended to combine with the left holes (h⁺) in AgI valence band maintaining h⁺ in the valence band of the N–TiO₂ and e⁻ in the conduction band of the AgI. The remained e⁻ and h⁺ have high redox potential to initiate for photocatalytic decomposition of TC. Thus, the TC degradation by the ANT materials were significant greater than those by single components (AgI or N–TiO₂). We also investigated that the TC degradation by the ANT-30 material, which the AgI: N–TiO₂ molar ratio was 30%, exhibited that highest degradation efficiency. Finally, the ANT photocatalyst exhibited excellent stability during TC degradation processes supporting for its promising potential application in practical systems.     

Wind modelling,validation and sensitivity study using Weather Research and Forecasting model in complex terrain

A physically-based wind model is applied to determine wind speed and direction and to conduct a model sensitivity analysis. The focus is the East African site of the Lake Turkana Wind Farm, characterized by complex terrain and high diurnal variability that creates a nocturnal jet of typically 15 m/s. Observations from three tall meteorological masts are compared with Weather Research and Forecast (WRF) model outputs. WRF is configured with four domains nested down to 900 m spatial resolution. The model is tested with initialization fields from two different sources, optimised using different grid configurations and parameterization schemes. Comparing model and data from 3 tall masts A, B and C yields that the primary source of error in WRF model simulation in a complex terrain is due to incorrect specification of boundary fields used to initialize the model. RMSEs achieved in this research are ≤2 m/s representing good model performance (Emery et al., 2001).     

Inhibitory effects of odor substances,geosmin and 2-methylisoborneol,on early development of sea urchins

Geosmin (1α,10β-dimethyl-9α-decanol) and 2-methylisoborneol ((1-R-exo)-1,2,7,7-tetramethyl-bicyclo-(2,2,1)-heptan-2-ol) (MIB) are volatile terpene derivatives, and have received a great deal of attention because they can cause musty/muddy off-flavor in water and food resources. By the Ames test, these metabolites showed no mutagenicity but antimicrobial activity toward tester strains. While these compounds are produced by various organisms living in aquatic environments, there are few reports of their effects on aquatic organisms. The effects of geosmin and MIB on sea urchin development were examined. The estimated IC₅₀ (50% inhibitory concentration) values for the formation of the fertilization membrane were 16.67 mg geosmin l⁻¹ and 68.77 mg MIB l⁻¹; those for the cell cleavage were 16.58 mg geosmin l⁻¹ and 66.86 mg MIB l⁻¹, suggesting that the toxicity of geosmin and MIB toward sea urchins are comparable to their toxicity toward Salmonella tester strains in the Ames test. These values are far greater than concentrations of these substances observed in aquatic environments with severe muddy off-flavor problems.