Optimization of artificial neural network models through genetic algorithms for surface ozone concentration forecasting

Introduction

This study proposes three methodologies to define artificial neural network models through genetic algorithms (GAs) to predict the next-day hourly average surface ozone (O₃) concentrations. GAs were applied to define the activation function in hidden layer and the number of hidden neurons.

Methods

Two of the methodologies define threshold models, which assume that the behaviour of the dependent variable (O₃ concentrations) changes when it enters in a different regime (two and four regimes were considered in this study). The change from one regime to another depends on a specific value (threshold value) of an explanatory variable (threshold variable), which is also defined by GAs. The predictor variables were the hourly average concentrations of carbon monoxide (CO), nitrogen oxide, nitrogen dioxide (NO₂), and O₃ (recorded in the previous day at an urban site with traffic influence) and also meteorological data (hourly averages of temperature, solar radiation, relative humidity and wind speed). The study was performed for the period from May to August 2004.

Results and discussion

Several models were achieved and only the best model of each methodology was analysed. In threshold models, the variables selected by GAs to define the O₃ regimes were temperature, CO and NO₂ concentrations, due to their importance in O₃ chemistry in an urban atmosphere.

Conclusion

In the prediction of O₃ concentrations, the threshold model that considers two regimes was the one that fitted the data most efficiently.     

Experimental and statistical validation of SPME-GC-MS analysis of phenol and chlorophenols in raw and treated water

A procedure based on solid-phase microextraction (SPME) and gas chromatography coupled with mass spectrometry (GC-MS) was developed and validated in order to analyse 10 phenols in water samples. The optimised conditions were obtained using polyacrylate fibre (PA), 20ml of sample volume, 10% NaCl, pH 4.0 and direct extraction at 35 degrees C and 1000rpm, for 40min. The linear range and quantification limits for these compounds by SPME-GC-MS were defined. An evaluation of the main uncertainty sources of this method is included, which allows expanded uncertainties in the 9.4-35% range for the majority of the compounds. The main source of uncertainty is associated with matrix effects. The validated method is suitable for monitoring the production and distribution of potable water and was used, in field trials, for the analysis of samples from main intakes of water (surface or underground) and from water supply system of a large area (Lisbon and neighbour municipalities).     

Frequency of hepatocellular fibrillar inclusions in European flounder (Platichthys flesus) from the Douro River estuary,Portugal

Liver lesions in wild fish have been associated with xenobiotic exposure. Facing reports of pollution in the Douro River estuary (north of Portugal), we have been making field surveys using fishes and targeting histopathological biomarkers of exposure and effect. Herein, we intended to better characterize and report the rate of one poorly understood lesion—hepatocellular fibrillar inclusions (HFI)—found in European flounder (Platichthys flesus). With this report, we aimed to establish sound baseline data that could be viewed as a starting point for future biomonitoring, while offering the world's second only pool of field data on such a liver toxicopathic lesion, which could be compared with data available from the UK estuaries. Sampling was done in the Douro River estuary over 1 year. A total of 72 animals were fished with nets, in spring–summer (SS) and autumn–winter (AW) campaigns. Livers were processed for histopathology and both routine and special staining procedures (alcian blue, periodic acid Schiff (PAS), tetrazonium coupling reaction). Immunohistochemistry targeted AE1/AE3 (pan cytokeratins). The severity of the HFI extent was graded using a system with four levels, varying from 0 (absence of HFI) to 3 (high relative density of cells with HFI). Cells (isolated/groups) with HFI appeared in 35 % or more of the fish, in the total samples of each season, and over 40 % in more homogeneous sub-samples. There were no significant differences when comparing samples versus sub-samples or SS versus AW. When merging the data sets from the two seasons, the frequency of fish with HFI was ≈36 % for the total sample and ≈49 % for the sub-sample. The extreme group (biggest and smallest fish) revealed a HFI frequency of only 16 %, which differed significantly from the total and sub-sampled groups. Immunostaining and PAS were negative for the HFI, and alcian blue could, at times, faintly stain the inclusions. These were positive with the tetrazonium reaction. We showed the presence of HFI in European flounder from the Douro River estuary, proving that they are essentially protein in nature, that no seasonal changes existed in the HFI frequency, and that it was rarer in the smallest and biggest fish groups. Within the ranges of weight/size of our total sample, we estimate that the frequency of HFI in the local flounder is ≈35 %. That rate stands as a baseline value for future assessments, namely for biomonitoring purposes targeting correlations with the estuary pollution status.     

Teratological effects of pesticides in vertebrates: a review

Abstract

In the last decades, the use and misuse of pesticides in the agriculture have increased, having a severe impact on ecosystems and their fauna. Although the various effects of pesticides on biodiversity have been already documented in several studies, to our knowledge no consistent overview of the impact of pesticides in vertebrates, both terrestrial and aquatic, is available. In this review, we try to present a concise compilation of the teratogenic effects of pesticides on the different classes of vertebrates – mammals, birds, reptiles, amphibians and fish.     

Production of irrigation water from bioremediation of acid mine drainage: comparing the performance of two representative systems

The ability of two active bioremediation systems, based on the activity of sulphate-reducing bacteria, for the production of irrigation water from acid mine drainage (AMD), was compared. System I consisted of a packed bed reactor containing calcite tailing followed by an anaerobic packed bed bioreactor. System II consisted of a settler, fed with AMD and treated water recycle, and a sulphidogenic anaerobic packed bed bioreactor, fed with clarified settler effluent. Although efficient in terms of pH neutralisation and metals removal, system I did not allow continuous production of suitable irrigation water, since the concentration of sulphate in the effluent frequently exceeds the maximum recommended value by local legislation. On the other hand, in system II very high sulphate reduction (>99.0%) was achieved during the whole experiment, being the values in the effluent much lower than legislated limits for irrigation water. System II, not requiring the addition of any neutralising agent, can also be advantageous from an economical perspective.     

Photodegradation of Diquat and Paraquat in aqueous solutions by titanium dioxide: evolution of degradation reactions and characterisation of intermediates

The titanium dioxide assisted photodegradation of Diquat and Paraquat herbicides solutions has been the subject of the present investigation, considering its direct application in the treatment of contaminated waters and soils. To have a better understanding of the photodegradation process, different types of TiO2, commercial and 'home prepared' Ti(1-x)FexO2 (x = 0% and 4%), were used as catalysts, using an UV light as radiation source. The degradation reactions were followed by UV spectroscopy and the intermediates and reaction products were characterised by electrospray ionisation mass spectrometry (ESIMS) combined with collision-induced dissociation (CID) and tandem mass spectrometry (MS/MS). The present study shows that, for photocatalytic degradation of Diquat and Paraquat solutions, a basic pH can be determinant, as well as the type of catalyst. The type of catalyst can also strongly influence the degradation pattern of the herbicide. Regarding complete degradation, we were able to show that Diquat is more persistent than Paraquat. During the photocatalytic processes, several intermediate and reaction products are sequentially formed, to which structures are proposed.     

Ferronickel slag as free-draining rockfill dike material: a novel waste solution for mining regions

Journal of Material Cycles and Waste Management - Mining sites are vulnerable to erosion and siltation of rivers. While the construction of rockfill dikes can mitigate siltation, existing rockfill...