Pavlina Simeonova

Water quality study of the Struma river basin, Bulgaria (1989–1998)

The present paper deals with an estimation of the water quality of the Struma river. Long-term trends, seasonal patterns and data set structures are studied by the use of statistical analysis. Nineteen sampling sites along the main river stream and different tributaries were included in the study. The sites are part of the monitoring net of the region of interest. Seventeen chemical indicators of the surface water have been measured in the period 1989–1998 in monthly intervals. It is shown that the water quality is relatively stable throughout the monitoring period, which is indicated by a lack of statistically significant trends for many of the sites and by chemical variables. Several seasonal patterns are observed at the sampling sites and four latent factors are identified as responsible for the data set structure.

An advanced multivariate statistical approach to study coastal sediment data

The present paper deals with the application of classical and fuzzy principal components analysis to a large data set from coastal sediment analysis. Altogether 126 sampling sites from the Atlantic Coast of the USA are considered and at each site 16 chemical parameters are measured. It is found that four latent factors are responsible for the data structure (“natural”, “anthropogenic”, “bioorganic”, and “organic anthropogenic”). Additionally, estimating the scatter plots for factor scores revealed the similarity between the sampling sites. Geographical and urban factors are found to contribute to the sediment chemical composition. It is shown that the use of fuzzy PCA helps for better data interpretation especially in case of outliers.

Environmetric data interpretation to assess the water quality of Maritsa River catchment

Maritsa River is one of the largest rivers flowing on Bulgarian territory. The quality of its waters is of substantial importance for irrigation, industrial, recreation and domestic use. Besides, part of the river is flowing on Turkish territory and the control and management of the Maritsa catchment is of mutual interst for the neighboring countires. Thus, performing interpretation and modeling of the river water quality is a major environmetric problem. Two multivariate statstical methods (Cluster analysis/CA/and Principal components analysis/PCA/) were applied for model assessment of the water quality of Maritsa River on Bulgarian territory. The study used long-term monitoring data from 21 sampling sites characterized by 8 surface water quality indicators. The application of CA to the indicators results in 3 significant clusters showing the impact of biological, anthropogenic and eutrophication sources. For further assessment of the monitoring data, PCA was implemented, which identified, again,three latent factors confirming, in principle, the clustering output. The latent factors were conditionally named “biologic”, “anthropogenic” and “eutrophication” source. Their identification coinside correctly to the location of real pollution sources along the Maritsa River catchment. The linkage of the sampling sites along the river flow by CA identified four special patterns separated by specific tracers levels: biological and anthropogenic major impact for pattern 1, euthrophication major impact for pattern 2, background levels for pattern 3 and eutrophication and agricultural major impact for pattern 4. The apportionment models of the pollution determined the contribution of each one of identified pollution factors to the total concentration of each one of the water quality parameters. Thus, a better risk management of the surface water quality is achieved both on local and national level.

Chemical and statistical interpretation of sized aerosol particles collected at an urban site in Thessaloniki, Greece

The size distribution of aerosol particles (PM0.015-PM18) in relation to their soluble inorganic species and total water soluble organic compounds (WSOC) was investigated at an urban site of Thessaloniki, Northern Greece. The sampling period was from February to July 2007. The determined compounds were compared with mass concentrations of the PM fractions for nano (N: 0.015 < Dp < 0.06), ultrafine (UFP: 0.015 < Dp < 0.125), fine (FP: 0.015 < Dp < 2.0) and coarse particles (CP: 2.0 < Dp < 8.0) in order to perform mass closure of the water soluble content for the respective fractions. Electrolytes were the dominant species in all fractions (24–27%), followed by WSOC (16–23%). The water soluble inorganic and organic content was found to account for 53% of the nanoparticle, 48% of the ultrafine particle, 45% of the fine particle and 44% of the coarse particle mass. Correlations between the analyzed species were performed and the effect of local and long-range transported emissions was examined by wind direction and backward air mass trajectories. Multivariate statistical analysis (cluster analysis and principal components analysis) of the collected data was performed in order to reveal the specific data structure. Possible sources of air pollution were identified and an attempt is made to find patterns of similarity between the different sized aerosols and the seasons of monitoring. It was proven that several major latent factors are responsible for the data structure despite the size of the aerosols – mineral (soil) dust, sea sprays, secondary emissions, combustion sources and industrial impact. The seasonal separation proved to be not very specific.

Environmetric approaches for lake pollution assessment.

The application of multivariate statistical methods to high mountain lake monitoring data has offered some important conclusions about the importance of environmetric approaches in lake water pollution assessment. Various methods like cluster analysis and principal components analysis were used for classification and projection of the data set from a large number of lakes from Rila Mountain in Bulgaria. Additionally, self-organizing maps of Kohonen were constructed in order to solve some classification tasks. An effort was made to relate the maps with the input data in order to detect classification patterns in the data set. Thus, discrimination chemical parameters for each pattern (cluster) identified were found, which enables better interpretation of the pollution situation. A methodology for application of a combination of different environmetric methods is suggested as a pathway to interpret high mountain lake water monitoring data.

Statistical Calibration of Model Solution of Analytes

Statistical Calibration of Model Solution of Analytes A new method based on spectrophotometric-partial least-squares procedure was proposed for simultaneously determination of thorium and zirconium using SPADNS (4,5-Dihydroxy-3-(p-sulfophenylazo)-2,7-naphthalene disulfonic acid, trisodium salt) as a color reagent. Absorbance measurements were made in the range of γ = 541÷620 nm with 1.0 nm steps in buffered solutions at pH 3.5. The linear ranges were obtained for 0.5÷11.5 and 1.5÷14.5 μg cm-3 for Th4+ and Zr4+ ions, respectively. The limits of detection were determined 0.4 and 1.2 μg cm-3 for thorium and zirconium, respectively. The standard deviation (n = 3) and recovery percent of 10 samples in the prediction set were obtained in the amplitude 0.22÷0.38 μg cm-3 and 91.3÷109.2, respectively. The proposed method was used for simultaneously determination of mentioned ions in spiked real water samples and wastewater. The results show that the method is applicable for the analysis of samples with similar matrix. Model Statystycznej Kalibracji Roztworów Analitów Nowa metoda opiera się na procedurze spektrofotometrycznej - najmniejszych kwadratów, która została zaproponowana do równoczesnego oznaczania toru i cyrkonu z wykorzystaniem SPADNS (kwas 4,5-Dihydroksy-3-(p-sulfofenylazo)-2,7-naftaleno disulfonowy, sól trisodowa) jako odczynnika koloru. Pomiarów absorbancji dokonano w zakresie λ = 541÷620 nm co 1,0 nm w roztworach buforowych o pH 3,5. Liniowy zakres uzyskano przy stężeniach jonów Th4+ i Zr4+ odpowiednio 0,5÷11,5 i 1,5÷14,5 μg cm-3. Granice wykrywalności dla toru i cyrkonu wynosiły odpowiednio 0,4 i 1,2 μg cm-3. Wyznaczono odchylenie standardowe (n = 3) i procent odzysku w serii 10 próbek odpowiednio w zakresie 0,22÷0,38 i 91,3÷109,2 μg cm-3. Proponowana metoda została zastosowana do równoczesnego oznaczania wymienionych jonów w wzbogaconych próbkach rzeczywistych wody i ścieków. Wyniki pokazują, że ta metoda może być wykorzystywana do analizy próbek o podobnej matrycy.

Some features of bulk melt-textured high-temperature superconductors subjected to alternating magnetic fields

Monolithic, large grain, (RE)Ba2Cu3O7 high-temperature superconductors (where RE denotes a rare-earth ion) are known to be able to trap fields in excess of several teslas and represent thus an extremely promising competing technology for permanent magnet in several applications, e.g. in motors and generators. In any rotating machine, however, the superconducting permanent magnet is subjected to variable (transient, or alternating) parasitic magnetic fields. These magnetic fields interact with the superconductor, which yields a reduction of the remnant magnetization. In the present work we quantify these effects by analysing selected experimental data on bulk melt-textured superconductors subjected to AC fields. Our results indicate that the non-uniformity of superconducting properties in rather large samples might lead to unusual features and need to be taken into account to analyse the experimental data. We also investigate the evolution of the DC remnant magnetization of the bulk sample when it is subjected to a large number of AC magnetic field cycles, and investigate the experimental errors that result from a misorientation of the sample or a mispositioning of the Hall probe. The time-dependence of the remnant magnetization over 100000 cycles of the AC field is shown to display distinct regimes which all differ strongly from the usual decay due to magnetic relaxation.

Statistical modeling of air pollution

The present communication deals with the application of several chemometric methods (principal components analysis, source apportioning on absolute principal components scores, chemical mass balance, self-organizing maps) to various aerosol data collections from different regions in Europe. It is shown that different latent factors explaining over 75 % of the total variance are responsible for the data structure and could be reliable identified and interpreted. Further, the contribution of each identified source to the formation of the particle total mass and chemical compounds total concentration is calculated. Thus, a reliable assessment of the air quality in the respective region is done. Classification by self-organizing maps makes it possible to better understand the role of different discriminating tracers in the air pollution. The use of chemical mass balance approach ensures a sound modeling of the pollution sources. The requirements of the sustainability concept for ecological indicators in this case is easily transformed to a multivariate statistical problem taking into account not separate indicators but the specific multivariate nature of the aerosol pollution.

Chemometrics as an option to assess clinical data from diabetes mellitus type 2 patients

The present study deals with the application of two major multivariate statistical approaches - Cluster Analysis (CA) and Principal Components Analysis (PCA) as an option for assessment of clinical data from diabetes mellitus type 2 patients. One hundred clinical cases of patients are considered as object of the statistical classification and modeling, each one of them characterized by 34 various clinical parameters. The goal of the study was to find patterns of similarity, both between the patients and the clinical tests. Each group of similarity is interpreted revealing at least five clusters of correlated parameters or five latent factors, which determine the data structure. Relevant explanation of the clustering is found based on the pattern of similarity like glucose level, anthropometric data, enzyme level, liver function, kidney function etc. It is assumed that this classification could be of help in optimizing the performance of clinical test for this type of patients and for designing a pattern for the role of the different groups of test in determining the metabolic syndrome of the patients.

Chemometric expertise of the quality of groundwater sources for domestic use.

In the present study 49 representative sites have been selected for the collection of water samples from central water supplies with different geographical locations in the region of Kavala, Northern Greece. Ten physicochemical parameters (pH, electric conductivity, nitrate, chloride, sodium, potassium, total alkalinity, total hardness, bicarbonate and calcium) were analyzed monthly, in the period from January 2010 to December 2010. Chemometric methods were used for monitoring data mining and interpretation (cluster analysis, principal components analysis and source apportioning by principal components regression). The clustering of the chemical indicators delivers two major clusters related to the water hardness and the mineral components (impacted by sea, bedrock and acidity factors). The sampling locations are separated into three major clusters corresponding to the spatial distribution of the sites – coastal, lowland and semi-mountainous. The principal components analysis reveals two latent factors responsible for the data structures, which are also an indication for the sources determining the groundwater quality of the region (conditionally named “mineral” factor and “water hardness” factor). By the apportionment approach it is shown what the contribution is of each of the identified sources to the formation of the total concentration of each one of the chemical parameters. The mean values of the studied physicochemical parameters were found to be within the limits given in the 98/83/EC Directive. The water samples are appropriate for human consumption. The results of this study provide an overview of the hydrogeological profile of water supply system for the studied area.