Significance of the factor analysis due to decreasing anthropogenic pollution--exemplified by river sediments.

Environmental datasets often consist of numerous features analyzed in many investigated samples. Therefore, the evaluation of those datasets is difficult. Chemometric methods like the factor analysis are useful tools to handle big datasets. In this paper, we discussed the relation between the geogenic background (noise) and anthropogenic pollution (source) for the suitability of environmental datasets for factor analytical methods. Thus, computed test datasets with different sources, diverse maximum of the sources, and various geogenic backgrounds were generated. Afterward, the maximum of the source was decreased stepwise, a factor analysis was computed, and the corresponding results were investigated in respect of the credibility. The major impacts on the evaluation of a feature are the mean value of the noise and the standard deviation of the noise. With the help of these two parameters, a pollution index can be calculated. The maximum of the source has to exceed that index in order to be usefully evaluable with the factor analyses. The evaluation of the results of the factor analysis would become increasingly complicated if the variability of a dataset decreases due to reduced maximum values or geogenic/anthropogenic sources which correspond to increasing environmental quality.

Improving detection power in trace analysis using wavelet transform.

Environmental analysis most often is trace analysis. Therefore, the concentrations are commonly in the lower working range near the limit of detection of the corresponding analytical method. However, whenever the instrument's analytical noise is too large, it dominates the signal curves and analytes cannot be detected anymore. Furthermore, the evaluation of peaks with defined baselines is hindered very much. One possibility for de-noising is wavelet transform which is presented in this work. Different wavelet functions are applied and Symlet4 is suggested as the most powerful for analytical peaks that resemble Gaussian distribution curves, as it improves limits of detection by factors 6 to 7. The comparison of different wavelet functions has been carried out for two modern analytical scopes. At first, chromatograms are de-noised for the speciation of four arsenic compounds via the coupling of HPLC and ICP-MS. Secondly, the determination of cadmium is shown by HR-CS AAS, which is one of the most recently developed devices in atomic absorption spectrometry and allows the registration of three-dimensional spectra in order to investigate the spectral vicinity of analytical lines. On the basis of these investigations, we recommend using wavelet transform with Symlet4 for all analytical techniques which are resulting in similar signal curves.

Comparing different means of signal treatment for improving the detection power in HPLC-ICP-MS.

The purpose of detecting trace concentrations of analytes often is hindered by occurring noise in the signal curves of analytical methods. This is also a problem when different arsenic species (inorganic As(III) and As(V) as well as organic dimethylarsinic acid and arsenobetaine) are to be determined in food and feeding stuff by HPLC-ICP-MS, which is the basis of this work. In order to improve the detection power, methods of signal treatment may be applied. We show a comparison of convolution with Gaussian distribution curves, Fourier transform, and wavelet transform. It is illustrated how to estimate decisive parameters for these techniques. All methods result in improved limits of detection. Furthermore, applying baselines and evaluating peaks thoroughly is facilitated. However, there are differences. Convolution with Gaussian distribution curves may be applied, but Fourier transform shows better results of improvement. The best of the three is wavelet transform, whereby the detection power is improved by factors of about 6.

Geostatistical and multivariate statistical analysis of heavily and manifoldly contaminated soil samples.

The surroundings of the former Kremikovtzi steel mill near Sofia (Bulgaria) are influenced by various emissions from the factory. In addition to steel and alloys, they produce different products based on inorganic compounds in different smelters. Soil in this region is multiply contaminated. We collected 65 soil samples and analyzed 15 elements by different methods of atomic spectroscopy for a survey of this field site. Here we present a novel hybrid approach for environmental risk assessment of polluted soil combining geostatistical methods and source apportionment modeling. We could distinguish areas with heavily and slightly polluted soils in the vicinity of the iron smelter by applying unsupervised pattern recognition methods. This result was supported by geostatistical methods such as semivariogram analysis and kriging. The modes of action of the metals examined differ significantly in such a way that iron and lead account for the main pollutants of the iron smelter, whereas, e.g., arsenic shows a haphazard distribution. The application of factor analysis and source-apportionment modeling on absolute principal component scores revealed novel information about the composition of the emissions from the different stacks. It is possible to estimate the impact of every element examined on the pollution due to their emission source. This investigation allows an objective assessment of the different spatial distributions of the elements examined in the soil of the Kremikovtzi region. The geostatistical analysis illustrates this distribution and is supported by multivariate statistical analysis revealing relations between the elements.

Wavelet transform as a new approach to the enhancement of signal-to-noise ratio in anodic stripping voltammetry.

De-noising signals is a frequent aim achieved by signal processing in analytical chemistry. The purpose is to enable the detection of trace concentrations of analytes. The limit of detection is defined as the lowest amount of analyte that still causes signals greater than the background noise. Appropriate de-noising decreases only the noise and maintains the measurement signal, so that signal-to-noise ratios are enhanced. One adequate mean of signal processing for this purpose is wavelet transform, which still is not a common tool in analytical chemistry. In this paper, the ability of de-noising by wavelet transform is shown for measurements in anodic stripping voltammetry using a hanging mercury drop electrode. The calculation of limits of detection and signal-to-noise ratios on the basis of peak-to-peak noise is exercised to quantify the performance of de-noising. Furthermore, signal shape with regard of easing the application of base lines is discussed. Different wavelet functions are used, and the results are compared also to Fourier transform. Coiflet2 was found out to reduce noise by the factor of 330 and is proposed as the adequate wavelet function for voltammetric and similar signals.

Chemometric tools for identification of volatile aroma-active compounds in oregano.

One of the purposes of chemical analysis is to find quick and efficient methods to answer complex analytical questions in the life sciences. New analytical methods, in particular, produce a flood of data which are often very badly arranged. An effective way to overcome this problem is to apply chemometric methods. As part of the following investigations, three brands of oregano were analysed to identify their volatile aroma-active compounds. Two techniques were applied--gas chromatograpy-olfactometry (GC-O) and human sensory evaluation. Aroma-impact compounds could be identified in the main brands of oregano with the aid of chemometric methods (principal-components analysis, hierarchical cluster analysis, linear discriminant analysis, partial least-squares regression). Therefore, it is possible to reduce the analysis of sensory and olfactometry to relevant attributes. This makes classifying new species easier, much faster, and less expensive and is the premise for quick and more economic identification of new potential genotypes for oregano plant breeding. A comprehensive list of oregano key odourants, determined by GC-O and human sensory evaluation using different methods of supervised and unsupervised pattern cognition, has not previously been published.

Study of traffic-emitted lead pollution of soil and plants using different fuzzy clustering algorithms.

We discuss the clustering of 234 environmental samples resulting from an extensive monitoring program concerning soil lead content, plant lead content, traffic density, and distance from the road at different sampling locations in former East Germany. Considering the structure of data and the unsatisfactory results obtained applying classical clustering and principal component analysis, it appeared evident that fuzzy clustering could be one of the best solutions. In the following order we used different fuzzy clustering algorithms, namely, the fuzzy c-means (FCM) algorithm, the Gustafson-Kessel (GK) algorithm, which may detect clusters of ellipsoidal shapes in data by introducing an adaptive distance norm for each cluster, and the fuzzy c-varieties (FCV) algorithm, which was developed for recognition of r-dimensional linear varieties in high-dimensional data (lines, planes or hyperplanes). Fuzzy clustering with convex combination of point prototypes and different multidimensional linear prototypes is also discussed and applied for the first time in analytical chemistry (environmetrics). The results obtained in this study show the advantages of the FCV and GK algorithms over the FCM algorithm. The performance of each algorithm is illustrated by graphs and evaluated by the values of some conventional cluster validity indices. The values of the validity indices are in very good agreement with the quality of the clustering results.

Discrimination of biofilm samples using pattern recognition techniques.

Biofilms are complex aggregates formed by microorganisms such as bacteria, fungi and algae, which grow at the interfaces between water and natural or artificial materials. They are actively involved in processes of sorption and desorption of metal ions in water and reflect the environmental conditions in the recent past. Therefore, biofilms can be used as bioindicators of water quality. The goal of this study was to determine whether the biofilms, developed in different aquatic systems, could be successfully discriminated using data on their elemental compositions. Biofilms were grown on natural or polycarbonate materials in flowing water, standing water and seawater bodies. Using an unsupervised technique such as principal component analysis (PCA) and several supervised methods like classification and regression trees (CART), discriminant partial least squares regression (DPLS) and uninformative variable elimination-DPLS (UVE-DPLS), we could confirm the uniqueness of sea biofilms and make a distinction between flowing water and standing water biofilms. The CART, DPLS and UVE-DPLS discriminant models were validated with an independent test set selected either by the Kennard and Stone method or the duplex algorithm. The best model was obtained from CART with 100% correct classification rate for the test set designed by the Kennard and Stone algorithm. With CART, one variable describing the Mg content in the biofilm water phase was found to be important for the discrimination of flowing water and standing water biofilms.

Small-scale variability of metals in soil and composite sampling.

Soil pollution data is also strongly scattering at small scale. Sampling of composite samples, therefore, is recommended for pollution assessment. Different statistical methods are available to provide information about the accuracy of the sampling process. Autocorrelation and variogram analysis can be applied to investigate spatial relationships. Analysis of variance is a useful method for homogeneity testing. The main source of the total measurement uncertainty is the uncertainty arising from sampling. The sample mass required for analysis can also be estimated using an analysis of variance. The number of increments to be taken for a composite sample can be estimated by means of simple statistical formulae. Analytical results of composite samples obtained from different fusion procedures of increments can be compared by means of multiple mean comparison. The applicability of statistical methods and their advantages are demonstrated for a case study investigating metals in soil at a very small spatial scale. The paper describes important statistical tools for the quantitative assessment of the sampling process. Detailed results clearly depend on the purpose of sampling, the spatial scale of the object under investigation and the specific case study, and have to be determined for each particular case.