Air pollutant analytical data

Application of Soft Independent Modeling of Class Analogy Pattern Recognition to Air Pollutant Analytical Data... 

Pattern recognition has been applied In many forms to various types of chemical data (1,2). In this paper the use of SIMCA pattern recognition to display data and detect outliers In different types of air pollutant analytical data Is Illustrated. Pattern recognition Is used In the sense of classification of objects Into sets with emphasis on graphical representations of data. Basic assumptions which are Implied In the use of this method are that objects In a class are similar and that the data examined are somehow related to this similarity. 

In the following discussion, three types of air pollutant analytical data will be examined using principal component analysis and the K-Nearest Neighbor (KNN) procedure. A set of Interlaboratory comparison data from X-ray emission trace element analysis, data from a comparison of two methods for determining lead In gasoline, and results from gas chromatography/mass spectrometry analysis for volatile organic compounds In ambient air will be used as Illustrations. 

Hundreds of chemical species are present in urban atmospheres. The gaseous air pollutants most commonly monitored are CO, O3, NO2, SO2, and nonmethane volatile organic compounds (NMVOCs), Measurement of specific hydrocarbon compounds is becoming routine in the United States for two reasons (1) their potential role as air toxics and (2) the need for detailed hydrocarbon data for control of urban ozone concentrations. Hydrochloric acid (HCl), ammonia (NH3), and hydrogen fluoride (HF) are occasionally measured. Calibration standards and procedures are available for all of these analytic techniques, ensuring the quality of the analytical results... 

Sawicki (13) used solid-surface fluorescence techniques extensively in the 1960 s for air pollution research. In 1967, Roth (14) reported the RTF of several pharmaceuticals adsorbed on filter paper. Schulman and Walling (15) showed that several organic compounds gave RTF when adsorbed on filter paper. Faynter et al. (16) reported the first detailed analytical data for RTF and gave limits of detection, linear dynamic ranges, and reproducibilities for the compounds. 

An example from the field of environmental analytical chemistry is chosen to demonstrate the application of principal component analysis. The data is taken from an investigation on air pollution in the city of Vienna (Austria) by polycyclic aromatic hydrocarbons (PAH) (ref. 13). 

Sampling and analytical methods for the collection and measurement of different size classes of PM as well as of particle-bound metals, organic compounds, and other substances are a major issue of the present book. Routine air monitoring networks based on physico-chemical measurements provide continuously data on ambient PMio and PM2.5 concentrations, but in most cases do not inform about the chemical composition of the dust load. If trace compounds of PM are monitored at aU, such measurements are restricted to few components. Hence, knowledge about the chemical composition of PM, the local and regional distribution of airborne particle-bound substances and their toxic, genotoxic and ecotoxic potential is still very limited. Moreover, data on atmospheric pollutant concentrations do not permit to draw conclusions on possible adverse effects on human beings and ecosystems as their sensitivity to air pollution is influenced by many abiotic and biotic factors. 

Analytic or individual-level studies have been directed either at the effect of air pollution exposure on respiratory health generally, or on the status of persons with conditions (e.g., asthma) that make them more susceptible to air pollution than the population in general. The Six-Cities Study, a prospective cohort study, and the 24-Cities Study, a cross-sectional study, were of the general design. The panel study, a short-term cohort study involving relatively intensive assessment of outcome, has been used to assess the effects of air pollution on susceptible persons. Typically, a panel of participants is enrolled and asked to maintain a diary of symptom status and medication use, and physiological measurements, such as peak expiratory flow rate (PEFR) may be made. New methods for data analysis have also made this design more informative than previously (7,8). 

The analytic principles that have been applied to accumulate air quality data are colorimetry, amperometry, chemiluminescence, and ultraviolet absorption. Calorimetric and amperometric continuous analyzers that use wet chemical techniques (reagent solutions) have been in use as ambient-air monitors for many years. Chemiluminescent analyzers, which measure the amount of chemiluminescence produced when ozone reacts with a gas or solid, were developed to provide a specific and sensitive analysis for ozone and have also been field-tested. Ultraviolet-absorption analyzers are based on a physical detection principle, the absorption of ultraviolet radiation by a substance. They do not use chemical reagents, gases, or solids in their operation and have only recently been field-tested. Ultraviolet-absorption analyzers are ideal as transfer standards, but, as discussed earlier, they have limitations as air monitors, because aerosols, mercury vapor, and some hydrocarbons could, interfere with the accuracy of ozone measurements made in polluted air. 

Analytical considerations complicate the interpretation of the existing data base of atmospheric measurements of DMS and H2S. In the case of DMS the observed reduction in polluted air masses may have, at least in part, resulted from to oxidative losses during sampling. For H2S, only one analytical method has been used, and much of the existing data may be biased by an OCS artifact. New methods, preferably chromatographic or spectroscopic, are needed for HjS. 

The environmental scientist has at his disposal a variety of sensitive, multi-elemental analytical methods that can lead to a massive amount of data on airborne metals. Optimum use of these tools for environmental monitoring calls for focusing resources only on those metals that are environmentally important. Considerations of toxicity along with their ability to interact in the air, leading to the formation of secondary pollutants, and their presence in air have led to the identification of 17 environmentally important metals nickel, beryllium, cadmium, tin, antimony, lead, vanadium, mercury, selenium, arsenic, copper, iron, magnesium, manganese, titanium, chromium, and zinc. In addition to the airborne concentration, the particle size of environmentally important metals is perhaps the major consideration in assessing their importance. 

---