Portal de Periódicos da CAPES

Chemical Mass Balance

2001; Linguagem: Inglês

10.1002/9780470057339.vac017

Philip K. Hopke,

Atmospheric chemistry and aerosols

The management of air quality is a difficult but important problem. In general it involves identification of the sources of materials emitted into the air (see Air Quality Indicators, Pollutant Specific), quantitative estimation of the emission rates of the pollutants, understanding of the transport of the substances from the sources to downwind locations, and knowledge of the physical and chemical transformation processes that can occur during that transport (see Atmospheric Dispersion: Chemistry). All of those elements can then be put together into a mathematical model that is used to estimate the changes in observable airborne concentrations that might be expected to occur if various actions were taken. Such actions could include the initiation of new sources as new industries are built and begin to function, and the imposition of emission controls of existing facilities to reduce the pollutant concentrations. However, the atmosphere is a very complex system (see Meteorology) and it is necessary to simplify greatly the descriptions of reality to produce a mathematical model capable of being calculated on even the largest and fastest computers. Thus, although significant improvements have been made over the past 20 years in the mathematical modeling of dispersion of pollutants in the atmosphere, there are still many instances when the models are insufficient to permit the full development of effective and efficient air quality management strategies. Thus, it is necessary to have other methods available to assist in the identification of sources and the apportionment of the observed pollutant concentrations to those sources (see Source Apportionment). Such methods are called receptor-oriented or receptor models since they are focused on the behavior of the ambient environment at the point of impact (see Concentration, Ambient), as opposed to the source-oriented dispersion models that focus on the transport, dilution and transformations that occur at the source and follow the pollutants to the sampling or receptor site. Initially ordinary least squares was employed [5] for the estimation. Since different elements have quite different scales for their values (major elements at µg m−3 concentrations, minor elements at concentrations of hundreds of ng m−3 and trace elements at ng m−3 values), a weighted least squares regression analysis (see Least Squares, General) has been used to fit six sources with eight elements for ten ambient samples [8]. In these analyses, the ambient elemental concentrations are weighted by the inverse of the square of the analytical uncertainty in that measurement. Because of limitations of performing the least-squares analysis, it is useful to have additional techniques that can help to determine the applicability of source profiles to the particulate apportionment problem being solved. These methods have been developed as solutions to the problem of calibrating multivariate chemical analysis instruments (see Calibration), but these methods can be applied to the receptor modeling problem. The methods applied to date include partial least squares [4, 13-15], simulated annealing [10], genetic algorithms [1] and backpropagation artificial neural networks [11]. In particular, the artificial neural network showed better results with respect to collinearity of sources. However, these methods have not been extensively tested in solving actual chemical mass balance problems. (See also Chemometrics; Atmospheric Dispersion: Heavy Gases; Emission Inventory; Turbulent Diffusion)