Environmental Pollution Analysis by Dynamic Structural Equation Models

As requested by the framework EU Directive on air quality assessment and management (96/62/EC) and related ‘daughter’ directives, air quality standards for specific pollutants are designed to protect public health and environment. Modeling is one of the main activities to evaluate air quality and to prepare future control programs. Of course, this is not an easy task since a variety of pollutants may undergo chemical reactions between themselves and with other species. Nevertheless, in this paper we attempt to discuss a framework to construct a multivariate model which is able to capture the dynamical interactions among pollutants and meteorological variables. Specifically, assuming that latent or background effects underlying the fluctuation of observations can be estimated, a dynamic structural equation model is developed in a state-space form. A research study on the Milan district for data provided by the Lombardia Environmental Protection Agency (ARPA) is presented.