Multidimensional data processing system for air quality system

  1. Nicolás Rubio Barragán
  2. Eliseo Pablo Vergara González
  3. Francisco Javier Martínez de Pisón Ascacíbar
  4. Fernando Alba Elías
VIII Congreso Internacional de Ingeniería de Proyectos: Bilbao 6-8 de octubre de 2004. Actas

Verlag: Asociación Española de Ingeniería de Proyectos (AEIPRO)

ISBN: 84-95809-22-2

Datum der Publikation: 2005

Kongress: CIDIP. Congreso Internacional de Ingeniería de Proyectos (8. 2004. Bilbao)

Art: Konferenz-Beitrag


The 1996/62/CE European directive for the quality of the atmosphere is been developed in regular succession by specific rules, well known as 'daughter' directives. The first of them is the 1999/30 directive on the limit values of sulphur dioxide, nitrogen dioxide and oxides, particles and lead in the atmosphere, was translated to the Spanish law system by the law R.D. 1073/2002 on October the 18th. This directive establishes limit values for the SO2, PM10 and lead polluting agents, that will not be effective until 2005; while for the nitrogen oxides they will not be effective until 2010. A key factor for this daughter directive is the relevance bestowed to the mathematical prediction models in order to estimate the concentration level. For such a goal, upper and lower thresholds of evaluation a for each polluting agent are defined, so that the determination of the concentrations could be performed just by modelling techniques provided that the frequent concentration levels were below the lower threshold. As a result, a pervasive reduction in the costs related to the management of the atmosphere quality survey network. The concentration levels of these polluting agents depend, not only on the clearly measurable parameters but on others whose determination, and specially the evaluation of their effect, turns to be way past more complex. Thus the need of the development of specific models for each location. That has lead to the development of a graphical environment based on the R project, to analyzed and develop numerical models with neural networks in the more systematical possible manner.