DETERMINATION OF SOURCES OF AIR POLLUTANTS IN A RURAL AREA OF BACKGROUND UNIX THROUGH MODELS - PMF MEASURES APPLIED TO TIME IN 3 YEARS 64 COVS

Abstract

Currently, knowledge on the composition of volatile organic compounds (VOCs) in rural areas is still low compared to urban or industrial areas. Their determination in ambient air is a topical issue and importance in recent years due mainly to the participation of these compounds in atmospheric chemical reactions that lead to the formation of tropospheric ozone and other photochemical oxidants.

Valderejo, Álava (Spain), was declared a natural park in January 1992. The origin of VOCs measured in the sampling area is both anthropogenic and biogenic. Among the anthropogenic sources of VOCs include: the issuance by vehicles, the use of solvents, industry and the use of natural gas and within vegetation mainly biogenic sources. From among the plurality of biogenic isoprene were studied and by clearly monoterpenes majority being in the zone. Paraffins are the major compounds detected in this environment, followed by olefins. Over 30 compounds have
minimum and average values below the detection limit of the analytical equipment, as is typical in an area of rural background. During the measurements the concentrations of anthropogenic compounds have been declining due to the decline in industrial activity in the region. However, certain periods with somewhat higher than usual concentrations reflect the impact of local traffic, human intervention and phenomena as atmospheric transport of polluted air masses.

The source-receptor models use statistical procedures to identify and quantify sources of pollution in a receiving location. The USEPA has developed three models: CMB (Chemical Mass Balance), UNMIX and PMF (Positive Matrix
Factorization). We have used this type of receptor models that do not require prior knowledge of the profile of each source, specifying data only environmental measures and knowledge of source tracers to interpret the factors. This involves applying statistical procedures which allow inferring, for example, the mixture of hydrocarbon sources in a given impacting with the advantage that it can analyze a set of multivariate data using fewer components point.