ARIMA Modelling Of NO, NO2 And 83 Time Series

ARIMA Modelling of NO, NO% and 83 Time Series Sanchez Gomez, M. L.*; Caballero Sahelices, C.**; and Lucio Garcia, J. H.**. * Departamento de Fisica Aplicada /, Facultad de Ciencias. Universidadde Valladolid (Spain). email: marisa@cpd.uva.es ** Departamento de Fisica, Facultad de CyTA y C. Quimicas. Universidad de Burgos (Spain), email: concesa@ubu.es and jlucio@ubu. es Abstract. Ozone and nitrogen oxides concentrations are interrelated and depend on a great deal of physical and chemical parameters (such as meteorological conditions, solar radiation, and the presence of hydrocarbons and complex reactions with numerous catalysts). For the sake of comprehension and prediction, frequently we want to know the inner structure of each temporal series, assumed that it can be aproached by a simple univariate mathematical model. Ozone episodes in Burgos (North of Spain) have led us to perform a complete univariate ARIMA analysis to NO, NOj and O, time series (hourly averages) corresponding