Daily surface ozone maximum concentrations at Taliarte, Canary Islands

Two different methods for using air quality data to estimate daily maximum concentrations along a year of surface ozone concentration in a new measurement station in Canary Islands are described. It is usually assumed that sequences of pollutant concentration show stationary behaviour. In this paper, it is shown that the application of such procedure could be useful to estimate probabilities of maximum concentration of surface ozone. Nevertheless, this method does not take into account the time structure of the original data, so that some caution should be taken into account. In this sense, an alternative procedure, assuming nonstationary behaviour of data, is developed to estimate empirically the probability distributional properties of the original series. Its application yields a substancial improve of the explanation of underlying series structure.