Keywords: air pollution, autoregressive model, autoregressive moving average, ARMA model, time series analysis, sulphur dioxide, SO2, particulate matter, Turkey, modelling
Time series analysis for the sulphur dioxide and particulate matter concentrations in the Aegean Region of Turkey
Air pollution has temporal and spatial variability. Time series analysis is commonly used to evolution of air pollutants in time. The sulphur dioxide (SO2) and particulate matter (PM10) concentrations in the cities of the Aegean Region for the period of 1990–2009 have been modelled by using autoregressive (AR) and autoregressive moving average (ARMA) models for the yearly, monthly, and winter season pattern. The AR(2) model was generally observed for the yearly SO2 data in the Aegean Region according to the all investigated and AR(p) models. The AR(1), AR(2) and ARMA(1,2) models could provide reliable and satisfactory predictions for the yearly PM10 data according to the all analysed models. The most suitable model was predominantly determined as ARMA(1,2) model according to the all investigated models, while the best fitted model was mainly determined as AR(1) model according to the AR(p) type models for the monthly SO2 and PM10 data. The AR and ARMA models could be used for reliable and satisfactory predictions for the winter season SO2 and PM10 data to all analysed models.