Author(s)

V.V. Anh, H. Duc & Q. Tieng

Abstract

This paper describes a fractional autoregressive model and a co-integration model for the prediction of maximum daily ozone concentration at Lidcombe The models accommodate long-range dependence (LRD) and second-order intermittency of the data. It is found that ozone and wind speed are co- integrated, and the resulting fractional co-integration model gives a much improved performance on predicting ozone episodes than the univariate model, which relies on the history of the daily ozone series alone. 1 Introduction An air quality management scheme requires a thorough understanding of the trends in monitoring dat

Keywords