Modelling and forecasting airport passengers: a case study for an introductory forecasting course

This case study utilises monthly data on the total number of airport passengers from the Central Illinois Regional Airport to illustrate the steps required to construct and estimate two low-cost time series models for use in generating in-sample and out-of-sample forecasts: an Autoregressive Integrated Moving Average (ARIMA) model and an autoregressive-seasonal-trend model. The case highlights the importance of the various stages in the development of the respective models: identification, estimation and residual diagnostics. The results of the case study reveal that the autoregressive-seasonal-trend model outperforms the ARIMA model with respect to both in-sample and out-of-sample forecasting performance.