Abstract:
Data showing external airline passengers' numbers through Khartoum Airport (in the period jan2002-dec2006) were analyzed, as a time series, in both the frequency domain and the time domain.
In the frequency domain, the methods of spectral analysis and classical decomposition methods were used to detect seasonality. In the time domain, exponential smoothing methods and decomposition method were used in forecasting future numbers of air passengers.
It is shown that the series has obvious seasonality and that Holt-Winter’s method gave the best result in forecasting airline passenger among the exponential smoothing methods, tried, but that the decomposition method is slightly superior to it.