We present a method for investigating the evolution of trend and seasonalit
y in an observed time series. A general model is fitted to a residual spect
rum, using components to represent the seasonality. We show graphically how
well the fitted spectrum captures the evidence for evolving seasonality as
sociated with the different seasonal frequencies. We apply the method to mo
del two time series and illustrate the resulting forecasts and seasonal adj
ustment for one series. Copyright (C) 2000 John Wiley & Sons, Ltd.