Josu Artech and Peter M Robinson

Published September 1998

---

There has recently been great interest in time series with long memory, namely series whose dependence decays slowly in the sense that autocovariances are not summable and the spectral density is unbounded. This concept has been extended to SCLM (Seasonal/Cyclical Long Memory) where the dependence between seasonal or cyclic observations decays similarly slowly. We discuss issues related to SCLM processes such as modelling, estimation, statistical inference, applications and extensions.