There is considerable interest in forecasting interannual hurricane activity for the Atlantic basin. Various predictors representing different components of the tropical Atlantic climate have been suggested. The choice of predictors is based on previous research into contemporaneous and lag relationships with seasonal hurricane and tropical storm frequency. Past research is divided into five distinct periods: the search for physical relationships, the use of composite charts, the use of satellite imagery and climatology of easterly waves, the emergence of recent ideas, and the development of prediction models. As an historical summary this paper describes the important research contributions in each period leading to our current understanding of yearly hurricane variability. The paper concludes by describing current methods for forecasting this variability and recommends an area for future investigations.
*Current address: National Weather Service/Alaska Region, Anchorage, Alaska.