On the modelling of seasonal fluctuations in fish landings using structural time-series approach

Abstract

For describing seasonal fluctuations in fish landings, a new promising approach, viz. structural time-series modelling (STSM) is discussed. Here trend, seasonal fluctuations, etc. exhibited by data dictate particular model from this family to be employed. Two types of forms, viz. Dummy seasonality, and Trigonometric seasonality are considered for describing seasonal fluctuations. As illustrations, these models are applied to quarterly landings of silverbellies and croakers in Tamil Nadu during 1975-96 Comparisons with well-known auto regressive integrated moving average (ARIMA) models reveal that former approach is much superior to latter. Finally, identified models are used for forecasting purpose.