RECURSIVE FORECASTING, SMOOTHING AND SEASONAL ADJUSTMENT OF NONSTATIONARY ENVIRONMENTAL DATA

The paper presents a unified, fully recursive approach to the modelling, forecasting and seasonal adjustment of non‐stationary time series and shows how it can be used as a flexible tool in the analysis of environmental data. The approach is based on time‐variable parameter (TVP) versions of various well‐known time‐series models and exploits the suite of novel, recursive filtering and fixed interval smoothing algorithms available in the microCAPTAIN computer program. The practical utility of the analysis is demonstrated by an example based on the analysis of atmospheric CO2 and sea surface temperature anomaly data. Copyright © 1991 John Wiley & Sons, Ltd.