ANALYSIS OF LONG-TERM VARIATIONS IN CONCE NTRATIONS OF ENVIRONMENTAL COMPONENTS - DEVELOPMENT OF TIME-SERIES ANALYSIS FOR LONG-TERM VARIATION WITH SEASONAL-VARIATION

We developed a new time series analysis [TSA] for long-term variation with seasonal variation in concentrations of environmental components. The regression curve with two terms is used in this TSA. One term mea ns long-term tendency of an increase or a decrease in concentration of the component. Another gives numerical information about seasonal var iation, i.e. time when the highest concentration is observed in a year and regularity on seasonal cycle. The proposed TSA was successfully a pplied for characterizing the long-term variation in the amount of sol uble and insoluble compounds in deposit samples collected for 12 years in Kanazawa. Variations in inorganic components in deposit samples we re divided into three classes and five groups. This classification was in good agreement with that obtained by the factor analysis [FA]. Our TSA seemd to be helpful in characterizing the factors in FA.