A spectral analysis primer

Publisher Summary The chapter describes spectral parameters, their interpretations, and some of their applications. The theoretical developments of time-series methods have followed a rather cyclical pattern by first emphasizing on one domain and then the other. Time-series spectral analysis is the mathematical equivalent of the decomposition of light into its color components by a prism. The idea that other physical phenomena could be similarly broken down into spectral components and analyzed using the concepts developed from the study of light has been and still is immensely appealing. The statistical methods of linear regression developed at that time, with their concomitant dimension of prediction, appeared to be a natural and logical methodology to apply to time series. A regression equation is simply fitted to the time series for a block of consecutive times.