- 2015 Elsevier Ltd. All rights reserved. A sequence of data entries is called a time series. Most time series have a random character. A time series where there is no correlation from one member of the series to another is called white noise. Models of time series can be constructed from white noise by having each new entry a sum of a white noise contribution and a linear combination of previous entries. Special insights can be gained by examining the Fourier representation of the time series by developing it into a series of sinusoids of different frequency. A brief review of the technique is presented with attention to errors due to the limited sample or record length of a given time series.