The geomagnetic field measured at any point on Earth as a function of time shows periodic variations due to atmospheric processes. A method of time series analysis based on discrete Fourier transforms is developed for the detection and estimation of lines in the frequency spectrum; the method gives estimates, with error limits, of the amplitudes of sinusoidal variations in the data, and these estimates are unbiased by noise. The method is used to determine the lunar and solar variations present in Abinger declination records, data for a period of 17 years (1927—56) being used. The coherence between Abinger data and time series that model the lunar and solar input functions to the atmosphere aids the identification of lines in the Abinger spectrum. The amplitude spectrum of magnetic declination at Abinger shows prominent solar peaks at 1, 2, 3, 4 cycles per day (c/d), and lunar peaks at 0.0703, 1.932, and 2.932 c/d. Sidebands of the solar diurnal peak at 1 ± 0.037 c/d are attributed to a solar rotation mechanism, and this is supported by a high resolution power spectrum that shows the width of these sidebands. The high resolution analysis also shows annual and semi-annual splitting of the solar diurnal and semi-diurnal lines, and of the lunar semi-diurnal line.
High‐resolution analysis of tomato leaf elongation: the application of novel time‐series analysis techniques