Identification of frequency components in electron content data by nonlinear filtering

A nonlinear filtering technique, called multiplicative homomorphic signal processing, is used to identify three groups of frequency components in electron content data. The first group of components results from the modulation of solar activity produced by the 27 day solar rotation. The second group is of lunar origin. The lunar frequencies are caused by nonlinear processes which probably occur in the dynamo region. The precise nature of these processes is not fully understood. The third group of frequency components, with a fundamental period of 13.9 days, has an unknown origin.