Abstract
THIS subject has been discussed by Archie Blake, who states that classical methods for detecting and measuring periodicity have suffered from two serious defects1. First, the Schuster periodogram does not provide adequately for non-sinusoidal types of variation. Secondly, the smoothness encountered in almost all time series and many other types of data introduces a spurious appearance of periodicity which vitiates the test of significance unless proper allowance is made for the smoothness. New statistics designed to detect effects not contemplated in the Schuster periodogram are being tried in a study of the series of aftershocks of the Helena Montana earthquakes of October 1935. This work is being done by use of punched cards, which greatly alleviate the labour of the computation.
Article PDF
Rights and permissions
About this article
Cite this article
SEISMIC PERIODICITY. Nature 149, 308 (1942). https://doi.org/10.1038/149308a0
Issue Date:
DOI: https://doi.org/10.1038/149308a0