Premium
Random fluctuations, persistence, and quasi‐persistence in geophysical and cosmical periodicities: A sequel
Author(s) -
Forbush S. E.,
Duggal S. P.,
Pomerantz M. A.,
Tsao C. H.
Publication year - 1982
Publication title -
reviews of geophysics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 8.087
H-Index - 156
eISSN - 1944-9208
pISSN - 8755-1209
DOI - 10.1029/rg020i004p00971
Subject(s) - persistence (discontinuity) , statistical hypothesis testing , statistical analysis , statistical physics , statistics , variation (astronomy) , geophysics , mathematics , geology , physics , astrophysics , geotechnical engineering
The method of Chree analysis (superposed epochs) has been used in many disciplines, including geophysics, astrophysics, and solar‐terrestrial relationships. However, procedures to test the statistical significance of the results obtained thereby have not been available heretofore. Claims for statistical reality of average variations (from Chree analyses or otherwise) are unacceptable without testing the assumption that deviations from the average variation are random and sequentially independent. In many phenomena this assumption is not valid. One objective of this paper is to expand established methods for the analysis of variance in order to provide a statistical procedure for testing the Chree analysis result from data with nonrandom deviations from average. In addition, the statistical method developed by Bartels (1935) to determine the quasi‐persistency of deviations from average signals in the form of sine waves is also applied to the Chree analysis problem. The two alternative procedures for evaluating the significance of Chree analysis results (or variations otherwise obtained) are then compared to determine the circumstances under which one of them may be preferable.