CYSTRATI - A COMPUTER-PROGRAM FOR SPECTRAL-ANALYSIS OF STRATIGRAPHIC SUCCESSIONS

Citation
E. Pardoiguzquiza et al., CYSTRATI - A COMPUTER-PROGRAM FOR SPECTRAL-ANALYSIS OF STRATIGRAPHIC SUCCESSIONS, Computers & geosciences, 20(4), 1994, pp. 511-584
Citations number
26
Categorie Soggetti
Mathematical Method, Physical Science","Geosciences, Interdisciplinary","Computer Science Interdisciplinary Applications
Journal title
ISSN journal
00983004
Volume
20
Issue
4
Year of publication
1994
Pages
511 - 584
Database
ISI
SICI code
0098-3004(1994)20:4<511:C-ACFS>2.0.ZU;2-J
Abstract
The spectral analysis of stratigraphic successions represents an objec tive approach for the detection of periodic components that give the s uccession its cyclical form, as for example, in the periodicity attrib uted to climatic changes and related to the Milankovitch cycles or wit h the decennial cyclicity of solar activity. We present the CYSTRATI p rogram, written in ANSI standard FORTRAN 77, which includes five of th e most widely established techniques for spectral analysis in the eart h sciences: Blackman-Tukey's classic approach; classic periodogram app roach; maximum entropy approach; Thompson multitaper approach for cont inuous variables; and finally Walsh's spectral analysis that lets us i nvestigate cyclicity rhythms for binary variables (for example the alt ernation of limestone-marl). Another important characteristic is the p rogram's versatility in providing different ways of processing experim ental data (windowing, tapering, smoothing) in order to better estimat e the power spectrum. Obviously any one-dimensional (1-D) succession c an be analyzed with CYSTRATI, but for stratigraphic successions the po ssibility of adopting a previous, more exhaustive treatment has been i ncluded, a treatment in which the lithology and the type of contact be tween the strata can be included. The estimated results of the power s pectrum in two different temporal successions will be provided, the fi rst corresponding to lacustrine varves and the second to the Wolfer nu mber of sunspots. such results can be used to check the program's impl ementation.