Book contents
- Frontmatter
- Contents
- Preface
- Acknowledgements
- Chapter 1 Introduction
- Chapter 2 Constructing time series in cyclostratigraphy
- Chapter 3 Spectral estimation
- Chapter 4 Additional methods of time-series analysis
- Chapter 5 Practical considerations
- Chapter 6 Environmental cycles recorded stratigraphically
- Appendix – published algorithms for time-series analysis
- References
- Index
Chapter 5 - Practical considerations
Published online by Cambridge University Press: 13 August 2009
- Frontmatter
- Contents
- Preface
- Acknowledgements
- Chapter 1 Introduction
- Chapter 2 Constructing time series in cyclostratigraphy
- Chapter 3 Spectral estimation
- Chapter 4 Additional methods of time-series analysis
- Chapter 5 Practical considerations
- Chapter 6 Environmental cycles recorded stratigraphically
- Appendix – published algorithms for time-series analysis
- References
- Index
Summary
Introduction
This chapter is concerned with various practical matters not considered elsewhere. Ideally the various types of environmental cyclicity discussed in Chapter 6 would produce cyclostratigraphic time series that faithfully reproduce the original temporal information. Unfortunately this is never the case because of a variety of factors discussed below. Cyclostratigraphic signals are invariably distorted records of the environmental variables due to modification of both the amplitude and frequency characteristics. It is important to be aware of these distortions, both so that interpretations are not pushed too far and so that, in some cases, remedial action can be taken. The distortions tend to reduce the chances of distinguishing spectral peaks from the spectral background. Additionally, sometimes the distortions introduce additional, often easily identifiable, spectral components that describe the non-sinusoidal shapes of time-series oscillations.
Once the distorting processes have been covered, the chapter continues with practical issues which are, at first sight, straightforward. These concern the number of data points needed for a meaningful time-series analysis, regularity and the interpretation of spectral peaks.
In order to illustrate the effects of distortion processes the same ‘test series’ has been used in several figures. This series consists of a 4 m sine wave plus a 0.75 m sine wave, both of constant amplitude, plus AR(1) noise (with the lag − 1 autocorrelation = 0.7). The figures allow comparison of depth records that are undistorted, with depth records that result from the various distortions.
- Type
- Chapter
- Information
- Time-Series Analysis and CyclostratigraphyExamining Stratigraphic Records of Environmental Cycles, pp. 129 - 160Publisher: Cambridge University PressPrint publication year: 2003
- 1
- Cited by