

Bayesian periodic signal detection applied to INTCAL98 data.
Abstract
A Bayesian multiple-frequency model has been developed for spectral analysis of data with unknown correlated noise. A description of the model is given and the method is applied to decadal atmospheric INTCAL98 delta (super 14) C data. The noise of the INTCAL98 data is found to be red, and there seems to be no support for continuous harmonic frequencies in the data.
Keywords
atmosphere ;Bayesian analysis;C 14;C 14 C 12;calibration ;carbon ;Cenozoic ;concentration ;equations ;frequency ;Holocene ;INTCAL98 ;isotopes ;Markov chain analysis;models ;Monte Carlo analysis;noise ;Pleistocene ;probability ;Quaternary ;radioactive isotopes;signals ;stable isotopes;statistical analysis;statistical distribution;upper Pleistocene