Geomagnetic reversal history

ABSTRACT The history of geomagnetic polarity reversals in the Cenozoic and Late Mesozoic is well known since the Late Jurassic (Oxfordian). A continuous record of polarity has been derived for this time interval from the interpretation of oceanic magnetic anomalies. Most of the polarity chrons in this oceanic record have been verified and dated in coordinated magnetostratigraphic and biostratigraphic studies. This has led to the generation of progressively refined and improved geomagnetic reversal time‐scales that provide a framework for absolute dating of palaeontological zonations. By serving as a basis for statistical analysis of reversal frequency they provide information relevant to processes in the Earth's core. The rate of reversals since the Late Cretaceous shows a steady increase on which a cyclical variation appears to be superposed. A stochastic model for reversals predicts a Poisson distribution of polarity interval lengths. The polarity time scales contain many fewer short (± 50 kyr) polarity chrons than a Poisson distribution, and it has been suggested that a gamma renewal process with index greater than unity is a more appropriate statistical model. The statistical arguments give no convincing reason for abandoning the model and other, physical reasons must be sought to explain the incompleteness of the reversal record. The discovery and verification of short chrons in the oceanic record may best be investigated by deep‐tow magnetometer surveys. The reversal history before the Late Jurassic is not well known. Magnetostratigraphy in coeval Early Jurassic sections has not given correlatable records and it has not been possible to compile a definitive polarity sequence. Evaluation of geomagnetic polarity history for the Early Mesozoic and the Palaeozoic will require unambiguous magnetostratigraphy in well‐dated sections where verification of the polarity pattern is possible at the fossil zone or stage level.