Extension of the astronomically calibrated (polarity) time scale to the Miocene/Pliocene boundary

Abstract

The early Pleistocene to late Pliocene astronormcally calibrated time scale of Shackleton et al. [1] and Hllgen [2] is extended to the Mlocene/Pllocene boundary This is done by correlating the detailed record of CaCO 3 cycles in the Trubl and the lower part of the overlying Narbone Formation (Rossello composite sechon, Sicily) to the astronomical record, using (1) mferred phase relations between these CaCO 3 cycles and the corresponding orhital cycles of precession and eccentricity, and (2) calibration points provided by the previously established astronomical cahbratlon of sapropeht~c layers which occur in the topmost part of the CaCO 3 record [2] This correlation allows all small-scale CaCO 3 cycles m the Rossello composite section to be coded after the correlative precession peak and to be dated astronomacally with an accuracy of l ky In combination with the available high-resohitmn magnetostratlgraphy, tt further allows the construction of an astronolrncally cahbrated Geomagnetic Polarity Time Scale for the major part of the Gilbert and Gauss Chrons This new time scale gives ages for the reversal boundaries of the Kaena (3 04-3 11 Ma) and Mammoth (3 22-3 33 Ma), the Gilbert/Gauss Chron boundary (3 58 Ma), and the reversal boundaries of the Cochiu (4 18-4.29 Ma), Numvak (4 48-4.62 Ma), Sldufjall (4 80-4 89 Ma) and Thvera (4 98-5 23 Ma) These ages are s~gmficantly older than the ages gwen by convenUonal t,me scales Age discrepancies gradually increase with increasing age the age of 5 32 Ma for the Miocene/Pllocene boundary, as defined by the base of the Truhi on Sicily, IS almost 0 5 m y (9 5%) older than the presently accepted age of 4 86 Ma [3] These considerable discrepancies with the conventional time scale can either be explained by an (consistent) error in the decay constants used in K/Ar radlometnc dating or, more hkely, by dlffuslonal loss of radlogemc argon