LASER 40AR/39AR AGES OF TEPHRA FROM INDIAN-OCEAN DEEP-SEA SEDIMENTS - TIE POINTS FOR THE ASTRONOMICAL AND GEOMAGNETIC POLARITY TIME SCALES

Tephra layers within deep-sea sediment cores are valuable stratigraphic markers which can be radiometrically dated with high precision by the laser 40Ar/39Ar technique. Two Neogene ash layers from Ocean Drilling Program (ODP) Site 758 have been dated. These layers are located within detailed oxygen isotope and paleomagnetic stratigraphies, which have previously been assigned ages based on the astronomically derived geomagnetic polarity time scale (ADGPTS) [7-10,28]. We use the radiometric dates from the ash layers to independently evaluate the ages assigned to oxygen isotope stage 19.1 and the paleomagnetic reversals marking the Brunhes-Matuyama boundary and the termination of the Nunivak event. Ash-D [20,21] yields an age of 800 +/- 20 ka, which is consistent with an ADGPTS age of 780 ka for the overlying Brunhes-Matuyama transition and with the astronomical age for oxygen isotope stage 19.1. Ash-I, located near the top of the Nunivak subchron, bears evidence of excess 40Ar. An error-weighted mean of five plateau ages gives an apparent age of 4.519 +/- .012 Ma, but an errorchron through a subset of the data suggests a possible eruption age of 4.43 +/- .03 Ma. The plateau age is in agreement with the ADGPTS [7-10,28] and the errorchron age is in agreement with the time scale of Baksi [29]. The laser 40Ar/39Ar ages of both ashes are significantly older than the ages from the radiometrically derived time scale (RDGPTS) of Mankinen and Dalrymple [3] which is primarily based upon a large dataset of K-Ar-dated. paleomagnetic poles. It is essential to reconcile the ADGPTS with the RDGPTS because radiometric age tie points are the best hope of extending the ADGPTS to times before 5 Ma. An analysis of the expected and actual variance for the astronomical and Mankinen and Dalrymple [3] time scales suggests that only a handful of low K-Ar ages may have biased the original GPTS. It is not necessary to invoke general systematic errors in the Ar dating system to explain the discrepancies between the RDGPTS and ADGPTS.