40Ar/39Ar ages of tephras intercalated in astronomically tuned Neogene sedimentary sequences in the eastern Mediterranean

We present new 40Ar/39Ar data for sanidine and biotite derived from volcanic ash layers that are intercalated in Pliocene and late Miocene astronomically dated sequences in the Mediterranean with the aim to solve existing inconsistencies in the intercalibration between the two independent absolute dating methods. 40Ar/39Ar sanidine ages are systematically younger by 0.7-2.3% than the astronomical ages for the same ash layers. The significance of the discrepancy disappears except for the upper Ptolemais ashes, which reveal the largest difference, if an improved full error propagation method is applied to calculate the absolute error in the 40Ar/39Ar ages. The total variance is dominated by that of the activity of the decay of K-40 to 40Ar ( similar to 70%) and that the amount of radiogenic 40Ar(p) in the primary standard GA1550 biotite ( similar to 15%). If the 40Ar/39Ar ages are calculated relative to an astronomically dated standard, the influence of these parameters is greatly reduced, resulting in a more reliable age and in a significant reduction of the error in 40Ar/39Ar dating. Astronomically calibrated ages for Taylor Creek Rhyolite (TCR) and Fish Canyon Tuff (FCT) sanidine are 28.53 +/- 0.02 and 28.21 +/- 0.04 Ma (+/- S.E.), respectively, if we start from the more reliable results of the Cretan A l ash layer. The most likely explanation for the large discrepancy found for the younger Ptolemais ash layers (equivalent to FCT of 28.61 Ma) is an error in the tuning of this part of the sequence. (C) 2004 Elsevier B.V. All rights reserved. We present new 40Ar/39Ar data for sanidine and biotite derived from volcanic ash layers that are intercalated in Pliocene and late Miocene astronomically dated sequences in the Mediterranean with the aim to solve existing inconsistencies in the intercalibration between the two independent absolute dating methods. 40Ar/39Ar sanidine ages are systematically younger by 0.7-2.3% than the astronomical ages for the same ash layers. The significance of the discrepancy disappears except for the upper Ptolemais ashes, which reveal the largest difference, if an improved full error propagation method is applied to calculate the absolute error in the 40Ar/39Ar ages. The total variance is dominated by that of the activity of the decay of K-40 to 40Ar ( similar to 70%) and that the amount of radiogenic 40Ar(p) in the primary standard GA1550 biotite ( similar to 15%). If the 40Ar/39Ar ages are calculated relative to an astronomically dated standard, the influence of these parameters is greatly reduced, resulting in a more reliable age and in a significant reduction of the error in 40Ar/39Ar dating. Astronomically calibrated ages for Taylor Creek Rhyolite (TCR) and Fish Canyon Tuff (FCT) sanidine are 28.53 +/- 0.02 and 28.21 +/- 0.04 Ma (+/- S.E.), respectively, if we start from the more reliable results of the Cretan A l ash layer. The most likely explanation for the large discrepancy found for the younger Ptolemais ash layers (equivalent to FCT of 28.61 Ma) is an error in the tuning of this part of the sequence. (C) 2004 Elsevier B.V. All rights reserved.