Single-crystal 40Ar/39Ar incremental heating reveals bimodal sanidine ages in the Bishop Tuff

Friday, 18 December 2015: 14:55
310 (Moscone South)
Nathan L Andersen, Brian R Jicha and Bradley S Singer, University of Wisconsin Madison, Madison, WI, United States
The 650 km3 Bishop Tuff (BT) is among the most studied volcanic deposits because it is an extensive marker bed deposited just after the Matuyama-Brunhes boundary. Reconstructions of the vast BT magma reservoir from which high-silica rhyolite erupted have long influenced thinking about how large silicic magma systems are assembled, crystallized, and mixed. Yet, the longevity of the high silica rhyolitic melt and exact timing of the eruption remain controversial due to recent conflicting 40Ar/39Ar sanidine vs. SIMS and ID-TIMS U-Pb zircon dates.

We have undertaken 21 40Ar/39Ar incremental heating ages on 2 mm BT sanidine crystals from pumice in 3 widely separated outcrops of early-erupted fall and flow units. Plateau ages yield a bimodal distribution: a younger group has a mean of 766 ka and an older group gives a range between 772 and 782 ka. The younger population is concordant with the youngest ID-TIMS and SIMS U-Pb zircon ages recently published, as well as the astronomical age of BT in marine sediment. Of 21 crystals, 17 yield older, non-plateau, steps likely affected by excess Ar that would bias traditional 40Ar/39Ar total crystal fusion ages.

The small spread in older sanidine ages, together with 25+ kyr of pre-eruptive zircon growth, suggest that the older sanidines are not partially outgassed xenocrysts. A bimodal 40Ar/39Ar age distribution implies that some fraction of rhyolitic melt cooled below the Ar closure temperature at least 10 ky prior to eruption. We propose that rapid "thawing" of a crystalline mush layer released older crystals into rhyolitic melt from which sanidine also nucleated and grew immediately prior to the eruption. High precision 40Ar/39Ar dating can thus provide essential information on thermo-physical processes at the millenial time scale that are critical to interpreting U-Pb zircon age distributions that are complicated by large uncertainties associated with zircon-melt U-Th systematics.