Oxygen isotope homogeneity and trace element variations in glass within 250-79 ka Central Plateau Member rhyolite lavas from the Yellowstone Volcanic System

The 250-79 ka Central Plateau Member CPM) rhyolites are represented mostly by large volume lava flows with less than 10% crystals and are the youngest eruptive products inside of the 620 ka Yellowstone Caldera, formed after eruption of Lava Creek Tuff (LCT). These flows are low-δ¹⁸O requiring tens of percent of shallowly remelted hydrothermally-altered material and provide evidence into how large silicic magma systems evolve before and after major caldera-forming cycles.

We have developed a technique to directly analyze small micro-domains of rhyolite glass for precise (better than 0.1 ‰) δ¹⁸O determination coupled with D/H, [H₂O], trace elements, and Pb isotopes. We present evidence for striking δ¹⁸O homogeneity (4.48 ± 0.12 ‰, 2 standard deviations of all analyses) both across small (1 km³) and large (up to 70 km³) flows and between flows erupted over almost 200 ka. D/H analyses in the glasses are highly degassed and are not affected by secondary hydration (H₂O = 0.05 to 0.28 wt. %, δD = -99 to -171 ‰). Trace element analyses show broad temporal compositional evolution consistent with increasing feldspar fractionation over time: the oldest Scaup Lake flow contains 8 ppm Sr, 330 ppm Rb, and 270 ppm Ba to the youngest Pitchstone Plateau flow contains 0.7-1.9 ppm Sr, 180-250 ppm Rb, 18-25 ppm Ba. We also show small but recognizable trace element heterogeneity within lava flows unrelated to long-term geochemical trends.

Homogeneity of oxygen in melt from individual lava flows across the LCT caldera is consistent with convective homogenization of a large magma body generated by remelting of post and pre LCT hydrothermally-altered and likely highly variable, low- δ¹⁸O rocks. In combination with trace element data we show that the CPM rhyolites have also undergone a broad, caldera-wide differentiation trend with no evidence for significant recharge. We are reconciling diverse geochemical data streams to develop a comprehensive petrologic model for the evolution of the most recent products of Yellowstone volcanism.